STWR - Share The World's Resources

Dr Edward C. Green

Senior Research Scientist in the School of Public Health, Center for Population and Development Studies, Harvard University

Purity, Pollution and the Invisible Snake in Southern Africa.

Abstract

There exists in parts of southern and east Africa an apparently widespread belief in the existence of an invisible, internal "snake", often described as a power or force of some kind that dwells in the stomach but that can move throughout the upper body. Although some anthropologists have described this snake as related to witchcraft, findings from diverse parts of Mozambique, South Africa and elsewhere suggest that it may (also) be thought of as a symbolic expression of the need to respect the human body and specifically to protect it against the introduction of impurity. Belief in nyoka, as Tsonga- and Shona-speakers call the invisible snake, suggests the importance of purity and pollution beliefs as they relate to health in a particular society; the presence of nyoka belief may even be taken as an empirical measure of their importance. Going beyond nyoka, it is argued that pollution beliefs are more central in southern African ethnomedicine than the literature suggests, perhaps more so than witchcraft and sorcery beliefs. It is hypothesized that pollution-related illnesses tend to be roughly coterminous with diseases biomedically classified as contagious.

Apart from ethnographic and theoretical significance, establishing the nature and centrality of pollution beliefs, aided by analysis of cultural metaphors such as the invisible snake, can point to culturally-appropriate ways of presenting health education messages in societies where pollution beliefs are important. Pollution beliefs may be characterized as quasi-naturalistic and they in fact represent an area of potential interface between indigenous and cosmopolitan medicine--far more than witchcraft beliefs.

Introduction

There exists in parts of southern and east Africa an apparently widespread belief in the existence of an invisible, internal snake, often described as a power or force of some kind that dwells in the stomach but that can move throughout the upper body. It is designated by the local term for snake (nyoka in at least Shona and Tsonga; other variants include Nyowa, ndjoka, ndzoka, nyoga, inyoka and Nyakwadi). Less often a local word for worm is used, perhaps signalling a different concept. If a comprehensive treatment of this subject exists in the anthropological or related literature, I have been unable to find it. Indeed there seems to be little published of the subject at all. Yet the invisible snake concept appears to be of fundamental significance for an anthropological understanding of illness causality beliefs that relate to diverse illnesses and conditions affecting both children and adults, including diarrhea and other stomach conditions, sexually transmitted illnesses, epileptic and other convulsions, mental retardation, growth abnormalities, digestion, helminthic infections, fecundity, pre-natal development and childbirth, and barrenness.

This paper summarizes findings from the author's research and from the literature about belief in an invisible snake in Southern Africa. To bring some order to diverse research findings, they are organized chronologically, the earliest presented first. There are some minor exceptions to this in the interest of presenting the findings from several Mozambican societies in the same sub-section. It is argued that belief in an invisible snake reflects among other things the importance--and can be an empirical measure--of pollution beliefs. According to Ngubane (1977:77) pollution is "conceptualized as a mystical force which diminishes resistance to disease and creates conditions of poor luck, misfortune ...' disagreeableness' and 'repulsiveness'...In its worst form (pollution) is contagious." Lacking in this definition (taken from Ngubane's very useful broader treatment of the subject) is emphasis that pollution is believed to cause illness, whether directly of indirectly. I argue that pollution beliefs are more important in southern African ethnomedicine than has been appreciated and I offer the tentative hypothesis that pollution-related illnesses tend to be roughly coterminous with diseases biomedically classified as contagious. This latter means that the "fit" between the invisible snake and pollution is less than perfect, since the former is related to non-contagious as well as contagious illnesses and conditions.

It is further suggested that an understanding of the invisible snake is important for anthropological theory-building in the form of typologies that identify central conceptual themes or cultural metaphors underlying ethnomedical beliefs. The paper specifically deals with the question, Do witchcraft explanations predominate in explanations of illness in Southern Africa, as some have suggested? Evidence is offered from the author's and others' research suggesting that pollution beliefs may predominate for several Southern African societies. It is suggested that pollution beliefs resemble "naturalistic" rather than "personalistic" health beliefs. Finally, apart from theoretical significance, an understanding of the invisible snake points to culturally-appropriate ways of presenting health education messages in societies where pollution beliefs are important.

Methodological Summary

My own research was in all cases "applied" and related to public health intervention objectives--child diarrhea and sexually transmitted diseases--rather than to theoretical or other academic issues. Research in Mozambique took place between 1991-5. It was assisted by Mozambican researchers who conducted semi-structured, key-informant interviews in local languages, taking about 15 person months in Manica, 2 person months in Nampula, and 4 person months in Inhambane. Some 275 Mozambican traditional healers were interviewed in the three provinces. I designed all instruments and interpreted all findings, and conducted in some interviews in each province. My research in Swaziland was similar in objective and methods but was longer (3.5 years) and included considerable participant observation with traditional healers. Swazi research focused primarily on environmental sanitation and diarrheal disease. Limited interviews with knowledgeable elders (non-healers) were conducted in Swaziland in 1995 to learn about the invisible snake; these findings were reconciled with earlier ones. All other research was of shorter periods (between 3 weeks and 3 months) and was also aimed at health interventions. Interviews were conducted mostly with traditional healers in all countries. This is because my applied health work in southern Africa has often focused on the potential role of healers as change agents. Healers presumably represent the beliefs of clients who consult them (some 80% or more of the community) and they are often better able than their clients to explain such beliefs, both because of their specialized knowledge and because their status in the community makes them less likely to be intimidated by an interviewer. They also seem too possess deeper knowledge about etiology, including causation, of locally-recognized illnesses, a point often made by "laypersons" when interviewed. Evidence supporting this assertion can be found in the body of this paper.

A number of studies from the literature, each with their own objectives and methodologies, are also reviewed. Their findings, along with my own, are uneven in details and sometimes contradictory, a point taken up later in the paper.

Although the literature on invisible or internal snakes appears to be scant, there is a fair amount from Mozambique, much of it unpublished. This country encompasses a great deal of vertical or north-south territory, cutting across a number of ethnolinguistic zones or belts. The various ethnic groups of Mozambique are closely related to groups in contiguous countries--South Africa, Swaziland, Zimbabwe, Zambia, Malawi and Tanzania--which gives an indication of the cultural diversity found within Mozambique. This also suggests that invisible snake beliefs found in Mozambique ought to be present among linguistically-related groups in neighboring countries. Limited research findings presented below tends to confirm this.

The Snake in the Stomach: Summary of Findings

Based on research conducted early in the 20th century, the missionary and respected "amateur" anthropologist Junod (1962b:474) describes a condition that "...might be gastritis, congestion of the liver, or dysentery" and notes:

Another missionary anthropologist (Earthy 1968) worked among the Lenge in Gaza, several hundred kilometers to the north of Junod's area in Mozambique, between 1917-30. She too provides findings on an invisible snake, introducing the subject much as Junod did: "And here I must describe the strange concept of Nyakwadi...

She notes further,

Based on fieldwork among the Sotho-speaking Lovedu of the Transvaal (South Africa) a half-century ago, anthropologists Krige and Krige (1943:212) speak of an "...internal snake believed to be intimately bound up with fecundity and childbirth. Every one is said to have a snake in his stomach." They continue:

The Kriges continue to say a bit more about the internal snake as it relates to human reproduction.

W.D. Hammond-Tooke, in a book that seeks to identify patterns underlying health beliefs in Southern Africa (1989:55), discusses the Kriges' findings under a topic heading of "traditional ideas about anatomy and physiology" rather than under an immediately preceding section on "life force and personality". As I will argue below, the internal or invisible snake may indeed be life force with a personality. Elsewhere in his book, Hammond-Tooke (1989:80) suggests "a snake in the belly" can be a distinguishing characteristic of witches in South Africa. This notion may come from the report of Booyens suggesting the Tswana of South Africa associate serious diarrheal illness with the concept of an "intestinal snake" called kokwana. According to Booyens (1989:11), "...it is said that the snake, 'sent' to the child through witchcraft, 'eats' the child's food and the child itself." The invisible snake concept that emerges here appears as something evil, a weapon of sorcery or witchcraft. Indeed snakes occupy a place in African cosmology as witches familiars, manifestations of ancestor and other spirits, and adjuncts to rain-making and other rituals. This helps obscure the invisible snake-pollution association for researchers.

When my colleagues and I first encountered the belief in central Mozambique of a snake that dwelt in the stomach, called nyoka, we were conducting ethnomedical research on child diarrhea (Green, Jurg and Tomas 1991:12-13; Green, Jurg and Dgedge 1993:267-8; Green, Jurg and Dgedge 1994). As we began to discuss our earliest findings with local health workers, we were told that nyoka represented a crude understanding on the part of uneducated people of the role of the human intestines and/or of parasitic worms that could be seen in the feces of infected people. Findings by Maina-Ahlberg (1979) and Yoder (1981) in fact suggest that ideas of snake-like creatures in the stomach may be related to intestinal worms elsewhere in Africa. This seemed plausible yet the traditional healers with whom we were conducting in-depth interviews specifically denied this. Other health workers thought nyoka was somehow related to sorcery. Informants also denied this. It may be noted that our sample of 104 healer informants represented not only Shona speakers, but the Ndua, Manica, Sena and Ute languages as well. After a series of repeat interviews specifically on nyoka, it emerged that this snake is conceived as an invisible force that somehow demands purity of the body it inhabits. If contaminants enter the body, the nyoka react with displeasure, causing pain and discomfort. It is referred to as if it has a personality somewhat independent of the body it inhabits; for example it may be angry or calm. A useful English translation of the concept might be Guardian of Bodily Purity. Nyoka may be thought of as a symbolic expression of the need to respect the human body--even as a personified immune system. We further found:

As a force that ensures purity, it was not surprising to discover that nyoka was associated with menstruation, reflecting the common belief among traditional Africans that menstrual blood is highly polluting. When "dirt" accumulates in a woman's body, her nyoka twists and turns in discomfort and irritation, which movements are felt in cramps that precede menstruation. Such cramps are taken as evidence that the nyoka is preparing to expel accumulated impurities from the body through the menstrual process. Another focus of our ethnomedical research in central Mozambique concerned STDs, or illnesses believed associated with sexual intercourse or genito-urinary symptoms. Menstrual pollution proved to play a role here. If a man has intercourse with a woman during her menstruation, he becomes contaminated with the impurities her nyoka is in the process of expelling, and he develops a disease known as nyoka khundu. If a man does not treat this nyoka khundu, not only will he remain sick but at the moment of conceiving a child with a woman, the unborn child's nyoka will be negatively affected. When the child is born, it will not only have symptoms of nyoka khundu; it will also be susceptible to various other illnesses. If we may attempt translation into biomedical terms, the child will have poor resistance--we might even say a weak immune system (Green, Jurg and Dgedge 1993:267-8).

About the same time as our fieldwork in central Mozambique, studies in connection with two separate theses were concluded, one on mental retardation and the other on diarrheal disease, both conducted primarily among Tsonga-speakers in Maputo province. Both Marrato (1991) and Sousa (1991) encountered belief in a snake or "worm" that lived in the stomach or intestines, the movements of which related to the production of convulsions and diarrhea, as well as to mental development of the child. Fieldwork in both studies was based on in-depth, anthropological type interviews with traditional healers as well as their clients (mothers).

In 1992 a thesis in medical geography appeared based on fieldwork in southern Mozambique, on the Ronga-speaking island of Inhaca (Gibbs 1992). A finding of interest here is that nyoka (the same term is used among the Ronga and closely-related Tsonga) is "a basic traditional concept" explaining disease causation. Nyoka is said to dwell in all people. It can be bad, i.e., cause illness; yet it serves to "give life" and one cannot survive without it (Gibbs 1992:31). When food is eaten, the "chief nyoka" either accepts the food, "putting it toward the life-giving force within the body"; or it rejects it in which case it is converted to poison. This poison makes people ill. There is medicine which "kills the poison" and changes the nyoka in a way such that it stops making poison.

According to Gibbs, there is belief among Ronga traditional healers that nyoka is related to worms that are visible when they leave the body. But this is explained as the nyoka "becoming many" in order to make much poison. Medicine, in the form of store-bought (anthelminthic) syrup, can expel the worms, yet in the words of a diviner, "But some must stay in--if they all come out you can't live." (Gibbs 1992:33). The researcher goes on to comment that Ronga laypersons cannot--or at least do not--give explanations such as the foregoing. They may talk in a general way about nyokas but they refer more detailed questions to traditional healers, sometimes noting: "Only the nyangas can know such things".

During a water and sanitation study among the Macua of Nyasa province, northern Mozambique, Jurg et al (1994:22-24) encountered belief in an invisible snake, locally called nyowa. As elsewhere, the term for this means snake or worm. According to traditional healer and other informants, all children are born with a nyowa and it remains in the body forever. It does not always provoke health problems; it may be either provoked or "silent". It can make noise and move around in the stomach--even upward causing a person to "spit constantly" to the point of vomiting. It can bite the stomach but this does not necessarily cause diarrhea; it causes "large stomach" in children. Nyowa seems more related to birth (inheritance of essences or qualities from parents or spirits?) than to the causes of specific diseases--yet "can be treated with roots and papaya leaves". The purpose or goal of nyowa treatment is not described. My colleagues and I conducted interviews on child diarrhea with some 85 Macua traditional healers in another northern Mozambique province, Nampula, in March 1995. We found that nyowa was more often called mihaco (also miona, nowa,or mihoua) and that it was conceived as illness-associated worm that becomes visible when expelled from the body through diarrhea or vomiting. Mihaco seems to have no positive (i.e., protective, health-promoting, cleansing) function; instead it provokes anemia, diarrhea, other illnesses. It "eats" or "suck up" a person's blood and "weakens the body;" it "bites until it causes death" There was disagreement over whether everyone is born with these worms or still has them in adulthood. One healer said, "Some you are born with; others you catch from outside..." Not surprisingly, the aim of therapy is to kill or expel the worm(s). Only one out of 83 healers held a different opinion, commenting: "It is necessary to have mihacos in order to live..there are some who only live in the stomach and do no harm."

In an ethnomedical study of seizures, loss of consciousness, and convulsions in Swaziland, Reis (1994:S40) mentions a snake in the stomach which causes convulsions by raising itself in the body. Reis notes that this seems to be interpreted by Swazi as a naturalistic or impersonal rather than personalistic, cause. A study of epilepsy among Tsonga-speakers of southern Mozambique that relied on traditional healer informants (Panizzo 1994) also found that an invisible snake, nyoka, was causally involved in seizures and convulsions, and that these were related to the local illness nyokane (see below).

In 1993 I had the occasion to interview traditional healers and health workers from many parts of South Africa in connection with evaluation of a collaborative program in AIDS (Green and Zokwe 1994). I used the opportunity to do some brief interviews on invisible snakes. According to my co-investigator, a Xhosa with deep knowledge of that group's traditions, all children are born with a "snake" in its stomach. If a child has diarrhea, or bites itself (wanting blood), it must be treated to "calm its nyoka." These symptoms also mean it is time to have an animal sacrificed for the child, which will let its nyoka know it's being recognized and appeased. Later in the life cycle, during certain ceremonies where traditional brew is being taken, one must take a little always for one's nyoka. On a superficial level, nyoka can refer to tapeworm. "But this is the kind of distorted interpretation Xhosas might give whites" to hide their culture from probing and unsympathetic outsiders. Nyoka is a dangerous force until it is "developed" and calmed ("settled") and turned into a positive force of health. It is like a special soul. It has a personality. My informant was emphatic that nyoka is not something bad nor something that should or can be removed. A Pedi traditional healer endorsed the foregoing description as true for the Pedi as well as the Xhosa. She maintained that Zulus understand nyoka in the same way, "unless they've forgotten it."

During an informal group discussion on nyoka involving six traditional healers (five Xhosas and one Zulu), all agreed that everyone is born with a nyoka. It is created at the moment a man's sperm fertilizes a woman. Nyoka must be "developed" in order to "nurture" it and make it calm. Zulus were said to begin to nurture the nyoka prenatally; Xhosas wait until the child is born. The nyoka tends to be reincarnated and carry traits patrilineally, therefore if a father--but even a mother--is unhealthy at the time of conception, the baby will be born unhealthy and must be treated at once. Later, if a child's fontanelle is depressed, its nyoka is said to be thirsty. It must be given medicine to relieve the thirst. The nyoka may at times pull a baby's fontanelle down. According to the six South African healers, the nyoka at the time of death partly stays with the corpse (requiring special rituals if the person dies violently or away from home); partly becomes reincarnated; and partly becomes a fully-participating and useful ancestor spirit after some 20-30 years. There seems to be a transitional period during which the nyoka learns how to be an ancestor spirit. In this last depiction, nyoka seems to or resemble or be part of the "multiple soul" found in many parts of Africa.

Late in 1993 I had an opportunity to train a group of interviewers in anthropological research methods in Tanzania. The interviewers were originally from diverse parts of that country, yet many were familiar with a belief in an invisible snake in their home areas. Some used the term ndzoka. However they were unsure whether the snake was essentially a destructive, negative force or something related to maintenance of one's health.

During my fieldwork in Swaziland (1981-85 and periodically later), I confesses overlooked the significance of the invisible snake. However from limited interviews with elders in Swaziland in 1994 I learned that all children are born with worms (tilo) in their stomach. These grow into a snake (inyoka), described as a kind of invisible master of the body. Inyoka keeps the person healthy; one must have it in order to be alive ("Once it comes out you are dead"). When someone is hungry, his inyoka makes noise ("krrro-krrro-krrro!"). When taking food that does not agree with inyoka, it dispels it through vomit. One elder said, "It helps analyze the food and put the dirty things outside (the body)....The inyoka instructs you in what it needs." Another commented, "This inyoka instructs you in what it needs. Some call it appetite." Thus in Swaziland the snake seems to represent an ethnophysiological concept of digestion and appetite, an essentially positive force.

Also in 1994 I conducted brief research focused on child diarrhea and STDs in Inhambane province, in southern Mozambique but north of Inhaca island and Gaza, already referred to. I conducted interviews with seven traditional healers and also participated in a week-long discussion with 33 healers in the course of a collaborative workshop. My Mozambican colleagues had just conducted interviews with healers over a several week period. The healers spoke Tsonga or related dialects such as Chitwa or Bithonga. Just the year before, a week-long workshop for speakers of a Tsonga dialect in Gaza province had concluded: "There is consensus (among traditional healers) that the nyoka of a person becomes disturbed when impurities and dirt accumulate inside the body". (GEMT 1993) According to healers in Inhambane, the most important locally-recognized children's illness is nyokani. A child may inherit a quality or condition from its mother and thus be born with this illness, the symptoms being foam in the baby's mouth as well as vomiting, fits, convulsions, and diarrhea. Another type of nyokani can result from a child being left on the ground, allowing "dirt" to enter the child via its anus. Although nyokani involves more symptoms than diarrhea, one healer commented, "Nyokani is the mother of all diarrheas." Once diagnosed, treatment of nyokani must begin immediately, continue for 2-3 weeks and be coordinated with the phases of the moon.

As the name suggests, nyokani is related to nyoka. The first is an illness while the second, according to our Inhambane informants, is something that all children are born with inside them that grows as the child grows. In fact there are said to be two nyokas, one male and one female. Nyoka may "eat the energy" of a child's food. It may provoke diarrhea, vomiting or fatigue. One healer said traditional medicine must be given to the child to kill the male nyoka, which in turn will cause the female nyoka to die. Other healers described somewhat different treatment approaches but seemed to agree that the aim of treatment is for the nyoka to be expelled through defecation. According to some, the purpose of the medicine is to slice the nyoka up into small pieces, which then come out a bit at a time and are visible. Yet other healers denied there can be any external or visible evidence of nyoka.

After the week-long workshop for traditional healers in Inhambane, I tried to reconcile some of the apparent inconsistencies between accounts of nyoka among speakers of Tsonga- related dialects in southern Mozambique. Is nyoka a force to be nurtured and strengthened? Or should it be killed and expelled? With these questions in mind I interviewed two traditional healers from Inhambane who had moved to the Maputo area, one a Chopi and the other a Tsonga. They both agreed that all people are born with a nyoka. If the nyoka becomes disturbed, it manifests itself as the disease nyokani. The attacks of this illness are related to the phases of the moon. The purpose of therapy is to kill or expel the nyoka. However this does not pertain to the "beneficial nyoka". This nyoka was said to be essential in the maintenance of health; indeed a person cannot live without it. It turns bad or becomes agitated or angry if one eats food that disagrees with it. Nyokas have their own food preferences: "My nyoka may like a food that your nyoka hates." The Tsonga healer volunteered that "good hygiene" is needed to keep nyoka undisturbed. When I asked for examples, both healers said that surroundings, including latrines, kitchens and cooking utensils, must be kept clean and free of flies. A nyoka may be happy or angry, depending upon the "purity" and "cleanliness" of the body. An unhappy nyoka makes noises in the stomach. There is treatment to "clean out the dirt from the stomach", which pleases the nyoka.

Finally, I interviewed a number of traditional healers and health workers in a Bemba-speaking area of Zambia in early 1995. I encountered belief in nsoka, which means snake but in the context of an entity dwelling within the stomach seems to refer to ordinary worms--visible when outside the human body--that cause symptoms of illness. Not everyone has these worms. Those who do, suffer from an illness called insokanda, believed caused by eating impure food or drinking dirty water although it can sometimes be caused by sorcery. It is considered contagious. Bemba healers provide medicines in order to expel the nsoka and thereby cure insokanda.

Summary of Findings and Discussion

Based on published and unpublished sources, as well as my own fieldwork, it appears that belief in an invisible snake can be found along a strip of southern and east Africa extending at least from the southern cape of South Africa (Xhosa) to Tanzania in the north. Janzen (1989:243) in fact suggests this strip may represent a meaningful cultural area, somewhat distinct in health beliefs and ritual practice from other Bantu-speaking areas. No doubt a more exhaustive literature search would discover additional references and a more accurate map could be constructed. There may be significant belief differences among societies in southern Africa regarding the nature of the invisible snake. Researchers need to determine if the invisible snake is conceived as a positive force, a neutral ethnophysiological concept, an expression of witchcraft, or something that may be both positive and harmful, depending on conditions. Other questions include: Must one always have one or more snakes in the body to remain alive? Is the snake essential to life and health--or is it a consumer of vital energy and/or a major source of illness? When the snake relates to the objective of therapy, is the aim to: (a) kill or expel it or "cut it into pieces"; (b) strengthen or nourish it; or (c) "calm" or "cool" it? Are there medicines to strengthen the nyoka of children? Is the concept related to purity or pollution-prevention? Future research should determine whether any differences in beliefs are actual or due to differences in research methods and resulting data validity, to institutionalized prevarication (as my Xhosa informant suggests), or to varying levels of causal explanations.

According to my own and others' findings, traditional healers may give deeper, more detailed explanations than laymen regarding complex concepts such as nyoka. The topic in fact might be guarded or tabu. Earthy (1968:63) wrote of the Valenge, "The young people are not usually taught about Nyakwadi, though they overhear their elders talking." In Reis' Swazi study, patients of traditional healers were found to be less certain than healers in their explanations of causes of seizures, although they offered "the same variation in etiological concepts" (Reis 1994:S40). This argues for anthropological, rather than survey, methods and specifically for in-depth interviews with traditional healers.

Returning to the invisible snake, although researchers such as Booyens and Hammond-Tooke have suggested this is related to witchcraft, in fact it appears to be quite the opposite for most (but not all) of the societies reviewed. The invisible snake emerges as a positive albeit dangerous force. Reviewing the essential qualities of the invisible snake in many of the societies for which we have information, we see it is a force that requires cleanliness and purity of body; it reacts to the introduction of "dirt" or impure, spoiled foods by provoking various bodily discharges such as diarrhea and vomiting as well as grumbling in the stomach; it guards the body against impurities or what Shona healers called "contamination"; and it requires clean surroundings external to the body. I offer the hypothesis that the internal snake is conceived as a protective life force when it is invisible under all circumstances; it is an ethnophysiological concept related to digestion and stomach disorder when it is conceived as a worm (less often, a snake or grown-up worm) that becomes visible outside the human body. These two roles are not mutually exclusive since even as a protective life force, the invisible snake seems to regulate diet and digestion.

Classification of Illness by Cause

As Foster (1983:20) observes, "Traditional medical systems are often marked by 'levels' of causality; we find both efficient and proximate causes...Until the causal agent has been discovered, therapy is believed to have little effect." Foster (1983:19) also suggests that broad classification of illness causality concepts into "natural" and "supernatural" is less useful than into the alternative categories of personalistic (aggression or punishment directed at a specific individual as a consequence of the will and power of a human or supernatural agent or being), and naturalistic (where illness is explained in impersonal, systemic terms). Most of Africa, Foster suggests, is characterized by personalistic explanations. Ayurveda, Unani, Chinese, and humoral pathology, on the other hand, are systems based essentially on naturalistic causality.

Yet in many parts of Africa an important health-related causality concept can be found that is not personalistic. This is the notion of pollution. There seems to be general agreement among scholars that pollution beliefs represent a form of naturalistic or impersonal causation that is distinct from attribution of illness to spirits, witches, sorcerers, or ancestors (Janzen 1989; Hammond-Tooke 1989, 1981a, 1981b; Ngubane 1977; Douglas 1966). Even the missionary Junod (1968b:475), writing about the Tsonga early in this century, observed "three great causes of disease", carefully distinguishing "...defilement from death or from impure persons" from attribution of illness to witchcraft and spirits. Although pollution beliefs meet the requirements of naturalistic thinking according to Foster, his is not the only definition in use by anthropologists. Since pollution involves "dangerous heat", negative forces associated with death and other empirically-nonverifyable elements, we might prefer to characterize pollution beliefs as quasi-naturalistic rather than fully naturalistic.

With pollution, people find themselves in "dangerous states," often through no fault of their own, that put them in a socially marginal condition (in part because they are contagious), subject to various taboos, "particularly concerning sexual intercourse" (Hammond-Tooke 1989:91). Pollution beliefs are highly developed among Nguni-speakers (e.g. the Zulu) for whom pollution is related to menstruation, miscarriage, the death of a husband or child, and sexual intercourse, the last said to generate "dangerous heat" (Hammond-Tooke 1989:92; Ngubane 1977). Sources of pollution for the Tsonga include menstrual blood and birth-related bodily fluids, death including abortion and miscarriage, the birth of twins, sickness itself and physical contact with a new environment (Junod 1962b; Honwana 1994; Green 1994b). I have also found illicit sexual acts, such as adultery, to be considered highly polluting among the Tsonga, Shona, Bemba and others.

Some anthropologists have developed typologies--as distinct from dichotomies--that seek to characterize African health belief systems in terms of causal theories of illness. One such is Hammond-Tooke (1989:46) who suggests:

the traditional world-view of South African Bantu-speakers can be broken down into four broadly-defined sets of theories purporting to explain the human condition. These are: the Supreme Being, ancestors, witches and pollution beliefs.

It may be noted that most anthropologists working among southern Bantu-speakers do not report the attribution of illness to a Supreme Being, except possibly among certain Christian and Moslem converts. In addition to ancestral displeasure, witchcraft/sorcery and pollution beliefs--about which there is general agreement--others list as a fourth type causal explanation: (1) belief in a hot/cold balance or other "equilibrium model" common in Asia, Latin America and the circum-Mediterranean as well as parts of Africa (Janzen 1989; Foster 1983:20-21;); (2) natural causes ("it just happens"), typified by the group of illnesses designated umkhuhlane by the Zulu (Ngubane 1977:23) and Swazi (Green 1985:279); (3) environmental dangers caused by "traces" or "tracks" of essences of people or animals, noxious substances left by sorcerers, or discarded elements of maladies that come from patients' bodies. Such environmental attributions of illness are found among at least the Tsonga, Swazi and Zulu (Honwana 1994; Reis 1994:S40; Green 1985; Ngubane 1977:24-29). There is disagreement whether these environmental dangers should be classified as "natural causes" or pollution-related, or indeed whether pollution beliefs are a sub-type of naturalistic belief. It should be evident that these categories are permeable and sometimes overlapping. Purification rites, for example, may relate to witchcraft or spirits as well as to pollution beliefs.

Hammond-Tooke (1989:89) suggests that among the four major ethnolinguistic groups of South Africa (Nguni, Sotho, Venda and Tsonga), witchcraft and sorcery is "by far the commonest basis for the diagnosis of major illness and misfortune". Note that the four major groups of South Africa all extend beyond South Africa itself, therefore this generalization should hold for the broader region. Hammond-Tooke (1983:89; 1970:29) even suggests a crude empirical measure of this hypothesis. He examined 100 cases of misfortune among a Xhosa population and found that 73% were explained in witchcraft-sorcery terms (only 44% of people from the same group living in an urban area gave such explanations). Hammond-Tooke offers another researchable measure specifically to assess the central importance of pollution beliefs, namely the use of enemas and emetics. He notes that this is the "technique par excellence" used by the Zulus to expel pollution (1989:93).

I feel this generalization about the predominance of witchcraft-related beliefs may not be true for the Swazi, Zulu, Tsonga and Shona, representing fully half of Hammond-Tooke's major ethnolinguistic groups (Tsonga and Nguni, which latter comprises Zulu and Swazi). For one thing, the invisible snake appears to be a widespread and important belief among at least the Shona and Tsonga societies and this cultural metaphor surely expresses major concern with pollution and purity. Moreover there seems to be ample ethnoghraphic evidence apart from anything to do with the invisible snake that pollution concerns are important--if not predominate--in these four societies. Let us consider some evidence.

Pollution Beliefs in Four Societies

In describing the Tsonga some 80 years ago, Junod (1962a:152-3) spoke of "dangerous impurity" which "contaminates" objects, individuals and whole villages, requiring complex rites of purification. For example, he mentions the "extreme danger attached to the defilement which accompanies death. This uncleanness contaminates the community and can only be removed by collective purification." In my own research in Inhambane, Mozambique, healers explained that a woman who has had an abortion or miscarriage must have no direct contact with any children. She cannot give a child food directly; she must put it on the ground in front of the child. If there is contact, an illness results disease called ndzaka. In biomedical language, the woman is contagious.

Honwana (1994) broadens our understanding of Tsonga pollution by illuminating their ideas of the "pollution" of environments by a number of contaminants: "foreign essences", "traces" of foreign people, "discarded elements" of illness that have been removed, as well as medicines and poisons of sorcery found in the atmosphere against which Tsonga have no immunity. Purification rituals are required to treat pollution of this sort. Such ideas are also common among the Swazi (Green 1992; Reis 1994) and the Zulu as well. Ngubane (1977:24-29) relates such "ecological health" dangers to naturally-caused illness.

It may be noted that the Tsonga and related groups practice enemas and emetics for diarrhea and other children's illness such as chicuna and zombo. In such cases, my healer informants in Inhambane province, Mozambique report that they must first "remove the dirt" from the child's stomach with enema, and then restore the child's strength with another medicine. A frequently-encountered ultimate cause of child diarrhea is adultery. Either the father commits adultery, and then touches the child, thereby passing contamination in the form of "heat" to the child; or the mother commits adultery and passes "contaminated breastmilk" to the child through nursing (see also Sousa 1991). Purgative teas may also be used for child diarrheas. Tsonga healers further report that the "dirt" of a sexually transmitted illness such as chicasameti can be passed to another person during intercourse, unless first removed by enema.

Although our best source for ethnomedical data on the Zulu, Ngubane (herself a Zulu), apparently does not attempt to measure the relative importance of the four main types of illness causation she identifies (natural causes, sorcery, ancestors and pollution), she devotes considerable discussion to pollution beliefs. Hammond-Tooke himself writes (1989:92) that the Zulu appear to have "the most developed conceptual system" related to pollution in South Africa.

Enemas and emetics are used by the Zulu for a wide range of conditions including diarrhea and stomach disorders, menstrual problems, barrenness, "chest" problems, nausea, general debility and/or "body pain believed to result from excessive accumulation of gall..." (Ngubane 1977:107-8). But perhaps other medicines are used even more frequently to treat witchcraft or sorcery conditions? Ethnobotanist A.B. Cunningham actually conducted a quantitative survey of traditional herbal medicines sold in Natal, a predominantly Zulu area, and found that:

...the largest volumes of herbal medicines for sale...were the four major categories of mixed species: ubulawu (1,966 bags/yr), intelezi (1,924 bags/yr), ikhubalo (1,883 bags/yr) and imbiza (1,211 bags/yr." (All of these) are used in cleansing the body of 'pollutants' encountered in the crowded, stressful and highly competitive urban environment through enemas...or emetics. (Cunningham 1994; 1993)

Cunningham notes, "A valuable feature of market surveys is that they provide a record of health concerns unbiased by reluctance on the part of respondents or patients to reveal what they might be reluctant to discuss with staff of a hospital..." It is rare to have this sort of ethnobotanical corroboration for hypotheses pertaining to ethnomedicine; data of this sort for the other societies we are considering would be most useful.

Based on my five years of fieldwork in Swaziland between 1981-92, I found enemas and emetics widely used among the Swazi for diarrheal and other stomach-centered illness (for prevention as well as treatment) including ridding the body of "bile"; for sexually-transmitted illness such as gcunsula (probably syphilis), idrop (probably gonorrhea), imbune, lugola, tilonza and timvilapo; to prepare a woman from childbirth, to cleanse impurities from an initiate diviner-medium in order to prepare for revelatory dreams, visions and spirit possession (cf. Janzen 1989:233); and to treat a variety of illness such as madness (tilwane), poisoning (sidliso), harmful spells (e.g. umklwebho), bilharzia (umthundungati), sterility, impotence, headaches, and influenza. Even "Zionist" Christian faith healers (baprofeti) may use holy water to administer enemas. Some prominent traditional healers list enemas (kucatseka) and purgatives (kuhlanta) as being generally their most common treatment method. Yet these practices are not only in the hands of traditional healers; there appears to be considerable purgative self-treatment for common conditions. Washing the external body with special purifying medicines, a process known as kugeza, is another commonly used method of treatment related to pollution/purity beliefs (Green 1992).

What of the relative importance of witchcraft and pollution beliefs in Swaziland? I have written about this:

Turning to the Shona of Mozambique, they appear not to rely on enemas and emetics, Hammond-Tooke's measure of the importance of pollution beliefs, in treatment of diarrheal disease. Yet pollution beliefs seem clearly to predominate over witchcraft beliefs in two presumably representative areas of ethnomedicine: diarrheal and sexually transmitted illness. During two months of data gathering in Mozambique, my colleagues and I only encountered one type of diarrheal illness whose etiology involved an evil or avenging spirit. There were likewise only one or two references to witchcraft in descriptions of the more serious diarrheas. The great majority of diarrheal illnesses was attributed to quasi-naturalistic causes, specifically to pollution (Green, Jurg and Dgedge 1994:17).

The same was true for sexually transmitted illness (STI). Traditional healers speaking Shona-related dialects recognize two broad categories of STIs: siki and nyoka-related. The former seem to correspond with the more serious common STDs of Western biomedicine--syphilis, gonorrhea, chlamydia and chancroid--and are believed to be caused by a common invisible, microscopic agent (khoma), or by direct contact with genital discharges that contain khoma. Khoma is conceptualized as a type of "dirt" which is contagious. This "dirt" can, for example, be "eaten" by a fetus in the womb and cause death of the fetus. Some healers treat siki by removing and burying the illness; when a person later passes over the spot where it is buried he can become "contaminated". Nyoka-related illnesses seem to include less serious, perhaps self-limiting, genito-urinary infections. These are also understood in terms of traditional ideas of pollution, as might be expected from their association with the Gaurdian of Bodily Purity (Green, Jurg and Dgedge 1993:261).

Turning to Shona speakers in Zimbabwe, Gelfand et al (1985:70) tell us that traditional healers believe that the cause of an illness must be determined before treatment. A survey of 250 urban and 49 rural patients of traditional healers showed that about 17% of illnesses were attributed to ancestors; 28% to witchcraft, and about 6% to angry or foreign spirits. But what of the remaining 49% of illnesses? "Natural" causes are specified most frequently but there is no specific discussion of pollution beliefs. I suggest this is because such beliefs have not been specifically researched or disaggregated from "natural causes". Gelfand et al (1985:85-6,94) do list a number of purgative and emetic herbs used by the Shona, presumably for conditions in addition to diarrhea.

Of course I have not disproved Hammond-Tooke's hypothesis about the predominance of witchcraft beliefs in South Africa. Yet I have presented evidence from four societies in or contiguous to South Africa where pollution beliefs seem to predominate over those relating to witchcraft. Maybe our view is skewed because of the nature of much of the research reviewed here. Diarrhea and STDs both involve bodily fluids and discharges related to digestion and procreation, which can be polluting due to their power to "symbolize social relations and social processes" (Douglas 1992:125). On the other hand the view of anthropologists who emphasize the importance of witchcraft and sorcery may be skewed. Hammond-Tooke himself admits that pollution-related beliefs have been under-researched. It would therefore seem prudent to await further research before concluding that witchcraft beliefs predominate in illness/misfortune explanations for South African or southern Bantu-speakers. I would suggest here that the presence of belief in an invisible snake (as distinct from a visible worm) may be as good a predictor of the importance of pollution beliefs as evidence of use of enemas and emetics. At least both factors should be considered, rather than only the latter. Our preliminary evidence suggests the invisible snake is a potent cultural metaphor the analysis of which can provide deep insights into ethnomedical thinking. I recommend that researchers seek to establish the existence and nature of the invisible snake in societies where ethnomedical research is undertaken.

Public Health Implications

What does it matter whether witchcraft or pollution beliefs predominate in a given society? Apart from the light shed on comparative ethnography and anthropological theory-building, we should not overlook the public health significance of answering such a question in Africa, where health education and other interventions have often met with little success. Pollution beliefs represent an area of potential inter-sectoral interface between indigenous and cosmopolitan medicine. Both are concerned with cleanliness, environmental sanitation, prevention of contact with impurity, contagion and its avoidance, the danger of a "dirt", the value of clean, pure food that agrees with the consumer, and the like. Both traditional and modern medicine agree that within this domain, the cause of illness is impersonal and in fact relates to conditions that may be modifiable. There seems to be potential interface in ideas about both contagion and prevention, although this needs to be much more fully explored in the context of individual ethnomedical systems. In any case, if public health messages can be framed in terms appropriate to those for whom pollution-causality beliefs are important, they would rest squarely on common ground that already exists.

Furthermore, there appears to be less danger or misgiving (compared to witchcraft-causality beliefs) that such messages would: (1) perpetuate witchcraft beliefs or the scapegoating of others--a fear often expressed by health officials in Africa; (2) impede acceptance of scientific thinking and perpetuate magical thinking (Motlana, quoted in Freeman and Motsei 1990:7; Velimirovic 1981; Velimirovic and Velimirovic 1978:182-3; Asuni 1979); or (3) retard a patient's self-awareness or psychological maturation, as Asuni (1979) and others believe is still another negative consequence of reliance on indigenous African healers. Undue focus of anthropologists and others on witchcraft beliefs and practices, probably the area of least inter-sectoral compatibility, has not contributed to the incorporation of ethnomedical findings in public health programs--something many anthropologists bemoan as a serious oversight. The view of health officials is something like, "We can't build on traditional heath beliefs because they rest mainly on witchcraft superstitions, which are dangerous and socially-divisive, so lets simply start with a clean slate and teach what we know from modern medicine." I suggest that if the importance of pollution beliefs in contagious diseases--the diseases accounting for most of the morbidity and mortality in Africa--were better understood, we could more easily promote practical use of ethnomedical findings in health programs. This is because there is relatively high compatibility between quasi-naturalistic pollution beliefs and the parts of medical science emphasized by public health.

It seems missionaries working in southern Africa have already adopted an approach that builds upon, rather than ignores or confronts, pollution beliefs. Janzen (1989:234) notes:

The issues raised in this paper are complex and clearly there needs to be more research focused on the topics suggested here in order to better answer questions pertaining to the nature of the invisible snake, its relationship to pollution beliefs, the validity and utility of illness causality typologies, and the relative importance of pollution as distinct from witchcraft beliefs in a given society. Certainly there needs to be more exploration on the part of health educators of the common ground between traditional and Western, cosmopolitan medicine that pollution beliefs seem to provide.

Acknowledgments The author's research in Mozambique was funded by the Swiss Development Cooperation, the European Community, and the Agency for International Development (AID). Research in Swaziland, South Africa and Zambia was funded by AID. Opinions expressed are strictly the author's. The author wishes to thank Josefa Marrato and other colleagues in the Mozambique Ministry of Health, as well as four anonymous reviewers for Medical Anthropology.

References Cited

Asuni, T. "The Dilemma of Traditional Healing with Special Reference to Nigeria," Social Science & Medicine, 1979, Vol. 13, pp. 33-9.

Booyens, J.H., Aspekte van Populere Opvattinge oor Diareesiektes onder Tswanasprekende Stedelinge (Aspects of the Popular Attitude about Diarrhea among Tswana-speaking Urbanites). Curationis Dec:12(3-4):11-16, 1989.

Cunningham, A.B., Personal communication, 12/11/94.

Cunningham, A.B., Imithi-isiZulu: Trade in Traditional Medicines in Natal/KwaZulu. M Soc Sci thesis, University of Natal, Durban, 1993.

Douglas, Mary, Purity and Danger: An Analysis of Concepts of Pollution and Taboo. London: Routledge, 1992 (1966).

Earthy, E.D., Valenge Women: The social and economic life of the Valenge women of Portuguese East Africa. London: Frank Cass & co LTD, 1968.

Foster, G., "Introduction to Ethnomedicine" in R. Bannerman, J. Burton & Ch'en Wen-Chieh, Traditional Medicine and Health Care Coverage. Geneva: WHO 1983, pp17-24.

Freeman, Melvyn and M. Motsei, "Is There a Role for Traditional Healers in Health Care in South Africa?" The Center for the Study of Health Policy, Department of Community Health, University of the Witwatersrand, Paper No. 20, Johannesburg, March 1990.

Gabinete de Estudos de Medicina Tradicional (GEMT), "Primeiro Seminário de Colaboraçao entre os Sectores de Saúde Moderno e Tradicional de 13 a 19 de Maio de 1993, Xai Xai, Moçambique."

Gelfand, M., S. Mavi, R.B. Drummond, and B. Ndemera, The Traditional Medical Practitioner in Zimbabwe. Harare: Mambo Press, 1985.

Gibbs, Sara, "Mozambique: from Dependency to Sustainability. What Role for traditional healers in the Move Towards a More Sustainable Health Service? BA thesis (geography) Univ. of Liverpool, U.K. 1992.

Green, E.C., AIDS and Sexually Transmitted Disease in Africa: Bridging the Gap Between Traditional Healers and Modern Medicine. Boulder, Co. and Oxford, U.K.: Westview Press, 1994a.

Green, E.C., unpublished Mozambique fieldnotes (1990-94), 1994b.

Green, E.C., unpublished Swaziland fieldnotes (1981-92), 1992.

Green, E.C., "Traditional Healers, Mothers and Childhood Diarrheal Disease in Swaziland: The Interface of Anthropology and Health Education," Social Science and Medicine, Vol. 20, No. 3, pp. 277-2-85, 1985.

Green, E.C., A. Jurg, and A. Dgedge, "The Snake in the Stomach: Child Diarrhea in Central Mozambique." Medical Anthropology Quarterly Vol. 8, no. 1, 1994, pp. 4-24.

Green, E.C., A. Jurg, an A. Dgedge, "Sexually Transmitted Diseases, AIDS and Traditional Healers in Mozambique" Medical Anthropology Vol 15, No 3, Spring 1993, pp. 261-281.

Green, E.C., A. Jurg, T. Tomas, and A. Dgedge, Traditional Health Beliefs and Practices Related to Child Diarrheal and Sexually-Transmitted Diseases. USAID and the Academy for Educational Development, Washington, D.C., Dec 1991.

Hammond-Tooke, W.D., Rituals and Medicines, Johannesburg: A.D. Donker ltd., 1989.

Hammond-Tooke, W.D., Boundaries and Belief: The Structure of a Sotho Worldview. Johannesburg: Witwatersrand University Press, 1981a.

Hammond-Tooke, W.D., "Patrolling the Herms: Social Structure, Cosmology and Pollution Concepts in Southern Africa."

Raymond Dart Lectures 18, Johannesburg: Witwatersrand Univ. April 1981b.

Hammond-Tooke, W.D., "Urbanization and the Interpretation of Misfortune: A Quantitative Analysis," Africa, Vol. II:1, 1970, p. 25-39.

Honwana, Alcinda, Nyamusoro: Spirit Possession, Ritual and Social Change Among the Tsonga of Southern Mozambique, Ph.D. dissertation (draft summary), University of London, 1994.

Janzen, John, "Health, Religion, and Medicine in Central and Southern African Traditions." in Caring and Curing: Health and Medicine in World Religious Traditions. L. Sullivan, ed. New York: MacMillan, pp. 225-254, 1989.

Junod, H., The Life of a South African Tribe. Vol. I. Social Life. New Hyde Park, NY: University Books, 1962a (orig. 1926).

Junod, H., The Life of a South African Tribe. Vol. II. Mental Life. New Hyde Park, NY: University Books, 1962b (orig. 1926).

Jurg, A., R. Albino, J. Benate, et al, Conhecimentos, Atitudes e Práticas no ambito de Água e Saneamento nos Arredores de Cuamba. Agua Rual/MCDI, Maputo, Mozambique: June 1994.

Krige, E.J., and J.D Krige The Realm of a Rain Queen. London: Oxford Univ. Press, 1943.

Maina-Ahlberg, B., Beliefs and Practices Concerning Treatment of Measles and Acute Diarrhea Among the Akamba. Trop. Geog. Med. 31:139-148, 1979.

Marrato, Josefa, "Condições de Enquadramento Familiar da Criança Deficiente Mental". Maputo: Instituto Superior Pedagógico, (tese), Nov., 1991.

Ngubane, Harriet, Body and Mind in Zulu Medicine. London: Academic Press, 1977.

Panizzo, Enny, "Pesquisa sobre Convulsões na Crança. 1a Parte: Maputo." Maputo, Moçambique" Ministry of Health, 1994.

Reis, R., "Evil in the Body, Disorder of the Brain" Tropical & Geographical Medicine, 46 (3) (suppl.):S40-43 1994.

de Sousa, Julio F., Traditional Beliefs and Practices Related to Childhood Diarrhoeal Disease in a High-Density Suburb of Maputo. B.A. Thesis. Harare: Department of Sociology, University of Zimbabwe, 1991.

Velimirovic, B., Traditional Medicine is not Primary Health Care: a Polemic. Curare, vol. 7(1) 1984:61-79.

Yoder, P. Stanley, Cultural Conceptions of Illness in the Measurement of Changes in Morbidity. in The Health Transition: Methods and Measures, J. Cleland and A. Hill, eds. Canberra: Australian National University, pp. 43-60, 1991.

Subscribe to this comment's feed

Write comment