The term ‘possession’ has been applied to Africa, the African diaspora (especially Brazil and the Caribbean), the Middle East, the Pacific, and sometimes South and Southeast Asia in contexts in which humans are said to be temporarily displaced, inhabited or ridden by particular spirits. During these often highly framed episodes, voice and agency are attributed to the spirit rather than the host; the host is not held accountable for what occurs and indeed may claim subsequently to have no knowledge of it, or at least no ability to have influenced its direction. The spirits are generally conceptualized and experienced as discrete persons, whether ancestors, foreigners, historical figures, gods or members of an alternate species. These persons may be viewed as more or less distinct from their hosts according to the particular performance tradition at issue, as well as the stage the relationship between an individual host and spirit has reached.
Some commonly recognized cultural forms (which may cover a range of local variations) include Zfl? (Northeast Africa), Bori (Hausa), Vodou (Haiti), and Umbanda and Candomble (Brazil). While some writers have sought to clarify differences between possession and shamanism (a term generally applied in Asia and indigenous America), a categorical distinction does not appear to be useful and much of the South and East Asian material can be framed in the same general terms as the African. Nor does it make sense to firmly distinguish possession as a sociological type from other instances in which people find themselves subject, in varying degrees, to the influence of disembodied external powers. The choice of whether to exorcize a demon or assimilate a spirit generally represents less a distinction of kind than a local politics of religion, as well as an informed reading of the immediate social context and personal circumstances of the particular host.
Spirit possession has long exercised the anthropological imagination. This may be, in part, because it is itself an imaginative construction. Indeed in many instances possession forms a sort of unrationalized counterpart to anthropology itself, whereby what is foreign or distant is appropriated, framed, inspected and used for local reflection. The main difference is that while anthropological knowledge is objectified, rationalized and reproduced in textual form (such as this topic), spirit possession is embodied, reproduced in the bodies of human hosts. The presence of spirits is made manifest in episodes of illness, dissociation, dreams and taboos, as well as in performances in which the spirits may dress in their own clothing, eat their own foods, speak in their own voices and generally display behaviour which contrasts with that of their human hosts. Spirit possession thus occurs in the real time of human life and is experienced as an impingement upon it.
Spirit possession forms a complex and exciting subject for anthropological analysis for two main reasons. First, it combines both an order of collective thought in the distinctions among the spirits as a semiotic system (myth, totemism) and an order of collective practice (ritual, therapy, oracles) whereby hosts are initiated, particular voices are legitimated, and spirits are provided with the spaces in which to perform. Possession thus provides a context in which contemporary experience can be actively mediated by past myth models and vice versa. Second, it provides an instance in which the collective and personal clearly interpenetrate. Collective forms are internalized by individuals and become self-transforming. Likewise, individual intentions are externalized and given form through the voices and acts of the spirits and the careers of their hosts (Obeyesekere 1981). Performances of possession therefore have a degree of unpredictability that many other kinds of rituals lack. Possession raises fascinating questions of agency and accountability for participants and anthropologists alike.
Performance and the play of agency
If we take seriously the idea of spirits and possession as imaginative phenomena, it follows there is no single essence that underlies them. Hence reductive attempts to explain particular local practices as simple expressions of illness, hysteria, or relative deprivation (Lewis 1971) have been replaced by more nuanced forms of understanding. To the degree that possession is embodied, it constitutes a specific performance. Performances may be comic or dramatic, sus-penseful or gay, establish hierarchies or invert them. They make use of music, dance, distinct clothing, and unusual speech patterns and body movements. Different features may be highlighted in different performance traditions but one of the most striking will be the evident discrepancy between the features of the host as a recognized member of the community and the transformation that overlays them as she is possessed by the spirit. Possession may be linked to broader ritual cycles or political processes. It may be displayed before a crowd, restricted to privileged audiences, or embedded in kinship relations. Some of the most significant episodes of possession may take place in private settings among intimates (Lambek 1981).
As mimetic forms and practices that are always also skewed, altered, distorted versions of their original objects (Taussig 1993), spirits are often taken to represent ‘Difference’ itself, an ‘other’ world against which the quotidian is set off and from which it can be reflected upon. In precolonial times the salient difference often depended heavily upon imagery drawn from nature. The critical gap that spirit possession established and mediated may have been, as Levi-Strauss had it, that between nature and culture. At the same time, the spirits also drew upon signs of social otherness or cultural distance. Since the colonial period, the salient opposition has often shifted to what might be called culture-superculture (local-global). Possession attempts to think through or capture the relationship between indigenous forms of authority, power and practice and those manifestly stronger, external ones, often exemplified by the seductiveness of foreign commodities. Since the object of mimesis is almost always conceived of as very powerful, an interpretation of possession must address questions of hegemony and resistance (Comaroff 1985). Moreover, since power is not merely represented but addressed, possession is almost certain to be at least implicitly a discourse on morality (Lambek 1981; Boddy 1989). As a mimesis of power and a discourse on morality, possession is attuned to contemporary circumstances. Frequently bringing the authority of the past to bear on the present, it is also highly flexible and ready to change.
If spirits are imagined, they may become like works of art that, in turn, seize our imaginations, overwhelm, and captivate us with their power, interest, beauty or sublimity. They may, indeed, ‘possess’ us. Drawing on the classic analysis of "Godfrey Lienhardt (1961), Kramer (1993) makes a strong case for treating possession as passiones, a term that has dropped out of ordinary usage but means something like the opposite of action. Kramer delineates African ‘spirits’ as the images of passiones embodied in people and shrines. To be possessed is to be ‘carried away’. Kramer argues, however, that modernity ultimately locates agency within the self, replacing cosmology with psychology, passion with action. Devotion to the other has been lost and our links to the external, but intrinsically interconnected world have been reconstituted as internal qualities of mind. This broad discursive transformation may be partially responsible for the spirits’ withdrawal to the shadows, in some parts of the world, although it cannot explain their persistence and even expansion in others.
There is a significant difference between possession and such captivating aesthetic forms as poetry or film that enfold us in their worlds. Possession demands to be taken literally; the displacement is marked. Possession is an embodied phenomenon, manifesting itself in physical pain, spiritual trauma, anxiety, hyperconvulsive behaviour, temporary and partial amnesia. It invites a therapeutic response, one that usually requires addressing the spirit as a distinct entity from the sufferer, from the one impassioned. Social recognition is granted to the displacement of agency. But this means that out of passion stems a course of action.
Any viable institution of possession must retain the social means to render its spirits ‘real’. Naturalization (vraisemblablisation) develops from the fact that negotiation between the host and spirit is a public, social event. Possession takes place in the presence of at least a third party, a therapist, mediator, spectator or, if the spirit’s appearance in the host is already well authorized, a client. The existence of the spirit is corroborated through public witnessing of the embodied transformation of the host, legitimated through collaboration in dialogue with the spirit, and, subsequent to the possession episode, by the actions and statements of the host who must exhibit competence at shifting frames. The performative constitution of the spirits as distinct persons is mystified to the participants so that they come to understand themselves as simply ‘meeting’ entities whose existence is independent of such acknowledgement (Lambek 1981).
The difference between disembodied and embodied spirits is one of focus. Bodies provide the vehicle by means of which spirits can be held in focus, as it were; through which they can be sustained in steady communication and from which they can be expected to respond. These are not necessarily the effects of radically different belief systems, but of weaker and stronger forms of collective authorizing. The more the passio is understood as an external, even autonomous entity, the less need there will be for recourse to a psychological discourse. But this in turn is enabling of greater agency.
The nature and transformations of agency are apparent when possession is understood as a system of communication, providing alternative, authoritative voices and critical distance, but also anti-language and ambiguity (Boddy 1989), and the possibility for the simultaneous transmission of opposed messages (Lambek 1981). As Besnier (1996) demonstrates, possession is truly "heteroglossic.
Selfhood and therapy
To the degree that spirits are embodied, they are lived through and with, and to this extent also they impinge on the selfhood of their hosts. Boddy (1989) describes how possession expands culturally overdetermined selfhood for Northern Sudanese women. Kapferer (1991) examines the unmaking and remaking of the self during the conduct of specific rituals in Sri Lanka, while authors in Crapanzano and Garrison (1977), and Obeyesekere (1981) drawing on psychoanalysis, view possession from biographical perspectives. However this need not be understood pathologically; possession implies weakly bounded egos, but not necessarily weak egos per se. Moreover possession can be a vehicle for the growth of psychological insight or maturity. Once again, the emphasis on passion needs to be complemented by noting the agency, especially moral agency, and dialogical qualities of spirits and hosts (Lambek 1993).
Spirits are often associated with trauma and healing. As such, possession is full of moral references and contextualized by accountability narratives. A diagnosis of possession may provide an entree into therapy or a rationalization for its inadequacy. But at times illness is simply an idiom of possession, one of its embodied, "indexical qualities that contributes to the legitimation of spirits as autonomous entities and marks their arrival in human bodies as independent of the wills of their hosts. Some of the possessed may become therapists or oracular voices in their turn. To perform as a medium requires psychological insight and social savvy as well as the capacities for empathy, context shifting and making full use of the communicative properties of possession.
There is a great deal of excellent work on possession from many parts of the world. Boddy (1994) provides an extensive bibliography.
