Vol. 9 (2008)
Possession and self-possession: spirit healing, tantric meditation and āveśa
While much recent scholarship on Tantra has tended to distance it from the world of pragmatic ritual healing, mediumship and spirit-possession, and treat it primarily as an elevated mode of pursuit of enlightened consciousness, Frederick Smith’s recent book The Self Possessed is the most comprehensive presentation so far of the reverse position, that all of these modes of interaction with the divine, whether to do with healing, sorcery or spiritual liberation, share common assumptions and a common idiom. Here, and in other recent writing, Smith presents the idea of entry, pervasion or possession (āveśa) as a fundamental trope in Indic thought, encompassing everything from conception (seen as the individual jīva taking over possession of the embryo) to Tantric ritual, the temporary occupation of another body and malevolent spirit-attack. If this is true, however, does this suggest that we are applying the wrong set of categories to understand Indic and perhaps also other religious traditions? This article sketches an alternative mode of looking at the field of “possession,” broadly defined, and explores some of its implications.
The terminology in English regarding how people relate to spirit-entities is confused and contradictory. The terms possession, trance, mediumship, and shamanism have individual histories and overlapping areas of application. Thus possession in English, in the sense of being controlled by a demon or spirit, goes back to the late 16th century; its root meaning, from Old French possesser and Latin possidere, is ‘to be occupied, to be held’. Trance originated at around the same time, deriving from Old French transir, to depart or to die, and originally referred to a state of extreme apprehension or dread, or an unconscious and stupefied condition. Mediumship, as one might expect, derives from the context of mid-19th century Spiritualism. Shamanism, as is well known, derives from the Siberian (Tungusic) term shaman, for a particular type of religious practitioner, and entered English in the late 18th century through Russian and German.
The origins of the terms reflect quite different senses of the nature of the various phenomena at issue (control by an external spirit, a state of unconsciousness, a mode of communication by a spirit, the activity of a particular kind of practitioner). Since the 1960s, however, the terms spirit-possession, spirit-mediumship and shamanism have all been used extensively within social and cultural anthropology, religious studies and related academic disciplines, to refer to categories of religious practice that were seen as occurring in a wide range of different cultures. Perhaps unsurprisingly, the usages of different scholars have been far from consistent. Shamanism is a particularly problematic term, as I have discovered through personal experience, and the question of whether shamanism and spirit-mediumship should be regarded as mutually exclusive categories has been bitterly contested by rival groups of scholars, with very little in the way of resolution. <1>
The same is true of the question of whether Tantric meditation and yogic practice on the one hand, and shamanism on the other, have anything in common. Here the late Mircea Eliade caused considerable confusion through two well-known books, on shamanism and yoga respectively. In Shamanism: Archaic Techniques of Ecstasy <2> he inflected the concept of shamanism towards the image of the heroic male shaman, fully in control, on a heavenly-journey to a sky god, and away from matters such as spirit-mediumship and spirit-possession, which he saw as inferior and feminine modes of practice. Communication with earthly or underground spirits was similarly an indication of the decline and decadence of a tradition more properly directed towards a heavenly deity.
The late Dan Noel argued plausibly in his Soul of Shamanism <3> that Eliade’s views were at least in part a projection of his personal commitment to right-wing Christian political groups in pre-war Romania, with their implicitly patriarchal attitudes. They have certainly had a major influence on how shamanism is seen, and also on how it has been practised by contemporary Western shamans. The dominant contemporary style of Western neo-shamanism, Michael Harner’s ‘core shamanism’ and its various derivatives <4>, is in many ways shamanism according to the Eliade book, though this is certainly not to suggest that Harner shares Eliade’s politics.
In Yoga: Immortality and Freedom <5> Eliade emphasised the contrast between the shamanic voyage, as he defined it, and what he saw as the essentially inward-looking process of yoga - going as far as to invent a new word, “entasy”, to emphasise that there was nothing ecstatic about yogic practice. It all went inwards, not outwards. In fact, Eliade did not deny the possibility of a shamanic origin for yoga, but he certainly wanted to draw as sharp a line as possible between yoga in its developed form and any putative shamanic origins: ‘As a developed spiritual technique (we are not discussing its possible “origins”), Yoga cannot possibly be confused with shamanism or classed among the techniques of ecstasy’ <6>.
This picture of yoga as purely inward looking has been heavily criticised in recent years, at least in relation to the period before the establishment of the classical “Vedāntic” perspective on yoga associated with Śaṃkara (probably 8th or 9th century CE). Some of the more significant arguments come from Ian Whicher <7> and David Gordon White (2006). White’s 2006 article considers the employment of yoga by warriors in the Mahābhārata to travel out of their bodies to heaven at death, and the well-known use of yoga to take over the bodies of other people or animals, a matter to which I shall return later. White uses these and similar examples to argue that yoga, at least in earlier times, was conceived of not as a purely inward-directed practice, based on a ‘closed’ model of the human body, but as one where the body is essentially ‘open’. He argues that this
As for the distinctions between shamanism, spirit-mediumship, and spirit-possession, leaving aside the possible ideological underpinnings of Eliade’s perspective, the problem is that these distinctions appear to make good sense in some ethnographic contexts but hardly any in others. The by now quite well-studied “shamans” of the Nepal highlands - mostly but by no means exclusively male - share many of the classic features of the Eliadean shaman, including the climbing of heavenly poles on which they are initiated, but they freely incorporate spirits into their own bodies for healing purposes, and in other ways behave much more like a classic spirit-medium <10>. Examples could be multiplied, but what all this amounts to is that there is no such thing as a unitary or ‘authentic’ form of shamanism. At most shamanism is what Rodney Needham called a polythetic class <11>, in which items share a ‘family resemblance’, with each possessing only some of a set of partly shared features. It is probably more useful to speak, as Jane Monnig Atkinson has suggested, of shamanisms, not shamanism <12>.
When it comes to Tantra, a messy and problematic enough term in its own right <13>, it becomes much harder to draw the line with such modes of practice, particularly given the processes of deity-yoga, where the Tantric practitioner interacts with and in many cases actively identifies with the Tantric deity, often with the intention of bringing about healing or other this-worldly results for himself or others <14>. Is this really radically different from shamanism or spirit possession? A series of eminent Tibetan Buddhist scholars have nevertheless clearly been very uncomfortable, to say the least, with the idea that the elite realms of the Vajrayāna might be associated with anything as grubby and primitive as a tribal shaman.
A variety of other scholars, myself included, have meanwhile argued that we are dealing with a wide range of phenomena with many features in common, and that the field within which they occur is better approached as a continuum than as a number of discrete boxes hermetically sealed from each other. Here I might note, for example, the work of Gavin Flood and Rich Freeman on teyyam possession cults and Tantric practice in Kerala <15>. Flood’s and Freeman’s work is of particular interest because, while a number of authors have suggested shamanic origins for yoga or tantra, Flood and Freeman argue for the considerably more interesting - and, I suspect, historically more plausible - position of mutual influence between developed Tantric cults and persisting folk possession traditions.
If I were to specify the field we are talking about in general and cross-cultural terms, it would be in terms of the deliberate and conscious cultivation of states of the mind-body complex, for any of a wide variety of purposes. The employment of a vocabulary of spirits or deities could be seen as a further shared characteristic of such processes, though ‘spirit’ and ‘deity’ are themselves terms that may be understood in a very wide variety of ways. <16> The cultivation of such states is certainly extremely widespread, and both shamanism and Tantra (Indian or Tibetan) could be seen as poorly-defined, overlapping areas within this general field.
The Indian material is nevertheless particularly interesting, and is the focus of this article, since India historically was a society, or rather a number of related societies, where sophisticated, literate techniques of cultivation of mind-body states have co-existed and interacted for many centuries with more down-to-earth, village-based and tribal practices. Throughout this history, Indian scholars, many themselves practitioners of the techniques they were analysing, developed their own vocabularies for dealing with them. I am particularly interested in one key Sanskrit term, āveśa, which was discussed some years ago by the distinguished senior French scholar of Śaiva Tantra, André Padoux <17>, and has been examined at much greater length by Frederick Smith in his monumental recent book The Self Possessed <18>.
Padoux, in his 1999 article, incidentally adopts an evolutionary model of Tantra:
Padoux thus sees the more literate and sophisticated forms of Tantra, such as that of the great 11th century Kashmirian scholar-practitioner Abhinavagupta, from whom many of his examples are taken, as essentially evolved and modified forms of wilder and more ecstatic approaches, a position with which I would generally agree.
As for āveśa, this is a term that at first sight looks rather similar to ‘possession’ in English; as Padoux notes, it is
The English term “possession” can of course be used both in terms of being controlled by an outside force - one is “possessed” by something, for example an evil spirit - or in terms of self-control - we speak of self-possession. <22> Things are perhaps superficially similar with āveśa, at least as used by the Kashmiri authorities whom Padoux cites, but they are not really the same; the contrast here is between possession by a negative force, a lower spirit one might say, and conscious entry into or invitation of possession by a higher entity. Thus āveśa becomes part of the vocabulary of Tantric practice, along with a series of related terms, among them samāveśa, the fusion of the individual with the divinity, and praveśa, which Padoux renders as entry or penetration. Thus the relation between deity and practitioner involves an entry of the deity into the practitioner, with the intention both of fusion with the divinity, and also, at the more pragmatic level, of exercising the powers of the divinity for this-worldly ends. <23>
The more orthodox Śaiva schools were and remain strongly opposed to such practices. As one might expect, they reject concepts such as āveśa and samāveśa as labels for positive spiritual experience, and they systematically reject the idea of a positive fusion with the divine <24>. One finds something very similar to the Śaivite Tantric picture, however, with a somewhat different vocabulary to be sure, in the Buddhist Tantras, where the process is that of union between the visualised deity which the practitioner evokes as a result of having undergone Tantric initiation and its associated vows - the samāyasattva - and the deity as a real aspect of the nature of the universe - the jñānasattva <25>. The samāyasattva is evoked within the practitioner and the jñānasattva outside, and they are then brought into a non-dual union. This is the key process in the ‘development stage’ of the so-called anuttarayogatantra-s. <26>. The Tibetan tradition certainly values such experiences very highly, although it should be noted too that it regards them, as did Abhinavagupta, as dangerous and advanced practices only to be approached after thorough training.
I have spoken so far of the individual practitioner, but there is also a relational aspect to these practices, as my colleague Louise Child has noted in her own recent book, Tantric Buddhism and Altered States of Consciousness <27>. This concerns in particular both the relationship to the guru and the relationship to the male practitioner’s female partner, who has a close connection to the yoginī, a class of wild, dangerous and powerful female spirits associated with the spirits of illness and misfortune. The female partner is a human representative of the yoginī, so that the relationship between male and female becomes potentially derivative from that between the male Tantric deity, who in the Śaivite context is generally a transform of Bhairava, and the yoginī-s or related female deities. Padoux notes that “in this masculine ritual world, the Yoginī are particularly supposed to possess women, who thus transmit their power, through sexual union, to the initiated yogins” <28>.
Padoux also points to the role of the female partner as an intermediary in the transmission of the teachings from the guru to the male disciple, a question on which David Gordon White has written extensively in his recent Kiss of the Yoginī <29>. One can go further here, and note that a whole social group is, ideally at least, being constituted through these practices, with male and female practitioners taking on the role of male and female deities within the maṇḍala of the principal deities. Ultimately, at least in the Tibetan context, one has the idea of the entire world being transformed, at least imaginatively, and ideally in reality too, into a human projection of the cosmic maṇḍala of the deities.
Smith’s book extends the discussion of āveśa and related terms much further, in a systematic exploration of vocabularies of ‘possession’ across a very wide range of Indic contexts, including both ethnographic material and various regional literatures. His historical perspective, as one might imagine of a treatment at such length, is more complex and nuanced than that of Padoux’s short article but the general picture of a common vocabulary which ranges from what we would describe as spirit-possession and spirit-mediumship, through a variety of shamanic and divinatory forms, through to the higher realms of Tantric practice, is reinforced and indeed drawn in considerable detail. Smith emphasizes however that, particularly in earlier contexts, ‘possession’ may be a misleading translation: in the Vedas, for example, the relevant Sanskrit root has rather the meanings of ‘pervasion,’ ‘immersion’ and ‘participation’ <30>.
Contrasts with Western Vocabularies
Is there, though, a basic contrast between these Indic and Indic-derived vocabularies, and the way in which we conceive of these matters in the West? I would suggest that there is. In part this is to do with Western constructions of body and soul, as against Indic constructions of the mind-body complex and its directing consciousness. We need, however, to step rather carefully here. This is not a straightforward question of Western dualism versus Indic non-dualism. Both Śaiva and Buddhist Tantra are in fact directed towards a non-dualist mode of awareness, whether conceived of in terms of the divine consciousness of Śiva or of the dharmakāya of the Buddha, but they are working, necessarily, with the dualist conceptions that are part of everyday consciousness in India and Tibet as they are in the West. However the dualism is phrased differently, and moves in a different direction.
Thus, in his 2007 article, Smith notes that both the early medical texts, the Āyurvedic saṃhitā-s, and the Mahābhārata, regard birth, or more precisely the animation of the embryo at conception, as a kind of possession of the embryo by the jīva or living being <31> . The terms used here for the jīva’spossession of the embryo are verbal forms of the same term āveśa which we have been discussing above. Correlatively, death is a withdrawal from possession of the mind-body complex. Thus the Anugītā, a section of the Mahābhārata, states in relation to birth that the “jīva or living being enters (āviśya) all the limbs of the embryo, one at a time, and at that moment conjoins with the mind, while abiding in the locations of the prāṇas”. At death, the jīva “is driven upwards, and having fully entered (āviśya) the heart quickly inhibits the existence (sattva) of the being.” The same terminology is used in relation to a whole series of other examples of ‘benevolent’ or ‘positive’ possession, for example in a ritual described in the Bṛhadāraṇyaka and Kauṣītaki Upaniṣad-s, through which a dying father transfers his senses and consciousness to his son, or in the incident in the Mahābhārata whereAśvatthāman invites possession by Śiva <32>.
I do not think that one can assimilate all of these Indic examples - and we are of course dealing only with a limited subset of Indian terminologies here - into a single pattern, any more perhaps than we can with shamanism, spirit-possession and related terms in the Western case, but I think that, despite the use of “possession” as a first approximation by Padoux and Smith - as well as by Monier-Williams - they tend in a rather different direction to the Western concepts of shamanism, spirit-mediumship and spirit-possession. The Western terms all imply a possession by or a relationship to an intrusive external spirit-entity of some kind. I am not suggesting, of course, that Indic thought does not conceive of external spirit-entities - it clearly has a wide variety of them - but it seems to me that one can read what is going on here in a rather different sense, and as tending towards a different kind of analysis of the relationship between spirit-entity and mind-body, or psycho-physical complex if you prefer. Here I find the image of birth (or rather conception) as possession, and of death as a relinquishing of possession, as particularly provocative, and as consonant with what is conceived of as happening within the process of Tantric practice, Śaivite or Buddhist.
It is rather hard to sketch exactly what is being implied here in a Western vocabulary, but one could perhaps say that the mind-body complex is one thing, and that what is happening to it is less a taking over by an alien spirit entity but a kind of patterning or shaping of it, the acquisition of a particular way of operating with it. This is clearest in the case of Tantric practice, since the point of the exercise is evidently not to “become Śiva” (or Vajrasattva, or whoever) in the sense that one might “become” Joe Bloggs or Mary Smith, but to reshape one’s consciousness and one’s total mode of being after the pattern represented by Śiva or Vajrasattva. It is explicit in some Tibetan texts at least that this reshaping of consciousness is also a patterning of the prāṇa flows, the currents of the subtle body that underlie and animate the functioning of the body-mind complex. This is perhaps hinted at too by the Anugītā’s reference to conception in terms of the prāṇas.
To put this somewhat differently, the issue is not so much that of one spirit or soul being taken over by another, but of incorporating or internalising - voluntarily and consciously, or involuntarily or unintentionally - a particular way of operating the mind-body complex. This may sound like a rather sophisticated mode of approach to the issue, and it is perhaps more explicit in an author like Abhinavagupta than in a village spirit-possession ritual, but I think that it is also worth recalling that village India operates with a very permeable boundary between human and divine; students worship gurus as gods, brothers worship sisters, wives worship husbands, young girls and young boys may serve as temporary habitations for the deity, gods, and spirits and for that matter demonic forces are thought of as interacting very freely with humans.
Perhaps we might suggest that this notorious fluidity of Indian presence of the divine, this constant openness to the divine, is not so much a question of credulity or of unscientific thinking as an implicit understanding of something that is worked out and argued far more consciously and rationally in the works of the great Tantric scholars - that the gods are less external spirit-entities than different ways in which individual human beings, as unified mind-body complexes, may comport themselves both towards each other and in relation to the wider cosmic pattern of which we all form a part?
This article originated as a paper for a session organized by Bettina Schmidt at the BASR Annual Conference, Edinburgh, September 3rd-6th 2007. I would like to acknowledge the comments of the participants on that occasion, and of Bettina Schmidt and Mathew J. Guest in relation to its revision for publication in DISKUS.
© Geoffrey Samuel 2008