Causes of Spirit Possession in African Women


Maria Cruz


            Anthropologists have long since endeavored to explain the institutions of African society to a world unfamiliar with the special type of culture that the continent offers.  Unfortunately, Western scholars have tended to make their findings subjective, often asserting their own value system where an objective viewpoint is required in order to truly understand how African village life functions.  The subject of spirit possession is a perfect example of this slanted type of research.  Since the turn of the century, researchers have attempted to explain the complicated enigma of ritual healing that is simultaneously medicinal, religious, and political.  Throughout a century of study, three major hypotheses have emerged to explain the cause of possession among African women.  In this paper, I will present a brief overview of the three main theories behind spirit possession and construct arguments both for and against their content in order to reach a better understanding of the reasons behind spirit possession. 

            Before criticisms of spirit possession can begin, it is integral that its concept must first be fully understood.  Spirit possession is both a symptom and a cure for various types of illnesses that descend mostly upon women, although men can sometimes fall victim as well.[i]  The women who become possessed are usually either married or prostitutes, depending on the particular culture discussed, but for both types of women, their use of makeup, henna, scented colognes, etc. makes them more prone to possession, for those are the attributes that attract the spirits, or zairan.  Within these zairan are three types of spirits, defined by color—white, which bring benign illnesses, black, those harbingers of grave disease or mental illness, and the most common, red, the ambivalent, capricious spirits that bring milder illnesses not resulting in death.  Unlike white, which do not require much treatment or black, which require violent exorcisms, red spirits must be constantly placated in order to be considered cured.  The items with which spirits are placated can also be divided into three categories: items normally forbidden to women (cigarettes, etc.), non-durable purchased goods (clothing, sugar, etc.), and items of foreign provenance (automobiles, fez, etc.).[ii]

Once both native and Western treatments have failed to cure a woman’s illness, it is commonly recognized that some supernatural force must be the cause rather than a natural, physical phenomenon.  Other possessed women then help her use the treatment of trances in order to control her symptoms, just as they themselves received aid when their own illnesses could not be cured in any other way.[iii]  Those symptoms, depending on the type of spirit present, may be any sort of unexplained illness from headaches to violent vomiting.  The purpose of the trance is for the woman to “enter into a contractual relationship with the spirit responsible for her lapse in health”.[iv]  In order to control future attacks from the spirits, trances must be used continually to prevent relapse into sickness, which requires regular attendance at meetings with other possessed members, and hence the formation of various spirit possession cults.  The groups, formed under the auspices of medical treatment and prevention, can be seen as either wholly therapeutic, a vehicle for women’s ascent into the political domain of men, or both.  It is my contention that while these cults offer a way for women to step outside of the status quo for brief periods, their foundation is largely medical in nature, and that foundation should not be overlooked by scholars searching for Western values in African society.          

The first theory of spirit possession is by far the oldest and most prevalent in literature on the subject.  The main argument is that possession occurs due to the instinctive tension between men and women in male-dominated societies.  Author I.M. Lewis contends that when Islam came to the African continent, this new faith zealously swept out any involvement that women bore in organized religion and left them without a way to worship ancient spirits, as in the past.  Praising Allah took precedence over worshipping ancestors, and therefore this largely female cult of spirit domination—a way to involve ancestral spirits in everyday life once more and subvert the power completely overtaken by men—evolved.[v]  While this argument does hold a modicum of truth, it leaves large gaps in reasoning by assuming deceitful intent by all possessed women. 

This long-standing and conventional theory claims that women, in an attempt to gain more social control within the village, use the facade of spirits as an excuse to act in a manner normally considered taboo by the society.  It does not explore the deeper realms of the medicinal qualities of trances, but merely makes an overall judgment about the objective of the female gender.  The mind-set of this hypothesis can be found within literature such as Michael Onwuejeogwu’s study on Hausa bori—“in the bori, Hausa women experience the trappings of officialdom…it is therefore a mechanism through which women escape the seclusion of the inner compound and enjoy, at least in fantasy, the public life outside the compound.  It is thus the symbolic way of escape from the role of female to that of male”.[vi] The wording of this phrase implies that women use the cult exclusively as a means to gain “officialdom”, a say in the politics of the village, rather than as a way to treat illness.             

However this idea should not be discounted completely. Spirits are sometimes used for the good of the village (if they are called on in times of famine, drought, or epidemics for help), and these special circumstances should not be forgotten when speaking of social control through spirit possession.  Social status theorists may also be correct in that it cannot be assumed that all women enter into these cults for medicinal purposes—there certainly may be women who are unhappy with their station in life and wish to gain as much control as possible.  But the concept of all women being unsatisfied with their social status is an illustration of Western ideals being used to judge non-Western situations.  Westerners largely see African women’s social status as backwards and may presume that they would look for any opportunity to change it.  But African women do not see themselves this way, and so they would have no reason to want such a change.  Gender equality is a Western concept, and cannot be used to form an opinion of other societies, especially in objective, empirical research.  To assume that females want the same status as males is assuming Freud’s “genital anxiety” theory to be correct, when it is merely a theory, not a proven fact, especially across every culture.  This is where the social status theory falls apart—its presumption of all women wanting the same type of equality that those in the West take for granted.

Anthropologist Susan Carol Rogers, after studying the social power schemes of a Kenyan tribe, takes a slightly different view on this theory and critiques the established use of political power being the only real power within village life.  She argues that women already possess another type of power, that of domestic control, which is ultimately more influential than any political clout.  Her theory begins with the statement “that women virtually everywhere play a subordinate role is a recurrent, implicit, or explicit assumption in anthropological literature”.[vii]  Within a peasant society, females have more power than males, for it is they who take care of their husbands, and teach their children the values of their society.  Most anthropologists overlook their “informal power” because it is within the closed element of society, the tightly knit domestic unit that outside researchers cannot penetrate.  Rogers does bring these interesting points to the forefront, but too much of her argument relies on the assumed control of women—she herself states that it is an area closed to outside scholars, so how does one know exactly how much influence is wielded within the inner compound by mothers and wives?  However, speaking in terms of disease, Rogers’ theory does not imply “deceitful feminine tactics”[viii], but instead allows the possession of women by spirits to be seen as a purely healing ritual.

At the far end of the spectrum from the social status theory of spirit possession lies the more contemporary fertility theory advocated by Janice Boddy in her work dealing with villagers in Northern Sudan.  She explicitly states, “it is not necessarily good or desirable to have a spirit”[ix] which completely places her within a new domain of those who advocate the Lewis model.  Boddy takes the opposite approach from the prior anthropological work and states that spirit possession cannot be defined solely by itself, but in context with a woman’s entire position in the community.  Boddy, like Rogers, realizes that nothing is straightforward, and that no society, even Western, is ever clearly divided and categorized.  To assume that an African society is any different would be preposterous. 

She views spirit possession instead as a feminine trait stemming from the fact that women are freer than men to pursue religious folk beliefs outside Islam, whereas in most other aspects of life, they are more confined than males.  In dealing with village matters, women are submissive to men, but possession is a chance for women to regain a piece of that freedom.  She does not advocate that women use this freedom to try to gain an advantage over men, but instead paints a picture of Sudanese society as a delicate balancing act between the sexes where “everything is resonant with implicit meanings”.[x]  In her theory, women are possessed only after an illness; they do not fabricate spirit possession in order to subvert the male population.  Everything is a balance between the sexes, and spirit possession is merely a part of that, not a result of it; it is a part of harmonizing various social institutions.  However, Boddy does go slightly too far with her metaphors, citing the process of mixing and baking bread as a significant example of the complimentary roles of women and men in the reproductive domain.[xi]  This is where her theory becomes unstable, because by actively searching for implicit examples in social norms, one is sure to find them—it does not necessarily mean the author and the reader will see these examples in the same light.       

Janice Boddy argues that women have preponderance in possession rituals because most of the illnesses that are being treated involve women’s issues—sterility, stillbirths, nursing, menstruation complications, etc.  Therefore it is the domain of women not because of any desire to have an exclusively female social group, but because men cannot possibly relate to the issues in question.  I believe that she is correct in this assumption; female medicinal issues are rarely addressed by local medicine men and Western physicians are sparse, consequently women must look to each other for guidance in such situations. She, unlike Lewis, never implies that women become possessed voluntarily in order to thwart the male power system, but instead directly attacks his argument by stating “[his] social status model is unidimensional”.[xii]  But she responds by involving so many social norms into her theory of possession that it loses its impact on the reader.  It takes on the air of merely another daily activity that women complete in order to keep village life running smoothly.  But this is not the case.  Possession is the result of a “psychological disturbance and mischief and misery for the whole community”[xiii]; it upsets village life, and does not appease it.  It is not relatable to courtship rituals or bread baking.  It is a phenomenon unto itself with many inputs, but nothing so closely related to it in social institutions that it is easily comparable, as Boddy attempts to do. 

Boddy also gleans over that fact that possessed women are in fact a sort of political institution that those in positions of power may use to attempt control of natural events such as bountiful harvests and good rains.  She also does not give much precedence to the fact that women do extort goods from their husbands while in transient states, which abolishes the idea of the ritual as sacrosanct and pure—it is too easily feasible that some women may be partaking in a mild form of racketeering.  If they kept solely to themselves and demanded nothing of no one, then possession could be viewed as completely untainted, but since it involves other citizens outside of the possession group, motives are debatable.

A final theory on the reasons behind spirit possession in Africa comes from a pair of researchers who choose a purely scientific approach to the issue and defined nutritional deficiency in women as the reason for their symptoms of cramps, seizures, etc.  Alice Kehoe and Dody Giletti recognized that both of the theories provided by other anthropologists were “handicapped by dependence upon inferences about unconscious processes or about unacknowledged deception”[xiv]; that is, too vague and reliant on [possibly] tainted information for any concrete conclusions to be made from them.  They instead set out to find a tangible reason for the preponderance of spirit possession among women.  Their studies showed that women involved in possession rituals showed a lack of calcium, magnesium, vitamin D, etc. in their regular diets for number of reasons, including “nutritional inequivalence between men and women”[xv], and also because of certain social institutions that deprive pregnant women of necessary milk and meat products.

The calcium deficiency theory, while neatly providing an empirical explanation for women’s possession, is too scientifically oriented to be applied practically to zar cults in Africa.  The reason these cults are so difficult to understand is because they incorporate so many different aspects of medicine into a single entity, and to attempt a separation between the medicinal and psychological aspects would be disadvantageous to any research on the subject.  While cults are founded on the basis of ritual healing (one must become ill due to spirit interference, not natural causes), members also use their minds to control their symptoms; by channeling the spirits of ancestors, they psychologically influence the disease in order to heal themselves when all other forms of medicine have failed.  Because spirit possession is also a form of folk religion, to apply a scientific explanation is futile—it would be just as ineffective to attempt to explain the relationship between the power of Christian faith and secular healing.  I believe that a researcher absolutely cannot explain one as separate from the other; to endeavor to do so in African society merely because it is African would be quite untenable.

            Trying to understand the reasoning behind spirit possession in Africa as if it is a simple black-and-white procedure is a wrong approach that leads to the downfall of most research on the topic.  Possession cults incorporate so many different social norms—marital relationships, female status, fertility, religion, etc.—that whenever a scholar concludes that one of these is the cause for all occurrences of possession, another scholar can easily find evidence to counter such an argument. A blending of ideas discussed in this paper may result in a more comprehensive view of possession causation; I will therefore endeavor to explain spirit possession in context with these different hypotheses.

            Several points must be clarified before critiques can begin, however, the first of which is the fact that researchers of opposite genders tend to view situations differently.  Lewis’ studies may overlook key points, not because they are not present, but because he, as a man, does not understand the relationships that form among women.  Boddy may conversely rely too heavily on her female point of view and also partake in one-sided objectiveness.  Also, the time period when studies were conducted is crucial, because Lewis was the first published of all researchers spoken of in this paper—Kehoe and Giletti therefore had the advantage of being able to rebut his work, and Boddy, being the most recent study, had the chance to absorb and dissect all previous examinations.  It must be kept in mind that Lewis, Onwuejeogwu, etc. did not have such an opportunity. 

That aside, the strongest point of Boddy’s fertility argument is the way in which she draws from so many dimensions of society in order to arrive at her point—unfortunately, she shows too much heavy-handedness in this regard to be taken seriously.  The other two theories, social status and calcium deficiency, rely too much on a single issue of a multifaceted entity.  Lewis’ dependency on the innate conflict between men and women, for example, rules out the idea of conflict between members of the same gender—it ignores the ever-present emotions of competition, jealousy, and rivalry that women must manage on a day-to-day basis.[xvi]  Likewise, Kehoe and Giletti’s hypothesis is far too abstract in context.  It also takes responsibility away from the possessed—nutrition deficiency infers that the women themselves have no control over their symptoms or cure, and that is certainly not the case.  It is similar to the Western answer for psychological complications of prescribing drugs to patients.  I believe that Boddy is correct to identify the preponderance of women in these cults to be because of fertility issues, and I believe that she is also correct to see past the possession itself and link it to the greater context of society.  However, if she were to concede on a few points of Lewis’ and Rogers, and associate them with her own theory, it would prove to be a more comprehensive model. 

A more ideal theory would realize that, because possessed women do influence the world outside their cult, members cannot be seen as segregated from society; they are in fact, deeply rooted within it.  Folk religion is the offset, the balance, of Islamic authority, and women’s power within the domestic unit is the counter to the men’s visible political power in the village.  But above all, it is medicinal.  Scholars searching for authoritative institutions in village life too often overlook this fact.  It can never be forgotten that spirits are the cause of illness in possession cases, and whether or not outside researchers believe in spirits, native Africans do.  This must be taken into account—if spirits are automatically assumed to be fictitious, the research will be null and void.  The fertility theory recognizes this, but it must also recognize that spirits influence the non-possessed (husbands, village elders, etc.), and so political power schemes do play a role in possession cases.  She must also recognize that women do have a firmer hold on domestic power than she describes, that the balancing act of which she speaks is neither delicate nor precarious.  Women are firmly established members of the social order and have no need to constantly dance around men in fear of upsetting the equilibrium. 

It is inevitable that Western anthropologists should have a subjective view of the society that they study, and it is inevitable that their papers and lectures should reflect this.  However, they should also be willing to appreciate other scholar’s observations and incorporate them into their studies, not reject them as completely erroneous.  With possession, all three theories provide a unique insight into the reasons behind the act, and it is disreputable to allow any theory to stand alone when they could be melded together to explain the difficult concept.  It is only in this way that a comprehensive idea of spirit possession can ever be attained by Western anthropologists looking into African civilizations.   

[i] I.M. Lewis states that possession also affects men of marginal social status, and that these men are accepted into mainly female cults [in Ethiopian zar and bori cults specifically].  Ecstatic Religion, 101

[ii] Harris, Grace, Possession “Hysteria” in a Kenyan Tribe, American Anthropologist 59, 1957, p. 1052.

[iii] Boddy, Janice, Spirits and Selves in Northern Sudan: The Cultural Theraputics of Possession and Trance, American Ethnologist, 15, 1988, p. 10-11.

[iv] Ibid. 11

[v] Lewis, I.M., Ecstatic Religion: An Anthropological Study of Spirit Possession and Shamanism, (Baltimore: Penguin Books, 1971), 126-148.

[vi] Onwuejeogwu, Michael, The Cult of the “Bori” Spirits Among the Hausa, in Man in Africa, Mary Douglas and Phyllis M. Kaberry, eds.  (New York: Anchor Books, 1969), 290.

[vii] Rogers, Susan, Female Forms of Power and the Myth of Male Dominance, American Ethnologist 2, 1975, p. 728.

[viii] Harris, 1050.

[ix] Boddy, Janice, Wombs and Alien Spirits: Women, Men, and the Zar Cult in Northern Sudan, (Madison WI: University of Wisconsin Press, 1989), 145.

[x] Boddy, Spirits and Selves, 8.

[xi] Ibid. 8

[xii] Boddy, Wombs and Alien Spirits, 139.

[xiii] Hurreiz, Sayyid, Zar as a Ritual Psychodrama: From Cult to Club, in Women’s Medicine: The Zar-Bori Cult in Africa and Beyond, I.M. Lewis, Ahmed Al-Safi, and Sayyid Hurreiz, eds. (Great Britain, Edinburgh University Press, 1991), 139.

[xiv] Kehoe, Alice B., and Giletti, Dody H., Women’s Preponderance in Possession Cults: The Calcium Deficiency Hypothesis Extended, American Anthropologist 83, 1981, p. 550.

[xv] Ibid. 553.

[xvi] Besmer, Fremont E., Horses, Musicians, and Gods: The Hausa Cult of Possession-Trance, (South Hadley, MA: Bergin and Garvey, 1983), 20.