MEDICAL ANTHROPOLOGY AND THE ETHNOGRAPHY OF SPIRIT POSSESSION

MEDICAL ANTHROPOLOGY AND THE ETHNOGRAPHY OF SPIRIT POSSESSION

Peter J. Claus (California State University East Bay, Hayward, CA., U.S.A.)

On Defining Problems
Ethnographic research over the past several decades has become increasingly problem oriented. There is a tendency in this sort of research to formulate models of a cultural phenomenon before the phenomenon is investigated. Indeed, the very problem is identified before the research is carried out. Research and description consist more or less of identifying features of the phenomenon with features of the proposed model. Verbal terms and behaviors are selectively chosen out of a large universe of similar forms, and are then translated in terms of the model. Causal relationships postulated in the model are used to explain terms and interaction in the phenomenon. Ability to do this not only demonstrates the efficacy of the model, but provides a scientific explanation of the phenomenon in that it appears to reduce a welter of particular facts from a large number of cultures to a smaller number of variables, usually having a more objective, naturalistic base. In this process of translating a particular phenomenon to a more generalized model, problem oriented research often finds it unnecessary to investigate the broad cultural context of the phenomenon in question. Problem oriented research is generally highly focused. Moreover, the focus is preset to take in pragmatic features: problem oriented research is closely allied with applied anthropology. There has been a wide-spread neglect of ethnography as an exploratory science, capable of discovering new perspectives with which to understand the human condition.



When the phenomenon of spirit possession is encountered in India it is usually regarded in the broad spectrum of medical problems. (1) It hardly seems to matter whether or not it is so defined in Indian systems of thought. In this regard, it might not be a bad idea to remind ourselves that Westerners have not always looked at possession that way. When early missionaries encountered possession cults in southern India they described this as a religious problem. Early accounts translated the various terms for the spirits in question as devils and demons. (2) The villagers were given to "devil worship" and "devil dancing." They were possessed by the Devil. We needn't labor over the gross inaccuracy of this translation of the phenomenon other than to note that the missionaries did give credence to the native belief in possession by a spirit agent. Where they differed with the natives was in the nature of the spirit involved.



I doubt that any serious anthropological study gives credence to the native claim that they are possessed by spirits, although we do sometimes acknowledge that we can proceed with the study "as if there were spirits." On the other hand, we still feel strongly obligated to recognize that some "real" experience is represented in their belief and behavior. We make a stab at apprehending that experience on the basis of our own understanding of similar behavior and beliefs in the West. Hence, our interpretation of the phenomenon as a medical problem or, more neutrally, as an altered state of consciousness. Perhaps we do not see it in terms of religious behavior because rational Christianity in the West today is more concerned with sin and ethical behavior in the world, and not with ecstatic religious experience. But, more likely, as scientists, it is our usual approach to abnormal behavior to explain it in terms of biological foundations. Placing spirit possession in the conceptual area of medical problems leads us in that direction from the very outset. This approach is supported by reference to certain selected instances of possession. These are primarily cases that possession often starts with expressions of anguish and ends with a "cure". That seems a lot like the hoped for progression of Western psychotherapy. (3) Furthermore, in some instances, native terminology and classification of spirits identify them with abnormal health conditions; specific diseases, madness, epidemics and the like. Some aspects of many diseases, to put it slightly differently, are associated with the influence or intrusion of a spirit. But these instances are hardly sufficient to regard all instances of spirit possession as native medical practices and transfer the phenomenon to a medical paradigm. What is worse, to do so merely leads us away from what is ethnographically relevant, which is, the documentation of the process by which individual experience comes to be interpreted through a system of collective symbols.



The Ideology of Tulu Spirit Possession
There are dozens of words in Tulu which are used in the context of what I have been calling spirit possession. (4) The most usual are bhuta pattundu ("the spirit caught...") and maytu battundu ("...came into the body"). Both verbs (pattunu and barpunu) have a large range of uses, entailed subtle contextual nuances and making them difficult to translate with regard to a specific experience. The verb barpunu ("to come"), for example, may be used to say "I will come to your house," "Good will come from that," "I will get angry (anger comes to me)," "I understand Tulu (Tulu comes to me)," and so on. Several verbs describe the physical appearance of the possessed (e.g. kumbarunu, "shaking") and the internal experience (e.g. ersuntu, "rising"; also barpanu, here, I believe, with the meaning "arising" or "becoming").

In a cermonial context possession is more usually referred to as darsana, "appearance, revelation." The English phrase used by my informants was "so and so had a darsana," although a moreliteral translation would have been the more ambiguous "darsana happened to so and so" (arugu darsana atundu). The intensity of teh possession--which is, incidentally, a frequently discussed matter, while whether a medium is "really" possessed, or not, is rarely questioned--is expressed in terms of edde battundu or edde pattundu) "it came (or caught) well."

The spirit in question may be one of a number of different categories: bhuta, daiva, kule, preta, sirikulu, cikku, and others. They include spirits of cultural heroes, animal spirits, ghosts, and anthropomorphic divinities of various sorts. Several publications are available which list over 300 or so spirits. (5) Many spirits and categories of spirits have only vague identities, but the character of some is elaborately described in long oral narratives (paddana) which describe their birth, "adventures", and death. These are spirits which tend to have a large cult following. Many other spirits are incorporated into these cults under specific conditions. The cults spirits are said to be able to control other spirits and command them to leave the body of a possessed person. The ritual relationships amongst these spirits constitute a vast subject, much too complex to pursue in this paper. However, it should be noted that a great deal of discussion revolving around an unexpected possession concerns the identification of the spirit and which of the cult spirits has the authority to command it to stop bothering the possessed.

The affliction caused by a spirit is usually described as upadra, "trouble." This term may be applied to certain cases of unwarranted possession, but ore frequently it is used to label violent diseases brought about by spiritual attack on humans, animals or crops. It is also used to describe unusual and annoying events, such as mushrooms growing inside a house, as well as slow, persistent wasting diseases. In many of these circumstances there is considerable ambiguity as to whether the spirit is an external agent of affliction or whether his internal presence is the cause of the trouble. Some afflictions are said to be caused by the spirit's "touch" or "scratch".



It remains to be said that by no means are all instances of spirit possession regarded negatively. In cult ceremonies possession by the appropriate spirit is regarded as normal, legitimate and desirable. In other circumstances it is often precisely because the spirit cannot easily possess a person that the person suffers discomfort. Resistence on the part of the person or members of his/her family, uncleanliness (ritual pollution) of the body, existence of the influence of another spirit or a curse are the primary reasons why a spirit might have difficulty possessing its human vehicle. Possession is regarded as a normal means for a being of the realm of maya to make its presence or, more specifically, its need, known in the world of joga (living humans). What causes problems is generally human interference in these matters. 


In Search of a Framework
In the past we have focused our attention on the emotional experience of possession. We have identified the phenomenon as a medical problem and have narrowly researched it in terms of a cause-and-effect paradigm which presumes that it has some sort of somatic starting point. There is little in terms or concepts used by the people to suggest this to be the case. For them it is a very religious phenomenon; it has a spiritual origin. We look at the act as a medical problem because we take biological features to be more basic than cultural or ideological ones.


Attention to the terms people use to describe unusual experiences leads us in a direction to view it as they do. For the ethnographer this is a worthy goal. But reference to terminology of this sort is not the end of the line. It is only the beginning. The terms themselves--to come, to seize, to occur, appearance, spirit, trouble--are vague and metaphorical. (6) The verbs involved are all-purpose words, used in numerous contexts in general conversation. At best, here, they indicated that the transformation of character which the person undergoes is the result of the action of an external agent. (7) While that is an important part of the whole notion, it does not tell us specifically who the agent is, why it happens, why this person, why at this time, how, and so on. All of these questions are important to those people around the possessed person. Some of the answers are revealed after lengthy discussion and ritual action. The answers to some of the questions, however, are so deeply embedded in the structure of the Tuluva ideology that they are hardly ever raised. Many of the answers are to be sought in social contexts and social processes. Others are to be discovered in the character of the spirit and even by asking the spirit itself. And still others are to be deciphered from ritual action and symbolic expression. Terminology is but a small, albeit initially important, part of a very large inquiry. 



My own research has led me to the investigation of the way possession conjoins the individual (sometimes representative of a larger group or category of which he is a part) with his/her society's moral order. The sentiments from which possession springs (along this line of thinking) are no doubt common to all societies and are handled in many societies in different ways. (8) They arise under a number of situations, largely culturally conditioned, in which the individual feels himself outside of the order of moral expectations in which he has been conditioned to see himself. The specific Tulu expression of this crisis is in terms of the encroachment of beings from the domain of maya (the strange, mysterious, supernatural) on the domain of joga (the ordinary manifest and ordered world humans live in). Social groups larger than the individual too, in times of crisis, question, or even as a course of periodic routine, reassert the primacy of an ideal order over the exigencies (however they may arise) of individual and daily life. The phenomenon of spirit possession---or rather a tradition of spirit possession--can provide a symbolic medium through which the individual (or group) re-adjusts himself to an appropriate order.



There are certain instances in which "non-professional" possession is regular and elaborate in its expression. The several with which I am most familiar are found in Tulunad. In an earlier paper (l973), I described a case of possession among the audience attending an annual village bhuta ceremony. As is often the case during such ceremonies, village legal matters were discussed in the presence of the village spirits during the stage when the spirit possesses the body of the cult priest (pujari). In the course of the proceedings a woman from one of the litigant's family was possessed by a household spirit, and this spirit then proceeded to argue the household's case. The woman began her possession with a frenzied trance, but was soon encouraged to come into more focused possession by the possessed priest. Possession brought the legal matter to the level of the spiritual realm and conjoined not only the village, but also the household, each through their possessed representative, to a moral order above the level of day-to-day interests.


Another example may be cited from the northern part of Tulunad in which a large proportion of the village "audience" routinely partakes in a variety of possession and trance forms at the annual village rituals. The ritual ca be regarded as a sort of renewal rite in which members of the village individually and as representatives of groups express through possession states their relationship to the spiritual order which ideally governs the village:


The village headman (he is still referred to by a royal title in the context of the ritual) is seated in front of the village along with the deity, who takes form in the body of a priest-medium. Women from each of the village households surround the pair and the women themselves become possessed by their household ancestors and spirits, who are considered to be in the village deity's retinue in the co-existent realm of maya, and who desire to be present at their lord's appearance during his"visit" to the realm of the living. After the headman pays his tribute to the deity, the men of the village file past the lords of the two realms, offering their loyalty and a small symbolic tribute. They receive in return the blessings of the royal pair and the promise of continued prosperity and protection. Women then walk barefoot through a pit of burning coals which has been constructed in front of the temple. In past times, it is said, the men performed acts of self-immolation to demonstrate their loyalty to the king and his deity. This act, on one hand, symbolizes the extremities to which the men would go in the service of their lords, and on the other hand, the mystical protection from pain and injury the king and his deity could extend to those who serve them (Claus l978:31)


The Siri cult ceremonies provide an instance in which non-professionals are drawn into possession rituals in large numbers. Most participants have little contact with one another in their ordinary lives. Most participants have little contact with one another in their ordinary lives and many participants each year are newcomers. All have a history of troublesome individual possession previous to their participation in the cult, a condition which was resolved by their promise to serve (seve) the Siri spirits as possession vehicles in subsequent years. At the annual Siri cult ceremonies the participants, mostly women, gather in small groups (some of them meet occasionally at a village shrine under the leadership of a male priest at other times in the year) and recite a ritual version of the legend of Siri. Soon all are possessed in one form or another. The attention of the adepts is given to helping novices to more easily and fully express the character of their possessing spirit. The violent, troublesome nature of the novice's possession is felt to be due to her inability to allow the spirit to fully possess her, and this is in turn due to either ineptitude on her part, or malicious resistence, often using mystical means, on the part of her relatives. A variety of means are used to facilitate her full possession. The ritual location, chanting the Siri legend, possession enactment of key episodes of the legend, the presence of possessed adepts and their use of kinship idiom to invite the novice into a spirit group, the power of the male priest to remove mystical obstructions, and so forth, all serve powerfully to encourage the novice to express her possession more articulately and elaborately in a accepted character. In short, recitation of myth and ritual action are brought together in a dramatic model through which the individual is reintegrated into a moral order.


We have seen then how ritual encourages and shapes the possession experience. In standard, periodic cult ceremonies the structure of the ritual performance and the formal recitation of myth easily transform the specialist into a spirit and the ceremony into an enactment of the exploits of a culture hero. Others may be drawn into performance as well, and in certain instances may be regularly encourages to fully express their possession by means of a variety of ritual aids.


What it is that they are drawn into is often most clearly seen in myth. This is especially true in cases where the possessed embodies one of the many "historic" culture heroes. This sort of possession tends to be both more elaborate and more intense. The imagination of the South Indian villager finds its most elaborate and creative expression in the lives of the one-living heroes. Often these are based on real historic personages. However, there can be little question that the bulk of legends and epics consists of stylized elaboration of imagery already a part of oral tradition and everyday speech. The number of imagery already a part of oral tradition and everyday speech. The number of epics and legends of this sort in the different regional cultures of southern India is vast and many of the individual pieces are enormous in length. Collection and assessment of this body of literature has barely begun.


Scholars from different disciplines have viewed these works of oral tradition from a number of different perspectives. However, most pertinent for the study of the relationship between these heroic epics and legends and their enactment in the form of possession is the 'moral imagination' these works exhibit. In this it is not unlike the oral tradition of many cultures. Beidelman, for example, writes of the Kaguru (an East African culture):


"Instead of considering situations and persons inextricably linked to a totality of phenomena over a long, even enduring, period of time, as in real life, the Kaguru story teller presents an extreme and limited case...While stories avoid the complexities of possibilities in these daily problems, they do point out implications and difficulties posed as one tries to succeed where conflicting and competing loyalties perplex people, and where yesterday's enemies may be tomorrow's friends. Their very simplicity gives stories their attraction, just as the simplicity of sociological models both reduces yet helps explain s social reality...These stories are odd, not in the sense that they do not represent recognized characteristics, feelings, motives and roles, but in the sense that, whereas in real life these cannot all be properly judged and met by the same person or in one situation, here they are clearly defined and resolved." (Beidelman l980:33)


Like Kaguru tales, the morality of South Indian oral traditions is not easy to discern, because the rightness or wrongness of the heroes' actions is conditioned by the circumstances. The Tulu legends, for example, relate the heroes' actions in desperate situations and what in the legends are regarded as courageous acts are often actions which would be at odds with the expectations of everyday life. Nor is the moral order abstractly, or directly, described. Instead it forms a set of assumptions, a backdrop against which the actions of the hero are tried in a series of dialectic relationships, and through which he emerges a hero.
There is a lot in common between the legends and the immediate lives of the possessed. Specific similarities can be found at all levels; structure, theme, mood, character, setting, action, etc. The legendary heroes provide a model for the possessed. Through the lives of heroes, in the limited and extreme contexts of their legends, people can find possible definitions and resolution of their own circumstances. The fact that the heroes are medial beings----apotheosized humans--who define their heroism in desperate situations homologous to the situations living people are apt to be disturbed by, facilitates identification and transposition of their character. This is so not only in regular cult ceremonies, but also for those who experience desperate situations in their daily lives. Material I collected on reasons why many Tulu women are attracted to the Siri cult, where their initially vague, disturbing possession is brought toward resolution, suggested the following observations:


The individual struggles to live with a condition which is threatening to her propitious conduct, upon which, in turn, the prosperity and reputation of those around her depend. The threat is almost invariably brought about by the lack of male protection which the cultural ideology required n order that women may be secure against the ravaging intrusion of supernatural and human enemies...Under such precarious conditions a woman is vulnerable not only in the material world in which she must have sustenance, but also in the supernatural realm in which she is defenseless. By giving herself as a vehicle for one who might be her savior, a spirit of one who had preserved the concept of feminine virtue in her own lifetime, a woman choose the lesser of two dangers. By becoming the vehicle of her spiritual protectress, she brings resolution to vague fears and apprehensions. The more clearly her tutelary spirit can be defined as virtuous, the more secure and powerful can be the woman's defense against those who seek to destroy her (Claus l979:49).





Conclusion
My argument in this paper is that the incidence of spirit possession is a complex ethnographic phenomena. To interpret it with a simplistic Western medical paradigm is insufficient. There is no doubt that the possessed individual exhibits interesting abnormal psychological behavior and further, that his or her actions often symbolically express in some way unresolved past experiences of an interpersonal nature. But this is largely expressed and interpreted in a culturally specific symbolic system and the individual is reintegrated into a moral community through ritual action. When we approach the phenomena through problem-oriented paradigms we short-circuit the ethnographic process. If we are to carry the investigation further we have to e careful that we touch base at all times with native concepts. Indigenous terms and metaphors point us in the right direction. Indigenous classification constitutes the landscape. Myth and ritual provide us with models of how people think about and experience what we are unable to observe.


This procedure will, I believe, lead us in a different direction than does the problem oriented research of medical anthropology. Problem oriented research of the sort characterized in the beginning of the paper has a tendency to lead us back to where we started. We learn nothing new, nothing from others and little of real usefulness and ourselves. According to these views spirit possession is universally associated with specific psychological or physical conditions (e.g., calcium deficiency in the diet, as one recent suggestion runs). The behavior of the possessed is equated with the behavior of patients with such conditions. Ritual treatment of the possessed, to the extent that it is effective, is seen as a primitive form of the treatment of patients in a modern scientific clinic. These views ignore most of what is ethnographically relevant. We feel secure, perhaps, in identifying a universal or biological basis for behavior and confident in our own progressiveness for treating scientifically what others do with magic and ritual. But what do we really learn?

Unfortunately, possession is often ignored even in the ethnographies of cultures in which it occurs with remarkably high frequency and even in ceremonies in which the ethnographer records its occurrence. AT best we are given a simple description of the possession behavior, or the costume, or the dance-like movements of the possessed, or the occurrence of the act in relation to the temporal and musical aspects of the ceremony. Almost nowhere do we find an investigation of the meaning of possession in the same sense as we find discussions of other major religious phenomena, such as sacrifice or witchcraft.


So far, ethnography has failed to bring to bear on this phenomena what is important from a larger culture perspective. Ethnographers have missed an opportunity to understand a rather remarkable set of cultural events, a set which only the ubiquitous ethnographer is apt to witness n its greatest variety of forms and contexts.


What is probably most significant about possession is that it allows us to witness in detail the relationship between the expressive, and the interpretative and the integrative aspects of religion. Each case of possession is a case in point, varying from the highly standardized, routine and collective rituals held periodically at the village level to the spontaneous occurrence of possession in the household in the midst of daily life. The whole range is linked through a tradition of possession belief from which individual instances are interpreted and to which they contribute. We know little about such religious processes in any religious context, and a study of possession from this perspective would seem to me to constitute a major contribution.



References Cited
BECK, B. E. F : 1982 The Three Twins: The Telling of a South Indian Polk Epic. Bloomington: Indiana University Press.

BEIDELMAN, T. 0. : 1980 "The Moral Imagination of the Kaguru: Some Thoughts on Trieksters, Translation, and Comparative Analysis." American Ethnologist 7: 27-42.

BHATT, P. G. : 1975 Studies in Tuluva History and Culture. Manipal (India): Manipal Power Press. 

BLACKBURN, S. : 1980 "Performance as Paradigm: A Rhythm in a Tamil Oral Tradition." Paper presented at the conference on Models and Metaphors in Indian Folklore, Berkeley, California, February 7-10,1980.

BOURGUIGNON, E. : 1968a "World Distribution and Patterns of Possession States." In R. Prince (ed.), Trance and Possession Slates. Montreal: R. M. Bucke Memorial Society.

1968h "A Cross-Cultural Study of Dissociational States: Final Report." Columbus: Ohio State University Research Foundation.



1973 "Introduction: A Framework for the Comparative study of Altered States of Consciousness." In E. Bourguignon (ed.), Religion, Altered States of Consciousness, and Social Change. Columbus: Ohio State University Press.

1976 Possession. Corte Madera, California: Chandler and Sharp.

BURNELL, A. E. : 1894 "The Devil Worship of the Tuluvas." Indian Antiquity 23.

CARSTAIRS, G. M. and R. L. KAPUR. : 1976 The Great Universe of Kota. London: The Hogarth Press.

CLAUS, P. : 1973 "Possession, Protection, and Punishment as Attributes of the Deities in a South Indian Village." Man in India 53:231-242.

1975 "The Sin Myth and Ritual: A Mass Possession Cult of South India." Ethnology 14:47-58.

1978 "Heroes and Heroines in the Conceptual Framework of Tulu Culture." Journal of Indian Folkloristics 1: 28-42.

1979a "Spirit Possession and Spirit Mediumship from the Perspective of Tulu Oral Traditions." Culture, Medicine, and Psychiatry 3: 29-52.

1979b "Mayndala: A Myth and Possession Cult of Tulunad." Journal of Asian Folklore Studies: 38-95-129.

FOSTER, G.: 1976 "Disease Etiologies in Nonwestern Medical Systems." American Anthropologist 78: 773-782.

FOSTER, G. and B. ANDERSON. 1978 Medical Anthropology. New York: John Wiley and Sons,

LAKOFF, G. and M. JOHNSON : 1980 Metaphors We Live By. Chicago: University of Chicago Press. MANNER, A.

1886 Pdddanolu. Mangalore: The Basel Mission Press. 

NICHTER, M. : 1979 "The Language of Illness in South Kanara (India)." Anthropos 74:181-201.

Obeyesekere, C. : 1976 "The Impact of Ayurvedic Ideas on the Culture and the [individual in Sri Lanka." In Charles Leslie (ed.), Asian Medical Systems. Berkeley: University of California Press. Pp.201-226.

1981 Medusa's Hair, Chicago: University of Chicago Press,


PADMANABHA, P. : 1971 "Special Study Report on Bhuta Cult in South Kanara District." Census of India, Series 14 Mysore (Karnataka), Miscellaneous (a). Delhi: Government of India Press.

ROGHAIR, G. : 1978 "The Role of Brahma Nayudu in the Epic of Palnadu, "Journal of Indian Folkloristics 1:15-26. 

 
1. A classic example of this may be seen in Carstairs and Kapur 1976. There possession is treated as a psychiatric symptom along with other such symptoms as anxiety, depression, psychosis, etc, True, they do place it in the context of the local South Kanara cultural setting (e.g., pp.59-61, 67-69) and qualify it as other than a mere symptom of mental illness by allowing that there is '4'.. a close association between cultural beliefs and psychopathology . , , "in the case of possession (1976:110).

A far more sophisticated discussion of the relationship between private symbols and public symbols having both personal and interpersonal significance may be found in Obeyesekere's recent book, Medusa's Hair (1981). In this book-too tightly argued and intricately documented to review here-Obeyesekere successfully demonstrates that phenomena such as possession can have deeply motivated unconscious meaning for the individual and at the same time fit meaningfully into a network of religious ideology. The two approaches (the public and the private) and the two interpretations (psychological and religious) need not be viewed contradictorily, as I, also, argue later in this paper. 

2. Defining the spirits worshiped by the Tuluvas as demons did not prohibit the 19th century missionary from studying the religious literature or practices of' 'demon worshipers," but it did, certainly, influence the way they treated the data, The Rev. A. Manner, for example, wrote in the preface to his fine collection of paddanolu (1886):

"As the physician in the course of his study requires in many cases to apply himself to unpleasant research, even so does the Mission-worker require often to study adverse literature, which may not be pleasant, to enable him efficiently to cope with the difficult task of meeting the heathen on his own ground. From a scientific point of view these stones are not of much worth, but they serve to give one an insight into the hollowness of demon-worship and enable the Mission-worker among this class of people to gain an insight into their ideas and to follow out their line of thought. ... It is therefore in the interest of the Mission-worker to whom, as aforesaid, such information is peculiarly valuable, and not with any intent to give wider publication to these stories that we have had them printed and I have in hand a small number of copies for sale at cost-price to missionaries and Mission-workers only, strictly prohibiting the loan or sale of such under any circumstances whatsoever, to the heathen" (1886: i). 

3. There are, of course, other aspects of the relation between this class of spirits and the human realm which could lead us to assume that spirit possession is essentially a medical phenomenon. In popular belief all around South Asia, diseases of many types are associated with the supernatural (smallpox with Mari or Sitala; leprosy with the cobra, etc.). Propitiation of supernatural agents of disease is a common means of ridding the community or the individual of many types of disease. There is no doubt here that many South Asians adhere to a medical belief system which falls into the broad class of what George Foster calls a personalistic etiology (Foster 1976; Foster and Anderson 1978). What is dangerous in this sort of labeling exercise, is to assume that the "real" diseases we recognize and explain by scientific medicine correspond to the same diseases another culture recognizes and explains by supernatural agents. If this were true, medical anthropology would have a rather simple task of translating one system into another. But this is not often the case. Not only are there different (from ours) and subtly varied notions of causality involved in any one system, but there are different ranges of phenomena identified as diseases. The very ideas of what constitutes disease, health, etc., are different. There are, for example, what Obeyesekere calls "cultural diseases," because "they are created, at least partly, by the cultural definition of the situation" (Obeyesekere 1976: 207). 

4. For more details, see Nichter's (1979) extensive discussion of the vocabulary of illness in this region of India. 

5. See Padmanabha 1971; Bhatt 1975; Burnell 1894. 

6. I am referring here to metaphor in the sense discussed in the book Metaphors We Live By, by Lakoff and Johnson (1980). 

7. The most usual general concept of the body which the metaphors for possession entail is that of a container. In this respect, the body is likened to an idol or a temple, both of which arc also containers for the supernatural. Pots, stones, trees, poles, and seats (benches, swings) are also common 'containers' for spirits and deities. Each of these items, including the body, emphasizes somewhat different aspects of the notion of containing: temples, for example, often stress a series of boundaries; idols stress visual form; long, flexible poles allow the vitality of the spirit contained therein to be expressed through a violent undulating motion. The human body, having a number of these features, is thought to be an especially appropriate container for the spirits of cultural heroes and deceased kinsmen. 

8. For a listing of comparative material on spirit possession in other cultures, see Bourguignon 1968a, 1968b, 1973, 1976, Bourguignon uses mental and physical states as the basis for ordering her data.