What Do We Mean by Tantrism?
The beginnings of the Hindu Tantric traditions are all the more difficult to find in that Tantrism is a protean phenomenon, so complex and elusive that it is practically impossible to define it or, at least, to agree on its definition. Is not this difficulty due to the fact that we see and try to define an entity that does not really exist as such? Even if we do not go that far, even if we do not endorse H.V.
Guenther’s remark that Tantrism is “probably one of the haziest notions and misconceptions the Western mind has evolved,”2
the fact remains that Tantrism is,
to a large extent, “a category of discourse in the West,” and not, strictly speaking, an Indian one. As a category, Tantrism is not—or at any rate was not until our days—an entity in the minds of those inside. It is a category in the minds of observers from outside. To use the fashionable jargon of today: it is an etic, not an emic, entity.
The term Tantrism was coined by Western Indologists of the latter part of
the nineteenth century whose knowledge of India was limited and who could not realize the real nature, let alone the extent, of the Tantric phenomenon. They believed that the practices and notions they discovered in Hindu and Buddhist texts named Tantra (hence Tantrism) were something very particular, exceptional, and limited, contrasting sharply with the general, respectable, field of Indian philosophy and religion, a particular domain one could easily circumscribe.
But with the progress of studies in these fields one came to realize that, far from being a limited phenomenon, Tantrism was in fact something vast, diffuse, diverse and very difficult to define satisfactorily.
Mircea Eliade was perhaps the first to point it out, when he wrote in a book published in 1948 that, after the fifth century .., Tantrism became a pervasive Indian “fashion” (une “mode” pan-indienne).3 Neither in traditional India nor in Sanskrit texts is there a term for Tantrism; no description or definition of such a category is to be found anywhere. We know also that, more often than not, Tantric texts are not called Tantra.
As evidence that Tantrism was not considered a particular philosophical system, one may see that M¯adhava’s Sarvadar´sanasamgraha, ¸ a fourteenth century text, does not mention Tantra as one of the fifteen dar´sanas (schools of worship) it describes, although this work dates to a time usually considered as that of the fullest expansion of Tantric notions and practices.
There are, of course, Tantric elements in M¯adhava’s description of the Saiva ´ dar´sana where he quotes from such Tantric authors as Utpaladeva or Abhinavagupta, but Tantra as such is not mentioned. P.V. Kane explains this by saying that M¯adhava deliberately ignored Tantra because it was too scandalous. But it is more likely that it was, by that time, so pervasive that it was not regarded as being a distinct system.
The usual reference to the Indian use of the term t¯antrika derives from
Kull¯uka Bhat¸ta’s formula when commenting on the ¸ Manavadharma´sastra 2.1, where he juxtaposes vaidika/tantrika as two forms of revelation (´srutis ca dvidvidh¯a vaidikı tantrikıca) and, consequently, two different approaches to the ultimate reality (the first based formally on the Veda and the Brahmanic tradition and the second on other texts). The distinction has remained a basic one throughout Indian thought, but without a particular category of “Tantrism” evolving.
We may note here the use by Kulluka of the term ´sruti. Even outside Tantric circles, apparently, the Tantric tradition could be considered as ´sruti, that is, as a revelation valid in its own sphere. In fact, Kulluka’s formula shows on the one hand that, even though there is no inside definition of Tantrism, Tantrism was at least perceived by Indians outside it as different from the Vedic tradition. It evidently was similarly perceived by those inside who deprecated Vedic rites and notions. On the other hand, the quotation tends to show that the vaidika/tantrika relationship was not a clear-cut one since both could be called ´sruti. Such ambiguity, in fact, goes very far because Vedic and Tantric traditions, as time passed, tended to permeate each other in ritual, in concepts, and in scriptural references.
Not only are elements from the Atharva Veda important in some local Tantric traditions (in Orissa, for instance), but many Tantric authors quote freely from ´sruti. The assimilation went so far that, in Kashmir, some Va¸ısnava T¯antrikas of ¸ the P¯a˜ncaratra declared their scriptures to be the Vedic ´s¯akh¯a (school), the Ek¯anaya´s¯akh¯a.
Concerning the Indian textual use of the term tantrika, we should also take note of the fact that, in Saiva Tantric texts, ´ tantrika often is used instead of kaula to refer to the more exoteric texts and practices and as a way to distinguish from the esoteric kaula ones. Thus, those texts and practices called Tantric are the less Tantric ones. India, it appears, far from providing us with a definition of Tantrism as something specific, rather accumulates evidence showing the interpenetration, in Hindu thought and practice, of Tantric and non-Tantric elements.
A number of traits have been listed by authors writing on Tantrism as being constituent elements. Teun Goudriaan, for instance, in his Hindu Tantrism, lists eighteen such traits as “some constituents of Tantrism (in its wider sense).”4 The
trouble, however, with such lists is, first of all, that there is no consensus among scholars about these elements and, second, that there are no groups or texts usually considered Tantric where all these elements are to be found; also, some, if not most, of them can be found in non-Tantric contexts. This is not surprising because “Tantrism in its wider sense” is a hazy and ill-defined sort of notion; it can cover, in fact, so vast a field as to include almost all of Hinduism. Let us, however, examine some of these traits, limiting ourselves to those I believe to be the most obvious.
The first aspect to be examined, that is, a particular Tantric ideology, I
believe is important. I differ with those who consider Tantrism to be, in Jean Filliozat’s words, “merely the ritual and technical aspect of Hinduism.”5 This view,
however, is not to be dismissed too hastily since ritual, when the technical aspect is added to it, can go a long way toward characterizing Tantrism if you take it as a general Hindu phenomenon. Indeed, ritual may well provide one of the most
practical, but surely minimal, overall definitions of Tantrism. I shall refer again to this subject later on.
The ideological aspect of the Tantric vision is the cosmos as permeated by
power (or powers), a vision wherein energy (´sakti) is both cosmic and human and where microcosm and macrocosm correspond and interact. The ideology is important because it explains such Tantric features as the concept and practice of kun¸dalin¯ı, ¸ as well as a number of yogic and ritual practices for the use and control of that power. It also explains some aspects of the speculations and practices concerning the power of the word (v¯ac), e
Such an ideology is evident in the Bharirava Tantras, in those of K¯al¯ı, in such systems as the Krama and the Trika (all of the Saiva or ´ S¯´akta texts), in the esoteric Buddhist Yog¯anuttaratantras and in the Sahajiy¯a traditions. It is subdued, toned down, in other traditions normally considered Tantric, such as the P¯a˜ncaratra, or Agamic ¯ Saivasid- ´ dh¯anta. And, you would hardly find it among some dualist Saiva authors ´ (Sadyojyoti, for instance) even though the mantras and rituals used by these adepts are Tantric.
On the other hand, micro-macrocosmic correspondences are found in ancient, pre-Tantric texts; for example, they are fundamental in the Upanisads. The ¸ magical use of power is apparent in the Atharvaveda too and, later on, the Tantric vision permeated the Pur¯anas and seeped into most of Hinduism. We must add that ¸ the ideology of power, with its aspect of violence and transgression, is essential to the cults of the feminine aspect of the divinity, cults that cannot always be considered as Tantric.
In south India, for instance, the cults of local goddesses are surely autochthonous Dravidian and originally pre-Tantric. The ancient Indian practices of tapas (internal heat) and mastery of sexual energy (v¯ırya) for gaining supernatural powers also are examples of controlled uses of the power that are not Tantric.
In all these non-Tantric domains, there are elements identical or akin to
those constituting the Tantric vision. We can say, therefore, that in the domain of ideology or doctrine, we find the same situation as in the case of the other Tantric traits. Characteristics found clearly and fully in a few groups or in some texts only are found in a wide area. Furthermore, the origin or the seeds of many traits, ideological or otherwise, can be traced back to ancient, pre-Tantric, times.
Another element generally considered characteristic or constituent of Tantrism is the use of means pertaining to this world for supramundane ends, be it mukti (liberation) or the lesser rewards or enjoyments classified as bhoga. There is the attempt in Tantric traditions to achieve liberation and to gain supernatural powers, not by renunciation of all worldly desires or pleasures, but, to use Madeleine Biardeau’s words, “by harnessing desire—k¯ama [desire] in all the meanings of that word and with all its related values—to the service of liberation,”6 a liberation that is usually j¯ıvanmukti (liberation while living), a transcendental condition of unity with the deity—total freedom from the world, but also triumphal plentitude and demiurgic power.
But liberation in a Tantric context is not necessarily jıvanmukti. Even in such a completely Tantric work as Abhinavagupta’s Tantraloka, the best and highest adept, who benefits from the most intense grace of Siva ( ´ tıvrasaktipata), is instantly liberated and dies: a condition considered higher than j¯ıvanmukti. The typical Tantric jıvanmukta, totally free of a world he dominates and transcends, is to be found in some Tantras only—for example, in the Bhairava or KalıTantras, in Sahajiya Vais¸navism or in Buddhism too—that can be viewed as “hard core” ¸ Tantrism. In the more staid Saiddhantika Agamas where the term ¯ jıvanmukti seldom occurs, the liberated adept acquires ´sivatva, the condition of Siva, a condition of similarity (s¯amant¯a) with Siva, not one of total fusion ( ´ ekatva). This permits the liberated soul to go on loving God. It is evidently still more so in the Pancaratra, where devotion (bhakti) is essential.
Since I mention bhakti, I may note here that, gaining liberation while active in this world, being in this world but not of it, being entirely dedicated to God, is the basic teaching of bhakti from the Bhagavad-Gıta onward. Since, however, the love of God and the essential role of God’s grace to gain liberation are insisted upon in such Tantric works as those of Abhinavagupta, where does bhakti end and Tantra begin? There is a problematical relationship between Tantrism and bhakti.
A particular Tantric way of making use of this world for supramundane ends
is the ritual and soteriological use of things that are normally forbidden, that is, the transgression of norms. The main reason for this antinomian behavior appears to be the wish, by so doing, to participate in the dark, chaotic, undisciplined, and very powerful forces that are normally repressed and kept outside the pure, orderly, circumscribed world of the Brahmin. This wish, incidentally, implies a belief in a world pervaded by power, a power supposedly at its utmost in that outside world.
Such transgressive practices include the transgressive ritual use of sex.
The use of sex is not found in all Tantric traditions. It is not prevalent, but present nonetheless, in the Saiva and ´ Sakta groups that have a Kapalika origin or background and that have kept, if only symbolically, the Kapalika culture of the cremation ground with its cult of the Yoginıs and its erotico-mystic rites and notions. It is also found in Sahajiya circles, Hindu and Buddhist. But, all this is conspicuously absent from the less intensely Tantric traditions, whether Saiva or ´ Vaisnava. Transgression is characteristic of “hard core” Tantrism only.
On the ¸ other hand, transgression is a universal category of human behavior. In India it is older than Tantrism, as proved, for instance, by the Pasupatas and Lakulas. Erotic rites and sexo-yogic practices surely antedate Tantra. Here again, we see elements either not found in all Tantric groups or texts, or that exist outside Tantrism and have existed before it. For instance, the conception of the body as a structured receptacle of power and animated by that power and the somato-cosmic vision upon which these practices are based are certainly pre-Tantric or extra-Tantric.
The same thing can be said about most, if not of all, of the other elements considered characteristic or constituent of Tantrism. The ubiquitous use of mantras, for instance, together with all the notions concerning the power of the word (v¯ac) and with the relevant practices (nyasa, japa, mantra-sadhana) is so typically Tantric that mantra´sastra is often taken as synonymous with tantrasastra; however, Tantric mantras are used in non-Tantric rites. Some Vaisnana Samhitas, the ritual ¸ of which is Tantric, nevertheless, consider the Vedic mantras as higher than the Tantric ones. In a similar fashion ritual diagrams (man¸dalas, yantras, cakras ¸ ) or ritual gestures (mudras) are to be found variously used in and outside Tantrism. If one looks at the Tantric ritual (puja or dıksa), one would notice some of the constitiuent elements as deriving from groups outside Tantrism also. The same applies to another element considered typically Tantric, namely the polarization of the godhead into a male pole (usually higher, but inactive) and a female one (´sakti), which is active but theologically lower except in some Sakta traditions.
Such polarization is not stressed equally everywhere. The role of ´sakti is limited not only in Vaisnava Samhitas, but in the Siddhanta ¸ Saivagamas. There are, ´ futhermore, Saiva pantheons that are either entirely male or entirely female. ´ It is thus very difficult to gather traits that are both typically Tantric and found in most Tantric traditions, but not found outside these traditions when we limit ourselves to Hinduism. The difficulty becomes even greater—indeed, it becomes an impossibility—if we wish to include Tantric Buddhism as it developed in India and spread to China, Tibet, or Japan.
We could try to bypass the difficulty by choosing from among those constituent elements only a few that, when present in a text, in the practice or doctrine of a given group, would suffice for us to declare that text or group as Tantric. But, which elements ought we to choose—ideological, ritual, or practical ones? Should we limit those elements to observances (vrata) or cult ritual (Tantric p¯uj¯a)? Complicating the picture is that Tantric ritual is not always exclusive of a Vedic practice. Several texts or authors prescribe or admit both types of ritual. A Vedic public behavior may hide a Tantric domestic or secret practice. Furthermore, within the same tradition there are levels of esotericism and exclusivism, there are progressive, ascending levels of specificity (uttarottaravaisi¸styam ¸ ) and of “Tantricity.” Usually the more specific and esoteric the level of the Tantra, the more Tantric it is. It is clear that there are degrees in Tantrism.
A number of other elements could be adduced to show the uncertainty of
the criteria we can use to define Tantrism, the diversity within Tantric traditions and the problematic nature of the relationship of Tantrism with non-Tantric, “orthodox” Hinduism. On this last point, the judgment passed by each group on the other goes from utter condemnation to the admission of the validity of the other’s scriptures within their own field and for their particular purpose.
The
Sarvagamapram¯anyavada, ¸ for instance, was upheld by such authors as Yamunacarya. This being so, how can we concur on the definition of Tantrism?
Like Hinduism, “Tantrism” is made up of a number of groups, traditions, and texts sharing some common elements, especially ritual ones, and having some common beliefs and notions; the total of these elements or beliefs somehow
differentiate Tantric from non-Tantric Hinduism. Tantrism, however, includes practices or beliefs found in non-Tantric Hinduism too. Hence, we are faced with the uncertainty of the limits of Tantrism, its elusive nature, and if we take it comprehensively, its apparent pervasiveness. If it is pervasive, does it not loose its identity as something specific? If, therefore, we wish to keep the notion of Tantrism, we must take it as something specific and sufficiently distinguishable in spite of its uncertain limits. If so, we can consider as Tantric those groups or texts only where the main Tantric constituent elements previously mentioned are found.
This would limit the category of Tantrism to a few groups of people or of
texts, mainly the Saiva Sakta traditions with a Kapalika background, some of the Natha, the Sahajiya Vaisnavas, and, of course, the obvious Tantric forms of Mahayana Buddhism.
In such “hard core” Tantrism, we find a system of observances (often transgressive in nature) that are given meaning by a more or less poweroriented vision of man and the cosmos, a system where power is manipulated, where micro-macrocosmic correspondences play an essential role.
Also, there is usually a high degree of esotericism (the higher, the more esoteric, the more “Tantric”) together with a particular type of pantheon (not necessarily sexually differentiated however), and a particular and very developed type of ritual. Outside these qualifications, there may exist a varying number and proportion of Tantric traits, but not Tantrism as such.
With such a definition, could we include the Pancaratra? Possibly we could include the Pancaratra of some of the older Samhitas and the ¸ Laksmitantra, of course, since it is heavily influenced by Saivism. Surely, we would not include the ´ Kashmiri Pancaratrins who insisted they were a Vedic ´sakha, nor the more recent Srıvaisnavas. But how much of Saivism should we include apart from the Bhaira-vagamas? Is the Agamic ¯ Saivasiddhanta really Tantric?
Yes, it is, but mostly owing ´ to its ritual aspect and because it is the Sam¯anya´s¯astra of the followers of the Bhairav¯agamas. What about the modern “Vedantized” Sr¯ ´ ıvidy¯a of South India which traces its guruparampara ¸ to Sankaracarya and whose Tantric conceptions are so toned down in order to fit into orthodox Brahmin circles that it is hardly Tantric anymore?
Another approach to Tantrism and to its definition might be to stress its
ritual aspect without omitting entirely the ideological side, but subordinating it to ritual. One would underline the particular and proliferating nature of the ritual, and its conjunction with speculations and practices concerning the power of the word (mantras, etc.).
Ritual also involves the manipulation of power and a pattern that combines the assertion of the identity or fusion of worshiper and worshiped (nadevo devam arcayet) together with an ensemble of offerings and obeisances (upacara) to the deity (or Buddhist entity), who is treated as an honored guest.
It is important to stress that such rituals entail an intense inner participation—body, mind, and word—of the worshiper in the ritual he carries out. The main merit of an essentially ritual approach to the definition of Tantrism is that it applies equally well to Hindu and to Buddhist Tantrism and that it can apply not only to Indian Himalayan-Tibetan, but to Chinese Buddhism also. Its drawback is that ritual is not only that of the cult.
It includes other aspects of the Tantric adept’s life and observances. Furthermore, Tantric and non-Tantric rites are often not only performed by the same person (in different circumstances), but also sometimes during the same ritual. We cannot, therefore, content ourselves with the ritual approach to Tantrism.
But are we not facing a sea of troubles simply because we want to define
something that does not exist except in our minds? Having coined the term
Tantrism, we want it to mean something specific. Those in India or elsewhere, whose observances were Tantric, never used the term Tantrism nor did they give the term Tantric the same meaning as we do. Some would not even describe
themselves as t¯antrika. They simply followed the beliefs and practices that were current in their times in their own social groups. What were these? Simply they were the various forms taken by Hinduism and Buddhism as they evolved over the course of centuries, mainly under the influence of the Indian, Tibetan, or Chinese spheres and/or by a process of internal transformation.
Tantrism, thus, would be quite simply the various forms taken over the
course of time by large sections of Hinduism or Buddhism. Depending upon the background, the origins, and the local influences, the evolution was more or less marked by a rejection of the orthodox Vedic rules and notions; it included more or less local autochthonous cults and beliefs, local religious behaviors, and magical and/or other practices. All of this resulted in the more or less “Tantric” character of the different groups concerned. But, whatever the case, the variety of Tantra that baffles us might very well be nothing more than some of the ways in which
Hinduism or Buddhism were actually understood, believed, and practiced by
Indian, Tibetan, and Chinese practicioners during the last two millennia. These various religious forms we may decide to call Tantric in order to differentiate them from older or different forms of the same religions, but we ought not try to set them apart as a particular religious entity that we choose to call Tantrism, an entity that probably never existed as such.
We would thus be rid of the difficult notion of Tantrism. This would be very convenient! But is it possible? I am not sure. I fear we still have to toil to find a solution to the problem of Tantrism.
