struggles for control

Struggles for Control: The Social Roots of Health and Healing in Modern AfricaAuthor(s): Steven FeiermanSource: African Studies Review, Vol. 28, No. 2/3 (Jun. - Sep., 1985), pp. 73-147Published by: African Studies AssociationStable URL: http://www.jstor.org/stable/524604 .

Accessed: 05/12/2013 23:28

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

African Studies Association is collaborating with JSTOR to digitize, preserve and extend access to AfricanStudies Review.

http://www.jstor.org

This content downloaded from 129.105.215.146 on Thu, 5 Dec 2013 23:28:21 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=afsta

http://www.jstor.org/stable/524604?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


STRUGGLES FOR CONTROL: THE SOCIAL ROOTS OF HEALTH AND HEALING

IN MODERN AFRICA

Steven Feierman

This paper is a general interpretation of the social determinants of health and health care in Africa over the past century. The evolution of health cannot be separated from the broader story of social change. The political and economic forces which shaped the continent's history also established the framework within which patterns of diagnosis and treatment, health and disease, emerged.

The implication of this is that healers of all kinds-whether doctors or "traditional healers"-have been less influential than we commonly think in shaping states of health or in healing the sick.

This position opens up a range of difficult problems which are addressed in the following pages. What is the exact nature of the link between the broad political-economic forces and the distribution of health or disease? Which of these forces have driven therapeutics along its historical path, and by what means? What role do healers actually play?

The body of the paper is divided into three main sections. The first explores the micro-sociology of changing treatment of illness. In most African communities several kinds of healers work side by side: physicians or medical assistants, specialists in sorcery or spirit possession, Christian or Muslim religious healers, and others. Multiple authorities co-exist, and therefore no one healer decides the cause or cure of illnesses in a way which others accept as beyond challenge. But treatment cannot exist without coordination. Someone must decide on a course of action when lives are threatened. It is most often a loose network of the patient's relatives and neighbors who make this decision. The history of therapeutics must therefore take account of all the forces which shape local networks, in other words everything which affects community and domestic organization. The history of health care is inseparable from the total history of communal organization and of the economy.

The second major section tries to understand the social context of health and disease, life and death, population growth and decline, in Africa over the past century. It is sub-divided into two halves. The first half reviews the literature on early colonial population decline, and more recent population growth. The second half, which is crucial for the entire argument of the paper, identifies types of political and economic decisions which have an impact on the distribution of

African Studies Review, vol. 28, nos. 2/3, June/September 1985.

73



74 AFRICAN STUDIES REVIEW

sickness and death. Death is not distributed in a random way. It comes sooner to poor people than to the rich, sooner to people in the country than in the city, and sooner to the children of absent migrant workers than to children of women who live with their husbands. These inequalities are a consequence of decisions by power holders on the distribution of social costs-who is to suffer the disease costs of large-scale irrigation, which children are to be malnourished, which workers are to lose their health at work, and which farmers to be hungriest in the months before they harvest their new crops. Readers who skim through this paper need to read the section titled "Social Costs of Production" if they are to understand the general argument.

The first two sections ask about what healers cannot do-the important things over which they have little influence. The third looks at the impact they have within the larger political and economic framework. It asks first whether the therapies of either biomedicine or popular healing are effective.' In what way? Is one body of knowledge and techniques more useful than the other? The essay then explores the history of healers over the past century, and their social role today. To what extent can healers change the distribution of social factors which shape health and disease? Healers, when they are at their most effective, play a mediating role in society. They have the freedom, on occasion, to ally themselves with groups of lay people who want to improve the social conditions of health. By choosing allies and issues, they can improve states of health in the wider society.

A study on national health policies, another on control over sexuality, and yet another on the psychiatric or psychological literature would have fitted into the perspective of this paper. They are not covered because the essay is too long as it is; I hope to write future essays on those subjects. The paper concentrates on literature published after 1970 to give a sense of current debates, but refers to earlier publications when necessary. It concentrates on Africa South of the Sahara. Its coverage of the social costs of production is limited to a few cases which are, nevertheless, important ones for the total picture of African health.

I. PATTERNS OF CONTROL:

THE INTEGRATION OF HEALTH, HEALING, AND IDEOLOGY

No colonial power and no independent African state has ever intervened decisively to destroy popular healing. Governments have been either unable or unwilling to provide biomedical care to their entire populations, and have therefore been forced to tolerate the survival of African healing. This fortunate outcome is also a consequence of the determination and cunning exercised by popular healers over generations. Healers have long cultivated secretiveness as a survival strategy. Harriet Ngubane (1981), for example, tells the story of Zulu healers who created tight guilds invisible to the South African authorities.

The survival of popular healing, the spread of competing popular therapies over wide areas, and the introduction of biomedicine, Christian healing, and Muslim healing, have combined to create a profusion of therapeutic forms in contemporary Africa. In a single place, many kinds of practitioners co-exist: African physicians, dispensary aides, Christian prophets, Muslim teachers, spirit possession mediums, specialists in sorcery, diviners of all kinds, herbalists, faith healers, and specialists in kinship therapy and in the removal of pollution.



HEALTH AND HEALING IN MODERN AFRICA 75

The patient has access to the broadest possible range of therapeutic alternatives, leaving the scholar to grapple with one of the most difficult and important puzzles in the study of African healing: who controls the therapeutic process?

Control over healing carries with it power over other practical matters. The person who makes an authoritative diagnosis decides when a person is too sick to work or to meet obligations to relatives and neighbors. The decision on sickness leads to others-that relatives ought to pay for treatments, for example, or to provide nursing care.

Control over healing is important also in shaping ideology. The power to name an illness, to identify its causes, is also the power to say which elements in the experience of life lead to suffering. This is not a new point. It was one of the central themes of a generation of anthropological studies which interpreted witchcraft as a system of beliefs sustaining and stabilizing the social order (reviewed in Douglas, 1970).

Recent writings by Allan Young (1982b), Michael Taussig (1980), and Jean Comaroff (1982) argue a similar case transposed into a different historical and sociological framework. According to Taussig the relationship between doctor and patient is a "social interaction which can reinforce the culture's basic premises in a most powerful manner." Serious illness interrupts the everyday routine, and the uncritical acceptance of the meaning of life. "This gives the doctor a powerful point of entry into the patient's psyche" for the doctor offers the patient an interpretation of what is happening in his or her own body. Taussig argues that American doctors working under capitalist conditions interpret disease as a thing, an object separate from social relations-one which resembles a commodity. The doctor's interpretation makes illness an individual experience rather than a social one, and denies the importance of the social conditions out of which disability emerges. According to Taussig, Zande healing, by contrast, accounts fully for the social context of health problems.

Taussig's general point holds, even if his central contrast is oversimplified in three important ways: he romanticizes Zande healing, in which divination served to support the domination of aristocrats over the common people (Young, 1982b: 276; McLeod, 1972); he does not discuss commoditized popular healing in contemporary Africa, where some healers work impersonally and entrepreneurs mass-produce herbal remedies (Frankenberg, 1980; Abdallah, 1981b); and he ignores the contribution of those who struggle to create therapeutic alternatives in industrial countries-alternatives which tend, admittedly, to treat health problems as individual disorders divorced from their social context (Guttmacher, 1979).

The struggle to create alternatives illustrates the ideological role of those who control healing. Therapy shapes ideology by interpreting the patient's experience of illness. But values are not narrowly restricted to the sphere of medicine-they pervade society as a whole. The person who controls therapy serves as a conduit transmitting general social values, but is also capable of reshaping and reinterpreting those values in the healing process.

The Cultural Interpretation

The literature of African therapeutic systems provides three very different sets of answers-overlapping in important ways-on how therapeutic choices are




made, and therefore on the nature of control. The first is cultural, with an emphasis on shared values rather than on who, in particular, is in control. Each society, the argument goes, has its own body of cultural knowledge for the interpretation of illness. The patient's condition is interpreted in light of this knowledge in order to name the condition and arrive at a course of therapy.

Scholarly observers differ on the extent to which the body of cultural knowledge forms a coherent medical system, ranging from Robin Horton, who finds a closed set of causal categories, to John Janzen (1983) who rejects what he calls "systems monism" and argues that multiple healing traditions co-exist side by side in a pluralistic system, to Murray Last (1981) who writes that the Hausa Maguzawa engage in therapy while neither knowing nor wanting to know about a coherent medical tradition.

Horton's "African Traditional Thought and Western Science" (1967) makes the clearest case for folk thought as a closed system. Horton returns to Evans- Pritchard's question: how can a system of causal explanation persist if it is not verifiable when tested against experience. Horton argues that African traditional thought forms a tight system from which escape is impossible. He quotes Evans- Pritchard on witchcraft and oracles: "In this web of belief every strand depends upon every other strand, and a Zande cannot get out of its meshes because it is the only world he knows." By contrast to this closed system, Horton's argument goes, western scientific thought constantly tests its assumptions against experience. Tola Olu Pearce (1983) and others maintain that African medicine can be tested against experience and found effective. In her view, the oral transmission of medical knowledge, because of its need for economy, may be more successful at passing on medical prescriptions than at recording their empirical basis. Certainly, work on African medicinal plants shows that an empirical basis exists (Soforowa, 1982; Watt and Breyer-Brandwijk, 1962).

On Horton's central point, contrasting closed and open systems, I argue elsewhere (Feierman, 1979) that European medicine is not a fully open system, nor is African medicine closed. Scientific medicine has achieved monopoly power in Europe and the U.S., where non-scientific therapeutic alternatives are systematically excluded. Medical authorities tend not to test alternative therapies for their efficacy. Those who actually manage the process of therapy in Africa, by contrast, move pragmatically from one type of healing to another. In other words, Horton stresses the coherence of closed patterns of causal explanation and does not acknowledge the coexistence of multiple explanations in each African society and in each individual's thought.

Even where scholars identify a core medical tradition, ethnic groups tend not to be clearly bounded entities, and practitioners frequently cross local lines of language and culture. For example, Charles Good (1980: 14), following Fabrega (1977) defines an ethnomedical care system (EMCS), "conceived as the whole approach of an ethnic community to disease and illness" including the work of diviners, M.D.'s, and herbalists. But Good (1980: 42) then finds "frequent consultation across ethnic lines and social classes."

It is not clear whether pluralism characterized the African healing of earlier generations, or whether it is a consequence of this century's medical competition. In one isolated and therefore inconclusive case described by Buxton (1973) the Mandari, who at the time of her study had almost no access to doctors or dispensaries, still identified therapeutic conceptions which were not coherently integrated into the core tradition. Pluralism in this case was not a consequence




of the intrusion of biomedical practice. One common goal of cultural studies of African medicine is to understand the

indigenous logic of classification so thoroughly that the analyst, given a set of symptoms, can predict the folk diagnosis. In practice this is nearly impossible, for in any therapeutic culture the healer who assesses the patient's condition must organize ambiguous information. In biomedical practice, for example, physical signs do not usually lead ineluctably to a particular diagnosis or treatment. In many actual cases the signs are not clear, or they are contradictory. A fascinating ethnography of medical experimentation in the U.S. concludes that "in a sense every clinical act is an investigation" (Fox, 1974: 238). The likelihood of ambiguity is at least as high in the practice of popular healers who must assess both the physical signs (Frake, 1961; Warren, 1974) and the totality of the patient's social situation (Turner, 1968). A healer might, for example, assess the patient's stomach cramps, identify herbs which might cure them, but then also consider the fact that the patient is a married woman whose husband never paid bridewealth.

In addition to the patient's physical signs and social relationships, a third factor is crucial to the unfolding of therapeutic action. This is the passage of time. The cultural interpretation of illness changes as the illness itself changes through time. In many cases, illness episodes which are interpreted as "natural" at an early stage are later explained in terms of sorcery or spirit causes (for a contrary case see Ngubane, 1977: chapter 2). Kramer and Thomas (1982) demonstrate that Kamba diagnosticians give radically different interpretations of cases which continue for less than one month, between one month and two years, and more than two years. Chavunduka (1978: 38-40) explores the comparative diagnosis, among the Shona, of critical incapacitating dysfunctions as opposed to chronic non-incapacitating dysfunctions, only to find that his analysis is confounded by differences in diagnosis of a single illness at an early or late stage. According to Christopher Davis-Roberts (1981: 315), the relationship of experience to time is at the core of the cultural construction of illness. "Illness brings to bear upon the family of the patient the pressure of time speeded up. Instead of basing their choices upon their own necessities and convenience, people must move to the pace set up by the progress of the disease . . . the body ... is but the artifact of the processes unfolding, blossom-like from within it." It is this unfolding which is the central organizing process-the unfolding rather than the symptoms at any one moment which the healers must explain.

Two of the most influential medical anthropologists (neither of whom writes about Africa) discuss illness and therapy as a broad interpretive process, and not simply the selection of the most appropriate taxonomic category for each set of physical signs. For Byron Good (1977) and for Arthur Kleinman (1980) healing is a process in which the individual's experience is interpreted in relationship to an entire network of symbols. According to Good's interpretation, as the illness unfolds through time therapists locate it in a network of core symbols. In the particular illness he studies the condition is linked to childbirth, miscarriage, pregnancy, age, sadness (see Young's interpretation of this, 1982b). For Good and for Kleinman the broader process of cultural interpretation is at the heart of healing-more crucial to it than finding the right narrow label for a particular kind of skin rash.

Sargent (1982: 93) brings Kleinman's interpretation to bear on the question of therapeutic choice. Because healing is a process of general cultural interpretation,




she argues, patients choose healers with whom they share fundamental assumptions regarding clinical reality. Sargent appears to assume that if the patient's view of clinical reality diverges significantly from that of the healer, it becomes impossible to locate the illness in the core semantic network.

But in most chronic illnesses, for most parts of Africa, patients move back and forth from the hospital, to local healers of one type and then another (Janzen, 1978; Janzen and Feierman, 1979; Janzen and Prins, 1981). How could a single patient, in a single illness, share a sense of clinical reality with such diverse therapists? Part of the answer is that the patient does not come to the therapeutic process as an isolated atom. The patient's close relatives make important judgments on the therapeutic process, whether or not patient and relatives agree on a single view of clinical reality. In Sargent's own study they appear to agree, but even here the sharing of clinical assumptions between parturients and birth attendants is likely to be a secondary consequence of the fact that 77 percent of birth attendants live in the same household as the parturient (Sargent, 1982: 99). The role of relatives in choosing a therapy is clear in the work of Chavunduka, who found that neither income nor education was a factor in determining whether Shona patients were treated in a hospital or by a popular healer (1978: 45). The reason is that "When a person is taken ill, he does not often act on his own. His kinsmen both educated and uneducated jointly take the decisions throughout the illness and are responsible for paying the medical fees." In the Shona case, and I suspect in most parts of Africa, it is not the patient as an individual whose view of clinical reality is in question; the entire group of relatives works out a shared view of clinical reality, with maximum tolerance of diversity.

The Interpretation Based on Control by Healers The second body of literature on therapeutic choice in Africa analyzes the

nature of competition among alternative sets of healers. Over the course of time, new healing alternatives become available to ordinary people living in any particular locality. Hospital medicine is, of course, not the only innovation. African healing churches, Islamic healing, new forms of witch-finding, and the diffusion of cults of affliction are all part of the changing picture (Redmayne, 1970; Parkin, 1968; MacGaffey, 1983; Abdallah, 1981b; Omoyajowo, 1982; Peel, 1968; Janzen, 1982). Only in a few limited contexts do governmental authorities impose a requirement that patients must use a particular therapy, as when national law requires that cases of cholera or of tuberculosis be treated biomedically, or that only doctors can excuse the work absences of government employees. But in most illnesses the range of therapeutic choice is very wide.

Much of the literature assumes that health specialists, whether physicians or popular healers, control therapy. Parsonian sociology plays an important part in this assumption. Parsons treated the therapeutic relationship as one of authority and legitimation. Since sickness exempts a person from some responsibilities, there needs to be a mechanism to separate legitimate from illegitimate exemptions, people who ought to be recognized as sick from those who ought not (1951: 436-37, 446-47). For Parsons it is the physician, with his autonomous technical knowledge, who is the ultimate arbiter. The sick person's family plays a role in care, but has no competence to assess the patient's condition or suggest treatment. The Parsonian interpretation has of course been modified, even in the




study of American medicine. Renee Fox and Judith Swazey explore the fascinating ambiguities which emerge when physicians balance ethical considerations against technical ones in acting as gatekeepers for organ transplantation (1974).

The literature about Africa often assigns the role of the Parsonian physician to the diviner, on the assumption that diviners make authoritative judgments about illness causes. But the parallel is not precise. The diviner usually does not make the initial differential diagnosis, does not define the sick role, and often does not prescribe a single course of therapy.

By the time the patient's relatives visit the diviner they have already decided that the illness is one which has a serious cause-one which in many Bantu languages would be characterized as "an illness of man," in other words an illness with a moral cause. The relatives narrow their diagnosis before divination. Even where divination is extremely important, as among the Yoruba, it is only one step in an extended series which family, neighbors, and healers take to help the patient (Maclean, 1971: 17-20).

In most places it is not the diviner's job to legitimize the sick role. The diviner answers the question, "Why is this person ill," and not "Is this person ill?" (Adler and Zempleni, 1972: 74).

The diviner often does not decide, in an autonomous way, on the course of therapy management. His or her conclusions are often ambiguous, either because of the structure of divinatory statements (Adler and Zempleni, 1972: 144-45), or because those who manage therapy are not satisfied with the diviner's assessment of cause, and consult a second diviner (Whyte, 1982: 2059). Ngubane (1977: 39), for example, describes a Zulu case in which a lay person makes a sorcery accusation but then angrily rejects the diviner's choice of sorcerer. Turner (1975: 215) heard of instances where diviners were speared by the angry relatives of people declared to be sorcerers.

Most important of all is the suggestion of Sindzingre and Zempleni (1981: 287) about the Senoufo, with applicability over a much wider area-that divination is separate from therapy. Those who organize therapy consult a diviner so as to understand the social, natural, or mystical context of an illness, but not for the purpose of organizing the healing process. They choose treatments by trying first one cure and then another, in a pragmatic way.

One possible way for scholars to sort out the pragmatic therapeutic choices is to ask which treatment works better in each particular case-which is the most efficacious? But this is difficult because efficacy is not limited to the effects of therapy on symptomatic behavior, or to effects which are reproducible under controlled conditions (Young, 1982). The efficacy of a therapy might be judged by its effect on social relations-as when an American who does not wish to be called mentally ill chooses to be treated by an internist for mild hypertension rather than by a psychiatrist for associated emotional distress. Patients can also judge efficacy on the basis of technical features of the therapy-whether the healer's purgatives and emetics have short term effects, and whether the practitioner is able to control dangerous side effects. They can also judge the accuracy of prognosis in the absence of cure-as when a physician accurately sums up what is about to happen in the life of an incurable cancer patient. Even more fundamental than this is the emphasis in many African therapies on restructuring social relations and on the emotional context of illness. This gets at the heart of the process of healing, a process poorly understood in "scientific"




terms (Kleinman, 1980; Pearce, 1983). The problem is not that therapeutic efficacy is lacking in African medicine; it

is that diverse healing traditions, each with legitimate claims to efficacy, co-exist with little capacity to exclude one another from the range of practical options. Ministries of health and national medical associations can define who is or is not a legitimate physician, but cannot define a legitimate popular healer. Popular healers themselves lack authority to exclude practitioners who are ill-trained, unethical, or incompetent. This is because popular healers do not have access (and possibly cannot have access) to government power, which alone is capable of prohibiting forms of medical practice (as discussed in Section III). In the absence of centralized public authority it is the patients and their relatives who make the crucial decisions on therapy.

The Interpretation Based on Lay Control

This, then, leads to the third approach to control over therapeutic decisions, one in which it is the sick person's relatives and neighbors who choose among the many therapeutic options. According to this approach the pattern of therapy management is an inversion of the one a Parsonian would expect, for the people most qualified to make therapeutic decisions have the least technical knowledge of medicine and the greatest personal knowledge of the patient.

Two books, both published in 1978, made a major breakthrough in understanding lay authority over African healing. In that year Gordon Chavunduka published Traditional Healers and the Shona Patient (a revision of a 1970 Ph.D. thesis) emphasizing the role of the patient's relatives in a large city. John Janzen published The Quest for Therapy in Lower Zaire painting a similar picture for a rural area. Janzen provides a series of extended case histories, so that other scholars who want to analyze the implications of his approach have materials on which to work. He also tries to describe the approach in terms which lead to comparative research, for he emphasizes the importance of what he calls the "therapy managing group." I intend, here, to subject Janzen's work to the most searching criticism possible, because it is such a strong work. If any of us are to build on Janzen's approach, we will need to begin with a critical understanding.

In the accounts given by both Janzen and Chavunduka, a patient whose illness lasts over a period of time moves through the broadest possible range of therapeutic alternatives. In the case of the young woman Lwezi Louise, for example (Janzen, 1978: chapter 7), Lwezi fell ill with periodic fever and chills, radiating back pain, pain in her joints, loss of appetite, and general malaise. Her relatives took her first to a dispensary, where she was treated for a couple of months, then to an nganga for daily treatments of razor-blade incisions rubbed in with medicines and combined with both counseling sessions and worship. Then her brothers offered to make a payment which they hoped would win Lwezi her father's blessing, and then a group of Lwezi's maternal and paternal kin accompanied her to the Christian prophet Mama Marie Kukunda. The case history continues through another prophet, further visits to the dispensary, and a period of hospitalization.

It is obvious in this case that the healers did not make the basic decisions on therapy management. Lwezi went to an nganga after discharge from the dispensary even though its staff had not suggested that she do this. The nganga's




recommendation at the end of his treatment was that Lweze return to him again after a three-week interval. Lwezi's relatives paid no attention to his request. They, and not the nganga, held ultimate authority over her therapy.

In Janzen's analysis the "therapy managing group" is the critical decision- making body. Janzen (1978: 134) explains that "unlike the lay referral system in

Europe and North America, which is described as discharging its duties after a professional takes over, the Kongo therapy managing group continues to exercise its authority and frequently even increases it while the sufferer is in the hands of a specialist." It plays this role only through a subtle process of social negotiation, for according to Janzen it must act collectively. "The therapy managing group," he writes (1978: 139), "must achieve internal cognitive agreement and social consensus in order to be an effective decision-making body."

The problem with this analysis is that the groups Janzen describes as managing therapy do not have clear boundaries-they are fluid, their members ever changing. Janzen talks about "quasi-groups" as a way of harmonizing analysis and description, but it is difficult to see how a group which might have different members on each of three successive days can act as an authoritative deliberative body. Some members express opinions which are never translated into action; other members act effectively without consultation. "Cognitive agreement" and "social consensus" are not relevant here except as consensus and agreement on the rules governing general procedures of therapy management. It is important to make a distinction here between Janzen's analysis of the extended case studies, which are subtle, nuanced, and fully contextualized, and the general abstract discussion which reifies therapy managing groups.

Janzen (1978: 130) himself provides the key to understanding the process when he talks about "conveying rights." It is worth quoting at length:

The process of conveying rights of therapeutic decision making has been identified by a MuKongo writer in the following terms: "When an illness occurs, the person who is jurally responsible for the sufferer (muntu vwidi mbevo ) quickly consults a doctor. If and when the consulting doctor has made his diagnosis, it is again this jurally responsible person who sends for the appropriate treating doctor." The term vwidi (from vwa, "to have or possess") denotes the guardianship of a mother's brother, brother, child's father, parents, or slavemaster. It implies proprietary rights over the sufferer in keeping with the legal definition of a conveyance. Thus the lay therapy manager retains the right to choose the therapist even after a consulting doctor-diviner-has made his diagnosis and recommended action. No professional or bureaucratic referral enforcement is possible in this setting.

The paragraph discusses a single therapy manager with jural rights, but then there are other individuals who do not have those rights, but who contribute to the process of therapy management.

This can be illustrated with further details from the case of Lwezi Louise. Lwezi fell ill at a time when she ought to have received her father's blessing, but had not. Her father was upset at not having received the proper gift from the man Lwezi had married in the city. Lwezi's father had a direct responsibility for the blessing. No one else could give it in his place. Lwezi's brothers tried to save their sister's health; they wanted to pay the upbringing gift and bridewealth on behalf of Lwezi's urban suitor. "The pastors on both sides rejected this, noting that it would set Lwezi free, with a blessing, to become a public woman" (Janzen, 1978: 107).




The implications of this series of events are important. They show that the loose unity of the therapy managing group conceals a fundamental division between one set of people who have jural responsibility for the patient's welfare in a given context, and others who offer voluntary assistance but have no rights or obligations in the particular transaction. Janzen does not discuss this distinction. In Lwezi's case her brothers had no jural responsibility for paying the gift or the bridewealth. They wanted to make a voluntary gesture, out of a moral or sentimental obligation to help their sister become well again. At a later point in the same case a nurse at a dispensary asked Lwezi her story and then made her much better by telling her "you can die and leave as a corpse, or you can get better and leave" (Janzen, 1978: 112). Anyone can offer help, but ultimate authority rests in the hands of the person with the clearest claim to jural responsibility.

The second implication of the abortive bridewealth transaction is that the therapy managers rejected, as unacceptable, a course of therapeutic action which promised success but which challenged fundamental authority patterns. The interpretation of the illness in terms of bridewealth and her father's blessing must have reinforced Lwezi's sense that her well-being was linked indissolubly with respect for her father's authority. If Lwezi's brothers had been allowed to pay the bridewealth it would have undermined her father's authority and established a social precedent for reducing the authority of husbands and fathers more generally. When healing and the stability of authority patterns came into conflict, authority won.

Lay therapy management does not have a distinctive institutional hierarchy; it is fully embedded within general patterns of control over domestic and community affairs. The values and assumptions communicated through therapy are usually the dominant local values-in this case patriarchal ones.

When therapeutic cults establish their own hierarchies of control, separate from general kinship authority, they have the capacity to sustain alternative patterns of values, which challenge dominant ideology. It is for this important reason that feminist scholars focus on cults of affliction (see Berger, 1981). Spring (1978) finds that Luvale women's therapeutic cults are based on an ideology which assigns power to women and to matrilineal ancestors.

An understanding of lay therapy management alters the basic terms in which we explain patterns of change in modern African healing. The history of healers, of their ideas and their practices, turns out to be of secondary importance. Healers of all kinds merely present therapeutic options from among which those in control choose. The people who make the most crucial therapeutic decisions tend not to have expert medical knowledge, whether in the biomedical or popular medical sphere. When deciding on therapy they tend not to separate considerations of the social effects of therapy from questions of the patient's physical condition and the likely therapeutic effects of particular treatments.

Because the authority of lay therapy managers is not separate from generalized authority in the domestic and community sphere, all the factors which shape local communities affect healing. Changes in transport, production techniques, marketing, or education profoundly affect the shape of local society, and therefore also healing.

Janzen, in The Quest for Therapy, ignores this central social process, although the gap is one which is clear to us only because of the excellence of his own research. He establishes the historical context of his account by writing about the




history of types of healing, and of government control over medicine. This is merely one part of the framework within which people shape therapy. The other part must be found in the institutions of domestic and community life-not separable in any way from those which manage the peasant economy, ritual, or peasant politics at the most local level. The most satisfying historical context for the evolution of Kongo therapeutics would be the total history of local social organization and ideology. It is not the history of a transformation from "traditional" to "scientific" medicine (Twumasi, 1975), nor is it a history of therapeutic pluralism. It is, instead, the history of the fundamental social institutions which control therapeutic choice.

This view of the history of therapy breaks down all barriers between the study of healing and the general study of social life.

The implications can best be understood through an illustration of relations among therapy, health, and other elements in local social change. The illustrative case is based on my own field research (in collaboration with Dr. Elizabeth Karlin) in northeastern Tanzania.

In the 1880s, in this particular region, each group of adult brothers together with their father (if he was alive) coordinated farming activities in a limited way, and cooperated fully in responding to major reproductive crises. Each fraternal group held a fund of common wealth to pay for food in famine, to ransom members taken captive, and to pay for communal therapeutic rituals. In the two decades after colonial conquest these groups fell rapidly into decline. Most of them divided the funds of common wealth, leaving each male-headed household on its own, with very strong moral obligations to help each other (for example to participate in therapy management), but without jural obligations.

As the cash crop economy grew in the twentieth century, two alternative labor patterns appeared. In one, the proprietor (for example of a coffee farm) used hired labor; in the other, much more common, he used family labor. Wealthy coffee farmers welcomed widow inheritance (practiced within a broad group of agnates) as a way of increasing the size of their dependent labor force. A man's most important way of attracting dependents was by demonstrating his generosity at moments of need. A man who took an active part in therapy management for relatives, or who paid school fees for the children of impoverished relatives, was more likely to inherit widows and to attract dependents in general. All this changed with the introduction of Universal Primary Education in the mid-1970s. Now virtually all children went to school, needed school clothes, and were absent from farm work, and family labor declined rapidly. It was replaced by hired labor. Widow inheritance, along with other measures by which men expanded the number of their dependent children, became unpopular. The proportion of households headed by women increased, with significant consequences for the health of their children.

In our village study of 1979-1980 we paid special attention to children of those widows or divorcees who did not live with their parents or brothers, and who had not been remarried or inherited. These children proved much more vulnerable than other children to severe malnutrition, to nutritionally induced complications of infectious disease, and were more likely to die. In the most extreme of these cases (some investigated in the village setting and some in the hospital pediatrics ward) the child would not be fed adequately when ill; no one would discuss therapy management. People did not make a conscious and callous decision to neglect these children. The children and their mothers were close to




invisible. When they were brought to the attention of their neighbors and distant relatives they were given care.

South African researchers have found that the kinship support network is important for levels of child nutrition. Thomas (1981), for example, found in the Ciskei that 70 percent of children in the malnourished portion of his sample came from households from which the father had deserted. The burden of the argument is not to lay the blame on the households themselves, which are the victims of extreme exploitation, but to understand the specific domestic mechanisms through which the general social conditions have their greatest impact. In Tanzania, Jakobsen (1978) found children of migrant laborers to be at risk of malnutrition. What reseachers have not explored, however, is the relationship between patterns of therapy management as conditioned by change in domestic organization and levels of health.

In addition to the intense problems of women cut off from extended support, my own research pointed to a second set of problems where mutual support networks were overwhelmed by health problems because of an imbalance between the number of sick and healthy people. Scholars of peasant society ever since Chayanov (1966) have understood the need for each household to balance the amount of productive labor of which its members are capable against the household's consumption needs. Local studies in Africa have challenged the definition of the household, but not the importance of the labor-consumption balance. We found that the network of support at times of illness must be larger than the unit which organizes production and consumption. Otherwise the small domestic units are overwhelmed. But when the network of mutually supportive domestic groups is small, and the illnesses numerous, it is possible for the system of care and of production to be overwhelmed, leading to health crises. In much of Africa therapy assistance networks are shrinking, leaving sick people stranded without care. In these cases as with invisible children, patterns of health care and of health are inseparable.

This extended example has been worth setting out because it demonstrates, even if too briefly, the way a single set of transformations shapes the division of labor, the power structure of the local community, the conditions of health for particular individuals, and the pattern of control over therapy. The analysis could obviously be extended further in a number of directions-into the relationship, for example, between cash crop production and nutrition.

Therapy management, like all therapeutic power in the African tradition, is ambivalent. It may, as in the case of Lwezi Louise, reinforce patriarchal assumptions. But it also clearly reassures the patients who can expect the support and care necessary for health, and even for survival.

The interrelated evolution of therapy management and peasant economy is one on which the literature of African studies is close to silent. Some important works have described the way large units for agricultural production and consumption have shrunk with the increasing penetration of capitalist relations of production (Haswell, 1953; 1963; 1975; Watts, 1983: chapter 5; Raynault, 1976). But even these excellent works limit their main attention to production and consumption. We need to expand Chayanov's calculus to take account of all the local community's reproductive functions-therapy management in all its aspects (including nursing and the payment of health care costs), child care, cooking, and so on.




"What changes, as the mode of production and productive relations change, is the experience of living men and women" (E. P. Thompson as quoted in Comaroff, 1982: 63). Because healing plays a basic role in framing and giving meaning to that experience, its content must change along with the changes in its organization. Once again, the literature shows this to us only in the most limited glimpses. Alongside Raynault's and Watts's description of the narrowing of Hausa economic cooperation, we have Murray Last's study (1979) of the decline of an illness which reinforces women's links with their natal homes, and the rise of a new illness treated by the woman alone in her husband's household by self help. On the eastern side of the continent, in Bunyole, Whyte (1982) describes the weakening of kin group organization attendant upon cotton cultivation, and the appearance of individualistic treatments in Nyole therapeutics, although she emphasizes that the process is not a clearly linear one. Collective therapies survive alongside individualistic ones. According to Corin's description of Zebola possession in Zaire (1979), the cult requires much less family participation in Kinshasa than in rural areas. Janzen (1983) suggests that cults of affliction may work in an urban setting to provide a therapy management group for those who do not live near a full network of relatives.

II. THE SOCIAL CONTEXT OF DISEASE IN AFRICA

I am not presenting the sections of this essay in their logical order. The discussion ought to have begun with disease and disability, and moved on to ways of coping with them. Instead, the discussion of control over therapy came first. The order of presentation reflects a judgment that the central point of the argument on control over therapy needs special emphasis because it is invisible to most health planners and to scholars in fields related to health. The point is that local laypeople are able to reorder the social framework of their daily lives so as to shape their own experience of health and healing. The essay will show that the course of this reordering affects the outcome of struggles to control society's overall direction. This happens within a framework of powerful political and economic forces. But it is important to begin by showing that the weight of the constraints does not reduce human actors to helplessness.

I will consider the constraints shaping health, disease, and population over the past century in two parts-the first on the consequences of colonialism (and especially of conquest), and the second on the effects of particular forms of production.

Colonialism and Population The very broad trends in African population over the past century are clear,

even though our knowledge is not built on careful censuses. Colonial conquest brought a period, lasting for several decades, during which the continent's population declined quite substantially. During the inter-war years that pattern reversed itself. Population began to grow with increasing rapidity.

Historians give three basic interpretations of early colonial population loss, of which the second and third are not mutually exclusive. The first has a long history going back to the time of conquest. Henderson (1965: 123), in the Oxford History of East Africa, states the position dramatically:




East Africa [before conquest] was ravaged by tribal wars and by the depredations of the Ngoni and the Arab slave traders. The lot of many of the native inhabitants was indescribably wretched. They lived in terror of the slave-raiders; they suffered from malaria, smallpox, typhus, sleeping sickness, and other diseases; their cattle were victims of the tsetse fly and of rinderpest; their fields were stripped by locusts. Slave raiding and epidemics increased recourse to witchcraft.

Henderson argues that Africa had always been a continent of horrendous diseases, that the epidemics of the early colonial period were in no way unusual, and that colonial rule was needed to change this situation. The record of East African population history does not support his case.

The second position focuses on changes in population distribution and movement in the colonial period. Urbanization, the building of roads and railways, labor migration, and the movement of armies all increased the possibilities of transmitting communicable diseases. Dawson, for example, in his disease-by-disease account of Kenya in the early colonial period, attributes the greatest burden of mortality to labor migration (1981: 134; see also 1983: 7). Hartwig and Patterson, in their introduction to Disease in African History (1978), combine the argument on mobility with a second that improved communications "disturbed the relative tolerance which many rural Africans had developed for local strains of the parasites causing malaria, dysentery, trypanosomiasis, and perhaps other indigenous diseases" (1978: 12). Good's chapter (1978) provides an especially clear example of this sort of localized tolerance. His approach is a useful variant of a much less defensible position taken by others: that Africa as a whole had been isolated in the pre-colonial period, and that diseases, newly introduced from overseas, swept across the continent in the African version of the Columbian Exchange. Breaching Africa's isolation, in this view, was necessary even though costly in human life (Azevedo, 1978: 188).

The third and final approach to population trends under early colonial rule says that conquest was a political event which deprived Africans of the capacity to control their own environment. The European conquerors forcibly instituted new patterns of settlement, labor, and land use despite their own ignorance of African ecology, and thereby destroyed the basis of survival. This is the essence of the position John Ford takes in his book on trypanosomiases (1971). Helge Kjekshus (1977), in a book on East African ecology which owes much to Ford, makes the same case.

It is unwise to evaluate the overall worth of the three interpretations without first looking at regional differences. The population trends which define the early colonial experience in East Africa did not exist in the west, and the health problems of the equatorial forest were different from those in either East or West Africa. The crucial comparative questions-why regions differed from one another-have not been studied carefully.

East Africa

The literature on East Africa offers a rich selection of interpretations of disease history along with a dearth of population figures. But all recent scholars agree that the population of Kenya, Uganda, and Tanganyika declined sharply during the years between colonial conquest and 1920.




Marc Dawson, in his valuable recent work on the history of disease in Kenya (1983, 1981, 1979, 1978), attributes an especially significant role to smallpox epidemics which raged in the aftermath of the major famines of 1892 and 1899- 1900. According to Dawson's interpretation, famine led to population movement and concentration because the search for food drove people out of their homes. The new population patterns then aided the disease's spread.

Helge Kjekshus, writing about German East Africa (1977), asks whether the famines of the 1890s were natural events, or consequences of colonial conquest. He argues that the region's population had been growing in the nineteenth century only to suffer disastrous decline after conquest. He takes the intensification of agriculture in nineteenth century East Africa as a sign of population growth. This interpretation underemphasizes the destructive impact of international trade in the years before conquest. It is more likely that warfare and insecurity drove people into dense defensive settlements where they were forced to intensify their agriculture. The density of population would then have been a result of population movement, not of population growth. Hartwig (1979, 1976) gave a reasoned rebuttal to Kjekshus's claims of population growth, showing that in the pre-colonial decades famine, cholera, and smallpox took a great toll.

Sleeping sickness brought some of the most substantial population losses in early colonial eastern Africa. This is the subject of one of the great (although difficult) books of African history, John Ford's The Role of the Trypanosomiases in African Ecology. A Study of the Tsetse-Fly Problem (1971). Ford, who had a long career as a colonial scientist, argued that African societies had developed effective ecological controls for trypanosomiasis. They had succeeded in isolating the most dangerous trypanosomiases in what Ford called Grenzwildnisse-wilderness areas at the borders of populated territory, arranged so that people and cattle did not come into frequent contact with tsetse flies. The European conquerors destroyed a whole range of controls, and unleashed a plague on Africa. Ford wrote (1971: 9):

It is a curious comment to make upon the effects of colonial scientists to control the trypanosomiases, that they almost entirely overlooked the very considerable achievements of the indigenous peoples in overcoming the obstacle of trypanosomiasis to tame and exploit the natural ecosystem of tropical Africa by cultural and physiological adjustment both in themselves and their domestic animals.

Later in the book he wrote (Ford, 1971: 143), Like their British neighbours across the Kagera in Ankole the Germans looked upon themselves as saviours of people sunk in centuries of barbaric misery. Few realized that they were the prime cause of the suffering they were trying to alleviate.

These are rare expressions of opinion in a massive work of detailed ecological analysis. Its central point is that the conquest destroyed African control of the ecosystem and thereby let loose diseases which had long been held in check.2

Ford emphasizes the role of two disasters in the breakdown of control-the rinderpest panzootic, which wiped out cattle and wildlife; and smallpox, which, together with famine, reduced the human population. These two sets of events, together with the colonial wars which brought them, upset the ecological controls which had long contained the threat of trypanosomiasis. Once this happened,




tsetse belts never stopped advancing. Colonial boundaries rigidified the Grenzwildnisse, and a host of ecologically ignorant authoritarian measures, many of them intended to control tsetse, led to the expansion of fly belts.

One of the core images of Ford's book is of the Semliki Valley, whose residents were forced to move from the upper hillsides to certain death of sleeping sickness in the valley bottom. It is no wonder that some Africans of the region saw colonial actions as a form of biological warfare (Nayenga, 1979).

Equatorial Africa

Large regions of Equatorial Africa, under French and Belgian control, experienced population decline more extreme than East Africa's (Coquery- Vidrovitch, 1977: 341, 346). Trypanosomiasis played a role here, too. In addition, the colonial disruptions of local economy and society were among the most sweeping and intense on the continent. The regime of enforced rubber collection, for example, won notoriety even at the time (Harms, 1975; Coquery- Vidrovitch, 1972). Infertility also threatened long term survival in Gabon, Congo, Central African Republic, and parts of Zaire and Cameroun. In some of the region's societies, even in recent times, 20 percent to 40 percent of fifty-year- old women have never had children (Belsey, 1976: 320-22).

Romaniuk (1980, 1967) and Retel-Laurentin (1974b; 1979a; 1979b) agree that sexually transmitted diseases were the major cause of infertility.3 Retel-Laurentin emphasizes the role of syphilis, but gonorrhea is a more likely cause of infertility (Belsey, 1976).

Sexually transmitted diseases appear not to have spread widely before the nineteenth century. Harms describes how syphilis, spreading from the coast in the 1880s, was so deeply feared that local judges could impose the death sentence on women who slept with Europeans (Harms, 1981: 181-83). The problem of infertility could not have been one of long standing, because the population could not have sustained itself over the long term while threatened by late nineteenth century levels of loss (Caldwell, 1985).

No one has written a serious analysis of patterns of sexuality in the transmission of disease. Such an analysis would need to take account of the economy, work patterns, and sexual division of labor for the period both before and after colonial conquest. Dennis Cordell (1983) points to the works of Thomas (1963) and Dupre (1982) as examples of studies working along these lines (see also Lux, 1976).

The areas of Equatorial Africa occupied by concessionary regimes tended to suffer high levels of infertility, for reasons which are not clear. Perhaps military forces played a role in spreading sexually transmitted diseases. Scholars could illuminate the problem by mapping military forces in the early colonial period in relation to the distribution of venereal disease and of infertility. In any event, population decline in that period was not solely a result of infertility, for extremely high infant and child mortality also played a role. Belsey (1976: 326) cites infant mortality figures ranging from 30 percent to 80 percent for nine villages in Zaire, as reported in 1911. This would have been one additional consequence of the upheavals brought by early colonialism.

The concessionary system of early colonial Equatorial Africa was one which would not make a substantial long term capital investment in colonies, but which aimed at short term profits. Concessionaires achieved these by terrorizing or




killing people unless they brought goods for export. The foreign enterprises which have been most disastrous for the survival of Africans on their own continent have been those which are capitalist in ethos, motivation, and organization, but which do not (or cannot) invest sufficient capital and which therefore intensify exploitation in order to survive. This is clear not only for the concessionary regimes, but also for the undercapitalized mines of Southern Rhodesia at the same time (Van Onselen, 1976).

West Africa West Africa saw few signs of the population decline which was so significant a

part of the early colonial period in Equatorial and East Africa. In this region as in other parts of the continent the statistics of the period are unsatisfactory, but scholars seem generally to agree at an impressionistic level that West Africa's population was relatively stable, possibly suffering small losses in some places, enjoying gains in others, but in no way similar to the drastic declines of Uganda or French Equatorial Africa (Caldwell, 1977: 8; Frishman, 1977: 227; Hill, 1977; Patterson, 1977; Patterson, 1981: 87, 96; Wilks, 1975: 90-93; Inikori, 1981: 299; Perrot, 1981; Wrigley, 1979: 128; for sources see Ajaegbu, 1977).

If African surrender of control over the local environment led to disastrous populaton loss in East Africa, why did it not have the same effect in the west? Inikori (1981) is one of the few scholars who tries to answer this question. He assigns an important role to the West African slave trade.4 He relies on Cissoko's description of sixteenth and seventeenth century epidemics at Timbuktu (1968) to argue that the slave trade led to political upheavals which in turn caused demographic disasters. As a result West Africa suffered the equivalent of its shock of conquest-its period of demographic decline-long before the colonial period. This interpretation does not pass the test of comparative analysis across regional lines. Angola suffered intense slave trading from an early date, yet this did not protect it from population loss in the nineteenth century when colonial pressures became intense (Dias, 1981; see also Thornton, 1977 and Miller, 1982).

Perhaps disease adaptation in nineteenth century West Africa was less localized than on the eastern side of the continent. The peoples of West Africa had probably never been as narrowly restricted as East Africans to particular zones. "By 1800," according to Patterson and Hartwig, "West Africans had had many generations to build up defenses to cope with their more complex disease environment" (1978: 8; see also Patterson, 1981: 2-3).

West African populations were in movement over centuries, especially in the savanna and at the desert edge. People living in the driest zones have always moved southward into the wetter savanna during the dry season and in dry years, with many returning northward again with the rains. This was more significant for biological adaptation than the movement of traders because it meant that whole populations interacted with one another and with alien environments. Pilgrimage to Mecca may also have been significant, for it was a form of movement in which West Africans left their homes, were exposed to alien disease environments over long periods, and then returned home. Al-Naqar estimates that 15,000 West Africans accompanied King Mansa Musa to Mecca. Even if the figure is not precise, it is clear that large numbers of people were involved. Other pilgrimages were also substantial (Al-Naqar, 1972: 12; personal communication, Ismail Abdalla).




The picture is less clear in the forest zone of West Africa, except for the effects of interaction with aliens at the coast, and the consequences of trade- induced warfare. The distribution of the sickle cell gene is one piece of evidence showing that West African populations were less localized in their adaptations to their environment than the peoples of East Africa. The gene confers some resistance to falciparum malaria. In parts of East Africa where falciparum malaria is a significant problem a much larger percentage of the population has the gene than in malaria-free areas. The equilibrium of gene and disease is much less precise in West Africa (Livingstone, 1967), showing that populations were probably not tied as narrowly to particular micro-environments as on the eastern side of the continent.

The final disastrous event which marked the end of early colonial population decline struck all the regions of the African continent. This was the influenza pandemic of 1918-1919 which, according to Patterson and Pyle (1983), took between one-and-a-half and two million lives in sub-Saharan Africa. War-time movement of soldiers and laborers helped to spread the disease, which moved rapidly among newly constructed rail lines.

At some point in the 1920s population began to grow. Since then, the rate of growth has increased rapidly. Gregory and Piche (1982), merging U.N. and U.S. statistics, estimate the continent's population as 164 million in 1930, 219 million in 1950, 352 million in 1970, and 458 million in 1979.

The high rates of population growth do not, however, indicate that people are healthy. The crude death rates of the 1970s varied between fifteen and forty per thousand. These were among the highest in the world, as was the infant mortality rate of about 200 per thousand (Vallin, 1976; Cantrelle, 1975; Gregory and Piche, 1982). Another sign of poor health is the prominence of infectious disease among causes of death. This was characteristic of West European mortality patterns before life expectancy began to increase, at which point degenerative diseases like cancer and cardiovascular disease became more important. In Africa today the great killers include malaria, diarrheal diseases, measles, tetanus, and respiratory diseases. It is possible that maternal mortality is also a significant cause of death (Ware, 1978; Cantrelle, 1975; Gregory and Piche, 1982).

The great growth in African population must be understood against a background of poor health, and cannot be taken (without more specific evidence) as a sign of generally improved conditions. The precise determinants of population growth are largely unknown. We do not yet have the detailed regional populaton histories which must precede works of broader synthesis.

Caldwell writes that a decline in mortality caused the population explosion. His argument grows inexorably out of his assumption (1975b; 1977; 1985; see also Pool, 1977) that precapitalist modes of production in Africa require high fertility, which is then unlikely to rise further as conditions change. The fertility levels recorded in the 1950s were, in his view, typical through the long span of African history. For Caldwell it was high fertility which sustained pre-colonial population despite the very high mortality, caused by an unhealthy environment made worse by the dangers of warfare.

Caldwell's assumptions about pre-colonial population are weak. There are two problems. First, warfare was not a major direct cause of mortality (Kjekshus, 1977; Wrigley, 1979). Political upheavals may have had an impact on population, but it was through disruption of food supplies and the impact on disease transmission. Second, pre-colonial populations probably did not share




mid-twentieth century fertility patterns (Swindell, 1981; Manning, 1981; Ganon, 1975: 697). The Caldwells themselves (Caldwell and Caldwell, 1977) showed the importance of the duration of post-partum sexual abstinence, which was long during the pre-colonial period but became progressively shorter as the twentieth century unfolded.

We do not have comprehensive data on fertility and mortality in the twentieth century. Public registration of births and deaths is still rare on the continent, leaving governments more dependent on censuses. Some colonies collected census information throughout the century, but in others the national governments made the first censuses only in the 1970s. At that late date a large proportion of censuses collected no information on fertility and mortality (Tabutin, 1984). Even when it is collected, the information is often inaccurate. In most cases the interviewer collects mortality data retrospectively, asking about all deaths within the preceding twelve months. This procedure almost invariably underestimates the number of deaths. In the Khombol survey zone in Senegal, for example, a retrospective survey in 1963-64 counted 126 deaths among one to four year-olds, whereas continuous observation over the following three years showed the same zone averaging 233 deaths each year in that age group (Cantrelle, 1975).

It is likely, nevertheless, that mortality has in fact been declining over the past sixty years, even though the causes of the decline are largely unknown. One possibility is that improvements in transportation reduced famine mortality by making it easier for people to move away from famine zones (Caldwell, 1975c). Yet people do continue to die during major famines (Watts, 1983). The question is whether the number of deaths has declined. The most common base-line for answering this is the early colonial period, a time of unusually high mortality.

A second possibility is that medical interventions have saved lives. Most people on the continent in this century have lived where there is no doctor. Where biomedical care exists it is usually a brief encounter between patient and doctor-rarely a matter of thorough diagnosis, treatment, and follow-up. Nevertheless, a few innovations seem important, including sulpha drugs, antibiotics, and anti-malarials. Immunizations also made a contribution. In one of the rare local case studies, Patterson (1981) estimates that colonial Gold Coasts's mortality declined because of chemotherapy for malaria, vector control and therapy for trypanosomiasis, sulpha drugs and then antibiotics for pneumonia, vaccination for smallpox, and treatment of traumatic injuries.

Geographical variations in mortality give some clues on the long-term mortality decline. Rural mortality is consistently higher than urban. Some of the most precise estimates are for Senegal, where Cantrelle placed infant mortality in Dakar at fifty-seven per 1,000, while it reached 247 per 1,000 in rural Thienaba. Sanders (1982) estimates similarly drastic urban-rural differentials for Zimbabwe. In Kenya, life expectancy for women in Nairobi in 1969 was 63.8, compared to 47.5 in Rift Valley Province, and 50.0 in Western Province (Monsted and Walji, 1978: 69).

These differentials show that mobility and urban crowding, which are sometimes taken as the major causes of disease and mortality, are less important than other factors. But which ones? Clean water supply is a possibility. Patterson (1979) shows that Accra's piped water supply expanded rapidly in the years after World War I, at a time when the city made few advances in the disposal of human waste. Access to clean water makes diseases of the gut less likely. These




are major killers and major causes of malnutrition. Political and economic power, which is concentrated in Africa's cities, shapes

the distribution of water supplies. Kenya in the late 1960s, for example, spent two dollars per capita on urban water supplies, and two-and-a-half cents per capita on rural water in a country where the vast majority of the population was rural (White et al., 1972: 13-14).

It is possible that nutrition, too, is better in the city, although urban food supplies have gotten worse recently with the intensification of economic crisis in many countries. Possible reasons for the differential in nutrition, aside from water supplies, include a softening of the effects of seasonal hunger for city- dwellers who purchase food in all seasons, and a shortening of women's work day making it possible for them to cook more frequently. One report (Benyoussef et al., 1974) compared rural and urban samples in Senegal with respect to a number of health indicators. They found that city women were heavier than their rural counterparts, and city dwellers had higher hematocrits (indicating less anemia). The only indicator on which the urban sample was worse off, because higher, was cholesterol.

City people also have easier access to medical care, which might make some difference because of immunizations, anti-malarials, antibiotics, and some surgical interventions. Orubuloye and Caldwell (1975), in an attempt to demonstrate the effects of medical care, studied two towns in Ekiti Division of Nigeria's Western State, one with biomedical facilities and one reliant only on popular healers for care. They found that children were much more likely to survive in the town with a hospital. But they ignored several possible explanatory factors. People in the hospital town drew water from shallow wells, as opposed to water in the other town from streams in which people defecated. Most important of all, men from the hospital town were more likely to live at home with their families, whereas many of the other town's men migrated to find work. In studies around the continent absent fathers and child malnutrition often go together. The authors lump together all medical interventions as "care" without distinguishing the effects of immunization and of therapy.

The assumption that declining mortality is the only source of African population growth in this century needs to be questioned. As we have already seen, it is unwise to presume that nineteenth century fertility patterns were identical to current ones. In a large part of the continent post-partum abstinence, which sometimes continued for long periods, was the major form of pre-colonial fertility control. It continues to play a significant role in some places. Caldwell and Caldwell (1977: 212) found that even in the 1970s the average woman in their study population "experienced sexual relations for less than half of her fecund married years." The general picture, wherever post-partum abstinence was important, is of declining duration and gradual disappearance in this century (Schoenmaeckers et al., 1981). The Caldwells claim that even in the days when customary abstinence lasted for long periods, fertility was not higher than it is now. In those days of higher mortality, they say, mothers often made an early end to abstinence after a child's death. The timing of declines in mortality and in abstinence are largely unknown, as are their effects on fertility. We cannot rule out a rise in fertility which contributed to the growth of twentieth century population.

Post-partum abstinence raises a question which is at the core of local level social change in the twentieth century- why did women's reproductive lives




change in this consistent way over large areas? The literature offers few considered hypotheses. Abstinence has something to do with control over women's bodies (Schoenmaeckers et al., 1981)-but what? Could its decline mean that patriarchal social groups lost some degree of control over women, who then chose to have more children? If this is so, why did they choose more children? The answer is not clear. Another possibility is that when large care- giving groups declined, as discussed above, small domestic groups became more insecure. For example, communal care for the aged might well have been better in the nineteenth century than it is today. These days old people need many children for support; then they did not. Perhaps in that case people a hundred years ago did not want maximum fertility. Alongside care-giving needs are needs for labor. The colonial economy, whether in places which supplied male migrant workers or where export crops were important, made great labor demands on each domestic group (Meillassoux, 1975). The only way to cope was to increase fertility (Turshen, 1984) by shortening abstinence periods. In this case maximum fertility is not, as Caldwell would have it, a pre-capitalist survival, but instead the creation of capitalism in its colonial form.

The problem with the explanation based on labor needs is that it would sound alien to those who actually abstain (Schoenmaeckers et al., 1981; Caldwell and Caldwell, 1981). The most common justification of the custom is that it is essential for the health of the infant, or that it is a sign of the mother's morality. The decline in abstinence can only be understood within a general understanding of the way people have re-shaped domestic and community organization in the twentieth century. The over-all process met colonial needs for labor, but was mediated by new forms of ideology and organization at the local level.

The larger pattern of rapid population growth alongside poor health is, nevertheless, in complete harmony with an enclave economy. The concentration of health investments in the cities, the need for maximum fertility, and the paradox that hungry farmers feed city people all need to be understood in this light. The concept of "social costs of production" will illuminate the central relationships.

The Social Costs of Production

Studying the social costs of production makes it possible to identify political and economic decisions which have an impact on the distribution of sickness and death-decisions, for example, on whether to invest in sanitation, education, health care, and family support. The social costs of production are not only the ones normally counted as factors of production, but take in a wide range of costs which in some societies and at some times are counted as production costs, and at other times are borne by the state, or workers' families, or the entire population. Social costs of production include the cost of making working conditions healthy, the cost of feeding workers and their families, of maintaining retired workers, and of either controlling or suffering the environmental effects of the production process. Questions about the distribution of social costs of production are relevant not only for understanding the fate of those employed for wages, but also for peasant producers who may benefit from extension services, public education and health services, or who may be compelled to produce cash crops at price levels which lead to immiserization, and for people who happen by chance to live near fields sprayed by harmful pesticides.




"Social cost of production" is closely related to the terms "social good" and "social cost" as used by liberal economists, but with crucial differences. A "social good" is one which cannot be divided and sold separately to individuals (Samuelson, 1970: 149-51); it is therefore provided by public policy, and treated by any one individual or firm as an externally caused shift in the utility curve (Samuelson, 1970: 454). "Social cost measures the value of the best alternative uses of resources that are available to the whole society, as evaluated by society" (Lipsey and Steiner, 1969: 219).

These liberal definitions of social cost and social good do not take account of any levels of experience or solidarity other than the whole of society on the one hand, and the individual or the firm on the other. The definitions furthermore assume that decisions taken by the whole of society benefit the whole of society. They therefore direct attention away from issues of class control. Questions about which groups within society pay the social costs and which enjoy the benefits are irrelevant from the point of view of liberal economic theory, which is concerned with external economies (benefits) and diseconomies (costs) only for the individual or firm. Yet from the point of view of health policy it is very relevant to know that urban dwellers derive benefit from irrigation without sanitation, while the rural population pays costs in ill health. "Social cost of production" differs from the liberal definitions of social cost in two ways: first by specifying that the relevant linkages are between production and social costs; second by specifying costs and benefits in terms of relevant social sub-groupings, whether based on class, gender, or the differentiation of rich and poor geographical regions.

Many recent works define costs of family support or worker health as costs of the reproduction of labor. I intend to retain both terms for use in different contexts, but to give greater emphasis to "social costs of production" for two reasons. The first is that reproduction is something we intuitively think of as being private, the concern of a wife, a husband, and their larger circle of kin. Social costs of production, on the other hand, must be paid at some level of society, but the term itself does not have strong connotations tying it to a particular level (whether domestic, the level of a business enterprise, a local community, or the state). Second, we easily forget that the assignment of costs to either the productive or the reproductive category is not objective or universal, but varies according to historical context. In the U.S. a cook who fries eggs in a restaurant is seen as producing, while one who fries eggs in a domestic kitchen is reproducing. In the U.S. the cost of protecting a worker from very high levels of exposure to lead is a productive cost; in South Africa the cost is assigned to the sphere of reproduction, for the worker is exposed and then, at a certain point, sent back to a rural home for relatives to support.

In any given historical setting the particular distribution of social costs of production among workers, corporate capital, the state, and consumers has a certain stability. Some elements in the distribution seem to those in a society to be almost a part of the natural order, and not to be questioned. For example, in colonial Africa women produced the food which sustained the families of migrant laborers. In the U.S. today, polluters do not usually pay the health costs of industrial pollution. These are paid either by health insurance subsidized by all employers, or by the afflicted individual, or by the state. At each historical moment political conflicts bring into question some previous decisions on the distribution of social costs of production, as in the Reagan administration's




attempt to reduce social security and health care for the elderly and poor in the U.S. Piven and Cloward (1971: 3) have shown that in the U.S., "when mass unemployment leads to periods of turmoil, relief programs are ordinarily initiated or expanded to absorb and control enough of the unemployed to restore order; then, as turbulence subsides, the relief system contracts, expelling those who are needed to populate the labor market." As this example makes clear, and as we can see from the radical difference between the distribution of social costs within the metropolitan countries and within their colonies in Africa, each concrete form of the organization of production and of political power has its own characteristic patterns of payment for the social costs of production. The patterns of morbidity and mortality which grow out of particular forms of production should therefore not be seen as inevitable consequences of the technology of production, because the effects of technology are mediated by decisions on which social costs of production should be paid, and by whom.

A colleague, on reading an early draft of this essay, informed me that K. W. Kapp had made the essential core of the argument more than thirty years earlier in a book about the U.S. economy entitled The Social Costs of Private Enterprise (1950). This is so. The obscurity into which the book and the concept have fallen show that social costs will never be redistributed by the invisible hand of the market or because of the rationality of this argument. The issues can only be addressed in practical ways by political action.

The four case studies which follow-on schistosomiasis, malaria, women's work and malnutrition, and occupational health in southern Africa-are selective illustrations of social costs and their distribution. Schistosomiasis radically changes its frequency when economic activities alter the landscape. Scholars often describe the disease as an inevitable consequence of development. The present discussion is meant to bring that inevitability, the "naturalness" of the disease, into question. The complex set of relations in schistosomiasis among government planning, the rural economy, hosts and parasites could as easily been illustrated with an account of trypanosomiasis, which is not discussed. The second case study, on malaria, reveals some of the same relationships among production, the landscape, and the distribution of disease. Malaria is also worth studying as one of the most important and rapidly growing disease problems in Africa today. The case study on women's work shows that malnutrition, which often interacts disastrously with infection, emerges from the way policies and economic pressures create the structures within which women shape their daily work. The final case study shows that South Africa's decision to use a migrant labor force has had significant consequences for health within the country and the southern African region. The costs of caring for ailments, the costs of old age support, child rearing, and food for workers' families, are borne to the greatest extent possible by the workers and their families.

The case studies define the sphere of production broadly. They are meant as a corrective to the ideology which finds its clearest expression in the migrant economy, in which conditions only count as work-related if they occur during the brief period the worker is receiving wages. Employers describe industrial accidents as work-related, but rarely accept responsibility for cancers which appear fifteen years later, or among non-workers living near the factory. In addition, the division between the work place and the rest of society has the effect of defining most women's work, even women's farming in Africa, as though it is neither labor nor production. If we accept this, then women ought to be




studied almost entirely in the reproductive sphere. The present definition of social costs and the case studies try to show, instead, that a broad interpretation of production comes closer to revealing the health consequences of production.

The Social Costs of Production: Schistosomiasis

Hughes and Hunter (1970) describe schistosomiasis as one of the "diseases of development." It is a parasitic disease for which snails are the intermediate hosts, and is therefore spread easily at large scale irrigation works where people live near snail-infested water. This sense of the inevitability of schistosomiasis as a cost of irrigation is misleading, however. Schistosomiasis is inevitable only if policy-makers decide not to pay for sanitation and water supplies. This disease provides us with a case study of how a clear decision not to pay one of the social costs of production created a substantial health problem.

The chain of events by which irrigation creates disease is best understood in terms of "simplification of the ecosystem." Dunn (1968) explored the implications of this process in his work on the health of hunter-gatherers. Recent articles on large-scale irrigation continue Dunn's line of thought (Kloos, DeSole and Lemma, 1981; Hill, Chandler, and Highton, 1977).

According to Dunn (1968: 225), "Parasitic and infectious disease rates of prevalence and incidence are related to ecosystem diversity and complexity." The more species of plants and animals per unit area, and the fewer the individuals per unit area, the greater the diversity. Tropical forests, which have many species of trees per hectare (60 species in one particular study cited by Dunn), with many of the species represented by single individuals, are very diverse ecosystems. Thorn woodlands are simple ecosystems. Large irrigation projects and large-scale commercial agriculture usually contribute to a process of simplification.

In Dunn's interpretation, simple ecosystems (as compared with complex ones) have fewer species of parasitic and infectious organisms. Sexually reproducing organisms which happen to find the particular environment favorable achieve very high densities-because of the uniformity of the environment-leading to intense infections. In environments where worms reproduce easily, burdens of worms are heavier. With simplification, there are also fewer species of potential intermediate hosts for parasitic and infectious organisms. But the process of indirect disease transmission (whether of sexual or asexual infectious agents) tends to be highly efficient.

In the case of schistosomiasis, for example, the snails are intermediate hosts, the irrigation projects are simplified ecosystems which provide an excellent uniform environment for snails, and the increases in infection are enormous. There is frequent contact between people and water in which the parasites live in their free-swimming stage, after they emerge from the snails. Infection rates of Schistosoma mansoni among schoolchildren in some irrigated areas reach 80 percent.

Irrigation does not, however, lead inevitably to schistosomiasis (Bruijning, 1980). Proper measures can break the cycle of transmission and lower the incidence of infection. People living near the irrigation ditches need adequate latrines so as not to evacuate schistosome eggs into the water, and they need piped domestic water. Those who plan the irrigation works can introduce measures to control the snails which are hosts to S. mansoni. The problem is




that African governments and aid donors (who want the largest and most visible projects for the least money) usually underestimate costs of resettling people. Final costs are often two to three times the amount budgeted (Scudder, 1973), even without latrines and piped water. It is often the poorest of the people in the irrigated areas who pay the cost in debilitation. Kloos, DeSole, and Lemma (1981: 463) studied 363 people living near irrigation schemes in the Awash Valley in Ethiopia. They found that none of the farmers who benefited from irrigated plots in settlement schemes were infected with S. mansoni, but 57 percent of crowded subsistence farmers were infected, as were an identical percentage of migrant laborers and their families, who form 90 percent of the population of the scheme. According to the authors of the study, "Availability of land for defecation is most limited in labor camps in irrigation farms, largely due to high population density and the use of all available land for irrigation agriculture." In this particular case it looks as if those who pay the social costs of production are not the people who gain the benefits.

Policy planners who make decisions on irrigation works rely on a limited range of social research in deciding whether to pay the social costs of production. One important study is a book by Weisbrod, Andreano, et al. (1973), which directly approaches the question of the costs and benefits of schistosomiasis control, and which has been quoted widely (see for example Stockard, 1978). The authors studied the effects of schistosomiasis on the economic production of banana plantation workers in the Caribbean. The authors show that schistosomiasis reduced the amount the male banana plantation workers earned each day by about 30 percent, but that infected workers "respond to a decrease in their daily productive capacity by working more days per week .... In short, the cost of schistosomiasis infection for males is a reduction, not in market production and earnings, but in leisure" (emphasis as in the original; Weisbrod, Andreano, et al., 1973: 75). In other words, the authors consider only the work done on multinational-owned banana plantations as economic, and all other work, whether on subsistence farms, in crafts, in health care, in food preparation, or in child care, as leisure. In my own view this method has the effect of systematically removing many of the social costs of production from the calculation of cost and benefit. The cost of caring for those who are debilitated in old age after a lifetime of schistosomiasis, the cost of reduced smallholder food production, the cost of reduced craft production-all of these the researchers account for as losses of leisure. In addition, they calculate the costs for one disease at a time, even though the conditions which lead to schistosomiasis also increase the prevalence and incidence of other infectious diseases." The most important effect of this distorted measurement of economic cost is to justify the policy of providing irrigation without significant sanitation or disease control.6

Simplification of the ecosystem combined with narrow limitations on paying social costs of production is characteristic not only of irrigation schemes but also of plantation agriculture. In the enclave pattern government and business invest only in what is directly productive (with production defined in the narrowest way), and shift the greatest proportion of social costs of production to those who are poorest. This pattern was at the heart of colonial policies, and remains a major strand of current development policies.




The Social Costs of Production: Malaria The spread of malaria in the twentieth century, like the spread of

schistosomiasis, is related to the enclave pattern of development and to the simplification of ecosystems. Many of the ecological changes resulting from the expansion of commercial agriculture have led to a change in the distribution of mosquito species, and especially to an enormous expansion of Anopheles gambiae, the major African malaria vector. In undisturbed forest, A. gambiae is one of the less common mosquito species. All over the continent the twentieth century has seen a process of clearing forest and brush, leading to an increase in the percentage of A. gambiae among mosquito species.7 In the Kano plains rice scheme in Kenya, A. gambiae went from 1 percent to 65 percent of the mosquito population (Desowitz, 1976; Hill, Chandler, and Highton, 1977). Quarrying, mining, brickworks, road construction, and urban building all lead to the creation of small accumulations of water, and therefore to the expansion of A. gambiae (Dutta and Dutt, 1978).

It would be impossible to argue that all forest clearing, farming, and irrigation should be abandond. As in the case of schistosomiasis, however, the economics of colonialism and the enclave pattern of development contributed to the seriousness of the problem. Rural poverty plays a role because quality of housing for both people and domestic animals affects prevalence rates. Measures for basic environmental control, such as the provision of drains, have been minimal. National governments (and later world health authorities) took vigorous measures before World War II to control malaria in Europe, and after the War used DDT to control the disease in India and Latin America. But in Africa none of the authorities-colonial, national, or international-made full-scale efforts at malaria control. The job seemed hopeless in a region without widely distributed health services, and where rates of prevalence were high (Bruce-Chwatt, 1974; Bruce-Chwatt and De Zulueta, 1980; Brown et al., 1976). Even in holoendemic regions, however, achievements are possible, as shown by the success of a mining company in Liberia in drastically reducing the prevalence of malaria. Hedman et al. (1979) estimate the cost of the control program at $4 per person (probably an underestimate), but the crucial fact is that when authorities decide to pay the costs, gains can be made. The one hopeful possibility is that malaria, because of its rapid expansion, will serve the same purpose in twentieth century Africa that cholera served in nineteenth century Europe-as a scourge which crosses class lines and therefore gives those with power an additional incentive to work vigorously for its control.

The failure of control efforts in India and Latin America and the consequent resurgence of malaria were direct results of measures which large scale commercial agriculture, especially plantations, took to deal with pest problems emerging from simplification of the ecosystem.8 The use of a single plant species as a crop over an extensive area leads to an increase in the density of the pests able to feed on that crop (Gillham, 1972). The characteristic response of capitalist agriculture on an industrial model is to find the silver bullet-the chemical capable of wiping out the pest wherever it is found. Most of the insecticide produced in the world is applied directly to the land for these sorts of agricultural purposes; much less is used for disease control (Busvine, 1978). In the years after World War II, DDT was the highly-capitalized farmer's panacea. Wide agricultural use led to pest resistance. Farmers sprayed entire fields, and then DDT entered agricultural run-off flowing into lakes or streams. Under these




conditions the weakened solution kills some but not all of the larvae present. Those which survive are more likely to become resistant mosquitoes. If DDT had been used only for malaria control-for spraying in houses or limited outdoor areas-the likelihood of insecticide resistance would have been much smaller. It would be wise to reserve some insecticides for mosquito control, but pesticide manufacturers and large farmers have been unwilling to do this (Agarwal, 1978, 1979). A second element in a rational approach would be integrated pest control, in which insecticides are "carefully chosen and applied to maximize their action on the target organisms and to minimize their impact on non-target species and on human and animal health" (Brader, 1979: 226). Biological controls would then be part of an integrated program (Ruesink, 1978; WHO Expert Committee on Insecticides, 1975; WHO Expert Committee on Insecticides, 1976; PAHO Advisory Committee on Medical Research, 1972; Davidson and Zahar, 1973; "Biological Control of Insect Vectors," 1978).

The Social Costs of Production: Malnutrition and Rural Women's Work

The majority of policy makers in twentieth century Africa (whether colonial or national) count women's non-wage work as domestic activity, of little interest for economic development. This is similar to the way most economists in the U.S. treat housework. In Africa, however, the sphere of uneconomic domestic activity is defined even more broadly to include women's farming for household consumption. Barbara Rogers (1980: 142-45) and others show that colonial regimes, international aid agencies, and national governments, have all tended to count commercial crops as masculine, and have directed credit and extension or marketing services towards men, while treating women's farming as uneconomic subsistence. According to Wolpe (1972), Murray (1981), Meillassoux (1975), and many others, when women's production is not counted as economic, employers can reduce wage levels, since working men do not need to support their families.

The general point about women's hidden subsidy of African labor systems is well known and widely discussed. Few scholars, however, have systematically explored how the organization of women's work under these conditions shapes basic patterns of malnutrition.9 The distribution of women's work time in peasant agriculture has changed drastically over the past century, with significant consequences for the distribution of infant and child malnutrition. The picture is not altogether clear, for scholars have only recently begun to study historical changes in women's work. Even now research tends to be restricted to farming work-only one part of a total picture which ought to include cooking, carrying firewood and water, child care, crafts, care of the sick, and a range of other activities.

My hypothesis, based on a broad but not overwhelming range of data, is that in much of Sub-Saharan Africa crop regimes have changed over the past century so as to intensify seasonal food shortages, and also to intensify seasonal variations in the demands on women's work time. This led in turn to higher seasonal peaks of both malnutrition and death among infants and children. If the hypothesis is correct, it means that today's characteristic pattern of malnutrition is a recent phenomenon related to changing demands on women's work time.

The tie between malnutrition and rural women's work is clear. In most African settings where malnutrition is a serious problem, fats and oils are in




short supply and foods are relatively low in energy. If convenience foods are unavailable, the level of child nutrition depends on the number of times a day children eat cooked foods. The way to improve child nutrition seems obvious: children need more meals a day, and if possible meals with higher energy content. At the best of times, however, peasant mothers must farm, care for children, carry firewood and water, cook, and do other chores. The multiple demands on their work time can make it impossible for them to provide adequate food. Rural mothers who are a bit better off might buy firewood or use piped water, but for the vast majority of peasant mothers the constraints of work time limit child nutrition, even if food is available in adequate quantities.

In rural areas, especially those with a single main rainy season, the amount of labor women devote to agriculture varies widely from one season to another. During the months which demand the heaviest labor, the nutrition of women and children suffers. Schofield (1974) surveyed the effects of seasonal labor peaks on child nutrition: women prepare meals less frequently; they leave the cooking pot to simmer, destroying vitamins; they come home to cook after the children have fallen asleep; women do not have time to gather green leafy vegetables; they do less housecleaning and limit fuel and water collection; and mothers devote less time to their children's care. The women themselves, at this time of year, often go hungry. The heaviest work seasons come at times when the previous year's food is nearly exhausted and the new year's food has not yet ripened. Women tend to get fewer calories than they need at the heavy working time, and to make up the deficits after harvest (Bayliss-Smith, 1981).

Nutritional levels and climatic conditions affect the disease picture at the same time. According to McGregor, 78 percent of childhood deaths in the Gambia occurred during the rainy season (1964; as reported in Hull, Williams, and Oldfield, 1983). Diarrhea, which is one of the commonest causes of childhood morbidity and mortality in Africa, appears to affect more children in the hungry rainy season (Rowland et al., 1981; Ndagala, 1981; Onchere and Slooff, 1981). Birthweights are also seasonally depressed. The evidence is strong for a correlation between an infant's birthweight and the probability that infant will survive (Mata, 1978; Rowland et al., 1981; Hull, Williams, and Oldfield, 1983).

Two sets of changes have had a profound effect on the seasonal distribution of women's work time. The first is change in the agricultural division of labor by gender. Men increasingly devote their working time to either wage employment or cash crop production. They therefore pay much less attention than they did a century ago to food crops for home consumption (Bukh, 1979; Linares, 1981; Johnny, Karimu, and Richards, 1981). Households use men's cash income to supplement food stocks in the hungry season. This means, of course, that women-centered households without men become especially vulnerable, as do debtor households. Haswell found in the Gambia that in the 1950s and 1960s men who grew cash crops withdrew from large mutual insurance kin-groups. The old system of hungry season sharing therefore declined, to be replaced by a new one in which the poor borrowed money for hungry season food, became indebted to those better off, and had even greater difficulty facing the next hungry season.o

The second major change is a reduction in the number of food crops each household grows. It is clear for the places on which we have detailed information that people grew many more varieties of food crops a hundred years ago than




they grow today. I found this to be true in northeastern Tanzania, where Germans at the turn of the century reported dozens of varieties of food crops which no longer appear in today's diet. Jan Vansina, in a personal communication, reports the same pattern for much of Zaire. Indeed it is clear on a world-wide scale that a few food crops are becoming ubiquitous while very many localized crops are no longer produced.

The loss of men's labor is only one reason for the change. A second is a basic alteration in peasant farming strategies. African agriculture in the nineteenth century placed heavy emphasis on hedging risks-on growing alternative foods in case drought or irregular rains caused a shortage of the main staple. The alternative crops often had slightly different moisture requirements from the main staple. The diversity of food crops made it possible to spread labor inputs relatively widely over the labor calendar, although the precise labor schedule must be examined in each locality. In addition, colonial (and later national) governments either forced or encouraged peasants to grow a few chosen crops-often the ones most easily transported and sold (Chauveau, Dozon, and Richard, 1981).

The implication of this somewhat speculative history is that patterns of malnutrition are probably quite different today from what they had been a hundred years ago. In those days the rare killing famine was a serious affair, but women would have been able to distribute their work more evenly in non-famine years because each of the many crops had its own work calendar, and because men spent more of their time growing food crops. Patterns of mutual assistance tended to spread the effects of famine relatively evenly over all those who had full rights in a particular locality. Quite probably the seasonal malnutrition which is so important a part of the health picture in today's Africa did not exist then. It is remarkable that after decades of mature scholarship on African history we are ignorant on so fundamental an issue.

The Social Costs of Production: Migrant Labor and Occupational Health in Southern Africa

The differential valuation of women's and men's labor, alongside the differential provision of health services (and in general the differential payment of the social costs of production), is particularly clear in migrant labor systems. These have assumed an extreme form in contemporary South Africa, but were dominant in most parts of Africa during the height of colonial influence, and continue to exist both in Africa and in the guest-worker segment of industrial economies. Scholars have known for decades that systems of male migrant labor affect the health of women and children, although the precise impact has usually been difficult to document." Most recently the authors of a WHO report on Apartheid and Health (1983) pieced together an assessment despite the biases in the data.

The gaps in our knowledge are not accidental; they are necessary consequences of the migrant labor system. Migrant labor is less skilled but cheaper than stable labor-cheaper because costs of health, education, and retirement are not paid. Governments which help to organize the use of male migrant labor on a large scale necessarily place a low value on the health of the migrants' wives and children. The governments therefore do not collect adequate




health statistics about rural areas in which migrants originate (which are sometimes beyond the national borders). Statistics are especially weak on the women and children from among the migrant population. Scholars then are caught in a peculiar trap of data and method, for they need precise data in order to be able to sketch the effects of the labor system. Those effects are very much better documented for parts of the population which do receive government services, and so there is a built-in tendency in all our writings to underestimate the damage done by a system of migrant labor. The bias in the data explains, in part, why Patterson's book on Gold Coast (1981) draws a more benign picture of the nutritional consequences of colonial rule than does Turshen's (1984) on Tanganyika. Turshen, who defines the focus of her reseach to match the contours of the labor system, concentrates on a labor-exporting area; Patterson defines his interests according to the availability of archives. He therefore focuses on the Gold Coast's capital city, for which government records are richest, and does not track working men of the capital back to their families in Northern Ghana or Upper Volta.

In the South African mines, where the migrant system is carried to a level of cynical perfection, a Human Sciences Laboratory works on the assumption that not all Africans raised in the mines' periphery will be strong enough to do strenuous work in the mines. It therefore developed a test chamber for sorting out potential workers (WHO, 1983: 188). Dr. C. H. Wyndham, Senior Epidemiologist for the South African Medical Research Council, commented on the importance of work capability data for health planning:

It is also apparent from these results that a much smaller percentage of rural Bantu males than urban Bantu males is capable of continuous high levels of physical effort. This fact must be borne in mind in the siting of Bantu homelands or border areas, or new industries which require hard physical work. In this context, consideration should be given to the improving of the physical work capacities of rural Bantu males. This could be done by better nutrition, particularly more calories and animal proteins, and by improving their health by eradicating endemic diseases, such as malaria and bilharzia, for it is unlikely that the health and welfare of rural populations will be improved by their own efforts (quoted in WHO, 1983: 188).

The authors of the WHO report on Apartheid and Health stress the importance of the assumptions in this statement that senior members of the South African medical establishment ought to promote medical interventions which maximize profits, and should aim at policies which draw the fittest workers out of the Bantustans. They also note the emphasis on the health and nutrition of "rural Bantu males," excluding African women and children who are not of direct use to South African employers (1983: 188-89).

This example supports the more general sense that the authorities do not collect data of the kind needed for health planning for the families of migrant laborers because they rarely do the planning. The crucial statistics are missing not only in South Africa, but also on the health (and even at times the existence) of migrants in the U.S., witness the law suit alleging that a significant part of New York City's population was not counted in the most recent census. Guest workers in Western Europe (in many cases coming from Africa) hold the dirtiest and most dangerous jobs, but the authorities rarely document their occupational diseases (ILO, 1977: 10, 15).




In South Africa, regulation of the work force defines the way statistics are collected, which then shapes scholarly thought. South Africa collects health data on migrant workers only so long as they are at the work place. These workers return to rural areas when they are no longer vigorous enough for employment, and they therefore die in places where their deaths are not counted-or where the people are counted as non-citizens of South Africa. Official statistics drastically under-report infant and child mortality levels among Africans (Unterhalter, 1982: 1113; Wyndham and Irwig, 1979) because the government does not collect mortality statistics for Africans in the whole of the country, but only for thirty-three magisterial districts of which thirty-one are urban (Simkins, 1979: 95). Rural areas, where mortality is the highest, and where women and children are concentrated, go unreported.

Local studies report very high levels of rural mortality and malnutrition, but it is impossible to construct a systematic account because there are no national statistics.

White (1980) discussed one of the reasons for raised levels of rural malnutrition. Rural families, in his study of two localities, received a much smaller proportion of men's wages than did illegal squatters living near the urban work place, and had a higher percentage of malnourished children. Westcott and Stott (1977: 967) report that 30 percent of Transkei children die of malnutrition before the age of two. According to Leary and Lewis, 50 percent of all children born in Sekhukuniland fail to reach their fifth birthday (both these studies are reported by Savage, 1979). Tobias (1975) reported that adult Venda, Southern Sotho, and Zulu appear to have become shorter over the three or four decades preceding the 1960s. His evidence was (in his own view) unsatisfying, but suggestive. The South African evidence on rural health is rarely satisfying. The government dealt with the problem of malnutrition by deciding, in the 1960s, that kwashiorkor would no longer be a condition of which the authorities must be notified (Unterhalter, 1982: 1112; Mechanic, 1973: 39). Despite all this, most writers on South African health, even critical writers of conscience, have no choice but to rely on the official statistics.

The pattern of data collection, and therefore of the scholarship which relies on it, is even more dramatically skewed when it comes to occupational health. The information base is (in South Africa as elsewhere) an integral part of the system for regulating labor. Most African workers are not covered by the system, and therefore their health problems go unreported. The occupational health system covers less than 30 percent of all workers (Green and Miller, 1979). Occupational health problems are defined in the narrowest possible way, during the period of employment, leading to underreporting of the health problems of migrants who return home. It was the brief of the Erasmus Commission, which investigated occupational health, to deal only with workers while they are employed, and not after they have left employment. Entering employment is difficult for sick men in South Africa-migrants are screened carefully before beginning work-but leaving it is easy. In some occupations migrant workers are simply sent home when they become too sick to work. Even when this is not the case, by the time symptoms appear migrants have often left the work place, and are being cared for by their wives and sisters in rural South Africa, or in Mozambique, Botswana, or Lesotho.

The South African labor system therefore creates systematic biases in the reporting of health problems as they relate to work, in three ways. First,




accidents tend to be reported more accurately than degenerative diseases, because the mines (where accidents are a major problem) keep relatively good records, and because accidents actually happen at the work place-it is difficult to defer their effects until workers return home. The toll of accidents is in fact heavy. In every year from 1970 to 1977, between 700 and 800 workers died in mine accidents, and nearly 28,000 were injured each year (Kooy, 1980; see also WHO, 1983: 190-94).

Second, the effects of carcinogens and of toxic substances are very strongly underreported. This happens in part because of very low standards, and in part because migrants leave the work place before symptoms appear. If lead in the blood were measured by U.S. standards, then 44 percent of the South African workers exposed to lead would have to be withdrawn from exposure. Over 150,000 workers are potentially exposed (Green and Miller, 1980: 148-49). The effects of time lag on migrant workers is dramatic in asbestos mining, and in the asbestos-based industries. The most damaging effects of exposure to asbestos begin to appear between thirteen and thirty years after exposure. How many migrant workers are still at the workplace to be examined after that period of time?

This is closely tied to the third set of biases in reporting. The more stable the work force, the more adequate are statistics on degenerative diseases. White workers in the asbestos industry are employed continuously over long periods, and are therefore still in employment when cancers appear. The crucial study by Irwig and Botha on asbestos related disease studied only whites and so-called "coloureds," because it was impossible to find adequate epidemiological data on the African workers (Flynn, 1982). In 1979 white workers accounted for 51.8 percent of all reported mesotheliomas (related to asbestos exposure), even though they were only 5 percent of the work force in the mines. Africans, who were 92 percent of the work force, reported only 28 percent of the cases (Myers, 1981). The weak position of migrant labor, and weak reporting on it, are not limited to South Africa. It is clear that the South African occupational health picture can only be understood as part of the overall international division of labor. But precisely because South African employers have greater arbitrary control over their racially disenfranchised workers, South Africa can serve as a recipient country for some of the most dangerous industries. In this respect Castleman's debate with Levenstein and Eller is particularly relevant here. Castleman (1979, 1981, 1983) argues that hazardous industries are exported to Third World countries because weak standards of occupational health (for example, the extremely weak South African standards for exposure to asbestos fibers-see Myers, 1981), substantially reduce the costs of industrial production. Levenstein and Eller (1981), who fear an attack on occupational health on the grounds of cost, argue that dangerous industries are exported to the third world not to follow lower occupational health standards (which are relatively inexpensive to enforce), but to follow cheaper labor costs. The argument here shows, I think, that the dichotomy between health standards and labor costs is a false one. Labor is cheap when neither the government nor the employer pays the costs of reproducing labor. This is part of a general system for reducing responsibility for the social costs of production, including the costs of occupational health at the work place, and health services for the families of workers.

Does this mean that there is no hope of improving health in South Africa without a total transformation of the labor system? Perhaps in one sense this is




so. Certainly a broad definition of occupational health (Laurell, 1981), concerned with the families of workers, with those living next to factories, and with the long term effects of carcinogens, is incompatible with an economy heavily dependent on migrant labor. Centralized health planning, by itself, is not likely to create significant change. The possibilities for change become clear only by examining the relation of the healing occupations to power, and to democratically-based movements for change.

III. THE HEALING OCCUPATIONS AND THEIR USES It is time to talk about healers. The essay so far has put them aside to

concentrate on other subjects-how networks of ordinary people take charge of therapy, and how political and economic power shape states of health and disease. But it would be wrong to correct a myth of professional dominance in Africa by creating a new countervailing myth of professional impotence. It is possible now, at this point in the essay-having explored what healers cannot do-to ask about their impact. Does their therapy work? To what extent do they act as an independent source of technical knowledge in society, and to what extent do they act in concert with established political and economic power- holders? Do they take socially effective action to promote health, and if so what are the circumstances of this effectiveness?

The discussion is divided into two parts. The first asks about the healing occupations in Africa, and especially the nature of their knowledge, the degree of their autonomy, and their relationship to political power. Is the position of popular healers in these respects fundamentally different from that of biomedical healers? The second part explores attempts by biomedical practitioners to affect the course of events at social levels where important decisions on health are made-either in the local intimate networks within which people organize therapy, or in the wider political arena where forces struggle to change the distribution of social costs.

Any policy position on the future of medicine in Africa and any careful interpretation of its past must decide whether popular medicine is effective. Does it work? Are there basic differences of knowledge and effectiveness between biomedicine and popular medicine, differences which make one preferable to the other?

Much of the literature about healing in Africa assumes that biomedicine is based on objective knowledge of real phenomena whereas popular medicine is not, and that biomedicine works whereas popular medicine does not, except perhaps as psychotherapy. These assumptions need to be examined carefully.

The judgment on African healing has deep roots. Evans-Pritchard (1937: 63), in his classic study on the Azande wrote that witchcraft "is not an objective reality." It is a system of thought, but the people who hold it never expound it systematically. Witchcraft, in his view, is not autonomous knowledge-coherent and useful irrespective of social context; it is an idiom in which everyday events happen. The anthropologist (and not the popular practitioners) creates a coherent and consistent natural philosophy by discovering and elucidating hidden interconnections.

Evans-Pritchard and later scholars find system at a second level-in the way witchcraft is embedded in the social fabric. These anthropologists use witchcraft to learn the locus of social tensions by exploring the social roles most likely to be




occupied by accuser and accused (Douglas, 1970). I. M. Lewis (1971) treats spirit possession in a similar way as a form of negotiation between the afflicted and their relatives, not as a phenomenon which exists in the real world or as a system of knowledge. Using witchcraft or possession for sociological diagnosis is useful so long as one does not reduce African medicine to the expression and palliation of social tensions. No matter whether the interpretation treats African medical practice as social negotiation or symbolism (in the works of Turner), or as imperfect theory (Horton, 1967), the clear implication of this strand of thought is that African medicine is deeply embedded in social life but not in biology, and that it has no medical efficacy.

By contrast, the vast majority of works on biomedicine in Africa treat its knowledge as both autonomous and efficacious. Medical knowledge according to this body of literature is impartial and expert. It can offer technical solutions to practical problems without choosing one side or another when social forces are in conflict. The strategies for dealing with social forces vary widely among authors-from the sense that scientific research can be formulated "solely on the merit of the problem regardless of practical considerations" (Beck, 1977: 35), to the claim that an integrated social/medical/agricultural science would be able to provide solutions to complex problems if only the field of expertise were defined broadly enough to take in all the variables (Hughes and Hunter, 1970; Coumbaras, 1977b).

I do not propose to address the literature which asks whether western thought is more rational than African, or the one which asks more appropriately whether scientific language or thought within any given society is more rational than ordinary popular language and thought.12 This debate has continued in various forms for a century, with some distillation of the issues along the way, but the result is a stalemate. The various sides are entrenched, under siege, and not giving an inch.'"

Nevertheless, there are clear and useful things one can say about the comparative effectiveness of popular- and bio-medicine. First, within any type of therapeutic practice, healing is a mysterious process (Kleinman, 1980: 33). Clinical knowledge in biomedicine specifies that a particular intervention-for example, administering a drug-leads to improvement in a certain proportion of patients, perhaps 600 out of a thousand. Researchers try to learn which 600, but an irreducible core of ignorance remains about why one patient gets better and another does not. It is clear that the patient's total emotional state, social relations, and culturally conditioned understanding of illness all have something to do with healing, but difficult to know how. The biggest body of research on this concerns placebos-inactive substances given to the patient with the false claim that they are active drugs. They have significant effects in making people better (Frank, 1974). Biomedical research on placebos normally puts to one side, quite characteristically, questions about a rich or satisfying web of personal relationships. In most experiments it must do so. If it did not, then investigators would be unable to say, "All other things being equal, a certain percentage of patients given placebos improve." It is the question of "all other things being equal" which makes healing mysterious.

A body of recent writing, much of it published in the journal Culture, Medicine and Psychiatry, gives careful attention to the symbolic, social, and emotional context of healing. But its founder does not claim that his own methods hold the key to understanding healing. He is interested in raising




questions which will make clinicians more broad-minded, more willing to accept that important elements of their patients' experience and knowledge escape their observation. His method is meant, in his own description of it, to be "quick and dirty," not to solve the problem of how it is that healing works (Kleinman, 1981).

Levi-Strauss writes about the nature of healing in his essay on a Central American treatment for women in difficult child-birth (1963). The healer tells the story of a quest through the supernatural world, but the mythical itinerary through that world becomes fused with the uterus and vagina of the pregnant woman so that narration is in fact manipulation, easing the birth. We do not know whether Levi-Strauss's account is true because he thinks of fact, reality, what actually happens, as irrelevant. He writes (1969: 13) in another work, "it is in the last resort immaterial whether in this book the thought processes of the South American Indians take shape through the medium of my thought, or whether mine take place through the medium of theirs." We have no way of knowing whether stories do move the birth organs. But Levi-Strauss's own narration has a core of validity; symbolic representations of the human body do in fact become entangled with the body itself in ways difficult to understand but important for healing.

Norman Cousins (1981), confined in a North American university hospital, having learned that he was dying of an incurable disease, decided that vitamins (of no demonstrated medical value) were the drug he required and that laughter was central to healing. He therefore consumed great quantities of vitamins, and asked his nurse to set up a motion picture projector so that he could watch funny films. He ate vitamins and laughed and in the end he went home, cured, to resume a normal life. Again the lesson is not a simple one. Healing is mysterious. We must therefore be tolerant of alien cultural strategies for healing unless we have good reason for not doing so.

The symbolic and social content of popular African therapies contributes to the process of healing. It also alleviates suffering in chronic illness, which biomedicine is weak at doing. But at the same time, popular African medicine has strong pragmatic elements and gives weight to natural explanation. Anthropologists interested in the differentness of African thought patterns have underemphasized the pragmatic strand of African therapeutics, creaing a false contrast between pragmatic biomedicine and supernatural popular medicine.

This is one issue on which the ethnoscientific method, because it is systematic even if clumsy, has shifted the terms of the debate in a useful way. Dennis M. Warren began ethnoscientific research in Ghana accepting Field's generalization that for Africans, sickness and health are ultimately of supernatural origin (Field, 1960: 112; Warren, 1979: 36). Warren therefore intended, in his research on Techiman-Bono healing, to look at how disease concepts were "related structurally and functionally to the indigenous system of religion" (1974: 3). Warren's (1974: 431) ethnoscientific collection of thousands of disease terms convinced him that most diseases "are classified as naturally caused . .. and as such have neither structural nor functional relations with Bono religion." Spring (1980), writing about the Luvale of Zambia and Maier (1979) on bone-setting and lancing infections in nineteenth century Asante, also find a strong tradition of pragmatic therapy and natural causation.

The classic texts on African therapeutics reported pragmatic treatments but gave them little emphasis. One of the central points of Witchcraft, Oracles and




Magic (Evans-Pritchard, 1937: 479) was that "Azande attribute sickness, whatever its nature, to witchcraft and sorcery." Gillies, however, shows that Evans-Pritchard says in the same book (488) that "When a Zande suffers from a mild ailment he doctors himself. There are always older men of his kin or vicinity who will tell him a suitable drug to take." Gillies explores similar contradictory tensions in the deservedly influential works of Turner and Horton.14 In each case, careful review of the ethnography shows that natural causation is a major part of the explanatory framework. The illness classification can give weight to natural causes and at the same time be embedded in the particular culture's wider set of assumptions about humanity and society. This is, in fact, the case in biomedicine.

Several of the most satisfying recent works in medical anthropology treat biomedicine as merely one more ethnomedical system, its assumptions to be respected and explored, but not to be treated as privileged or as objective in some special sense. This is the message of Foucault's historical writings (1973). Some historians accept the idea that past science was culturally bound while unfortunately preserving the illusion that today's science is true, objective, not to be questioned, and not contaminated by the prejudices of its historical era.

Several recent ethnographies of biomedical thought and practice in the U.S. have come to similar conclusions: biomedicine is permeated with the assumption that doctors can know the individual body, separate from the mind and from social relations, and can treat the individual through technical interventions. These are some of its central assumptions as an ethnomedical system. Hahn (1982) consults Stedman's Medical Dictionary for core definitions, and finds the following:

Medicine (1) A drug. (2) The art of preventing or curing disease; the science that treats of disease in all its relations. (3) The study and treatment of general diseases or those affecting the internal parts of the body, distinguished from surgery. Disease (1) Morbus; illness; sickness; an interruption, cessation, or disorder of body functions, systems, or organs. (2) A disease entity, characterized usually by at least two of these criteria: a recognized etiologic agent (or agents); an identifiable group of signs and symptoms; consistent anatomical alterations.

Hahn points out that "there is no mention here of non-bodily aspects of persons, or even of persons at all. ... Diseases are thought of as discrete 'entities,' with measurable 'lesions'," treated with medicines which are material artifacts. Hahn explores the significance of these definitions in an extended ethnography of one American internist's medical practice. The internist talks about treating the patient and not treating the isolated laboratory tests. But the patient, in this definition, is the set of relations among all lab tests, the overall syndrome, and not the "patient's world and his or her suffering in it." I would go both less far and further than Hahn. Physicians do get drawn in to a concern for the humanity of their patients, and for the larger context of illnesses, but at that point they usually leave systematic medical knowledge behind. It is difficult (often impossible) within the canons of medical knowledge for the physician to integrate an understanding of the patient's human relations with the numbers in the lab tests. The inherent difficulties are compounded by the division of knowledge among the various medical specialties and sub-specialties.




Corporeal individualism pervades biomedical knowledge. This is cultural, not natural or objective. Young (1980) shows that this is the case for research on stress. Researchers accept stressors as "given" ("all other things being equal," again), with little attention to the social organization of stress. They study stressful events (being fired from a job, marital separation) with no attention to hierarchies in the determination of stress. Why not start at some earlier point in the hierarchy, Young (1980: 142) asks, such as "the socioeconomic determinants of being fired," or at some later point such as "the employment consequences of becoming pregnant?" The discourse on stress depends on a tacit assumption that the individual can be abstracted from his or her social context. This assumption obscures the fact, demonstrated in the section on social costs of production in this essay, that decisions made by the powerful distribute stress in a systematically unequal way.

Michael Taussig (1980) studied the case of a forty-nine year old woman with a fatal degenerative disease of the muscles. She was being treated at a prestigious teaching hospital in the U.S. The woman described her personal circumstances to Taussig (1980:6)-her early marriage against her mother's wishes, and her long years of grinding poverty and hard work. She placed the illness in the context of her own life experience. "You can take a perfect piece of cloth," she said, "and if you rub it on the scrub board long enough, you're going to wear holes in it. It's going to be in shreds. You can take a healthy person and take away the things that they need that are essential, and they become thin and sickly. So I mean ... it all just comes together." She did not tell this to the doctors because "The would laugh at my ignorance." As seen by the doctors her disease was purely organic, not the product of social conditions. When the woman became distressed while in the hospital, the examining psychiatrist suggested that "evidence is strongly suggestive of an organic brain syndrome." Taussig writes, "Having stated that the evidence was strongly suggestive of an organic brain syndrome (i.e. a physical disease of the brain) the psychiatrist in his Recommendations wrote: 'Regarding the patient's organic brain syndrome. ... In other words, what was initially put forward as a suggestion (and what a suggestion!) now becomes a real thing. The denial of authorship could not be more patent" (1980: 9). Once again the patient and her disease were being treated as objects, torn out of their social context.

The woman described how patients helped one another at the hospital. The neighboring patient, who was able to get out of bed and move about, helped with the light switch, the food tray, and anything else which needed moving. But patients could not help one another with physical therapy which, as treatment of illness, was the work of professionals. Patients gave one another friendship. The hospital, which objectified them, "monitored" their emotions.

Kleinman, who is working to create a more humane form of clinical practice, insists that there is no necessary reason biomedicine must treat the individual as removed from the context of social relations.'" New orientations in biopsychosocial primary care in the U.S. and some forms of biomedical practice in Asia avoid decontextualization, and treat the patient within a social network. However, the obstacles to humane practice are to be found not only in the culture of biomedicine, but also in the legal and institutional framework of medical practice. Constraints on the social treatment of illness in the United States include rules of confidentiality which have positive value but which prevent doctors from sharing knowledge about a patient's condition with another




patient, a system of record-keeping which treats individuals as cases, a threat of malpractice suits which enforces rigid standards of case management, a set of rules by which insurance companies pay doctors for treating certain objectively defined clinical conditions, and so on. Changing the practitioners' culture, important as it is, can have only limited effects within this institutional structure.

Biomedicine in Africa is practiced within a very different legal and institutional framework-one in which depersonalization and decontextualization are equally extreme. In most cases a great cultural gap separates practitioners and patients. In earlier generations most practitioners were Europeans. Today, most practitioners are Africans drawn from a stratum of society whose members are in a small minority in terms of income and education. The majority of African patients receive treatment under conditions of extreme medical scarcity. This means that doctors must examine, assess, and prescribe in a very short time, to get on to other pressing cases. We can assume that the doctors do their best, but under these conditions patients are often treated like faceless cases.

"Tradition" and the "Traditional" in African Healing

Both popular medicine and biomedicine are forms of ethnomedicine: they are embedded within a system of social relations, and give concrete form to assumptions about reality drawn from the wider culture, which in turn influences the wider culture. They are products of history. It makes sense therefore to use the term "traditional" for both biomedicine and popular medicine, or alternatively to use it for neither. As they have actually been used, "traditional" and "tradition" are words onto which many writers have projected assumptions about African healing in its historical context. Examining these words and their uses is therefore a tool for uncovering the range of common assumptions before looking more carefully at the actual evolution of the healing occupations.

The large and representative sample of articles in African Therpeutic Systems is a convenient text for reviewing assumptions about "tradition" and the "traditional." The collection, edited by Z. A. Ademuwagun, John. A. A. Ayoade, Ira E. Harrison, and Dennis M. Warren (1979) is large (41 articles), eclectic, and not shaped to reflect a single carefully defined viewpoint. Many of the articles were published earlier in journals and have been cited frequently.

The assumptions are quite diverse, with five major strands. The first is that "traditional healing" is what "traditional Africans" do, and that these are special and peculiar people. Imperato and Traore (1979: 16), for example, say that the Dogon "are the most traditional of all Mali's peoples." Imperato (1979: 202) writes that some African cities like Djenne and Timbuktu have not experienced dynamic growth in the twentieth century and therefore traditional medical beliefs and practices in those cities "have tended to remain static." The authors make the unjustified assumption that where dramatic growth does not take place, change does not occur.

The second set of assumptions characterizes tradition by the presumed quality of its reasoning about cause and effect. For Imperato and Traore (1979: 16) traditional reasoning is "prescientific," and for Conco it is characterized by supernatural causation (1979: 61). Other authors in the same collection disagree with these characterizations. Janzen (1979: 216) and Young (1979: 132) define




traditional medicine as "rational." The assumption that what is traditional is not rational is built right into the

core concepts of modern sociology. Max Weber intended his distinction between rational-legal authority and traditional authority as an ideal-typical one, not as a description of any particular concrete reality. But the distinction has entered scholarly language and thought in a way which sometimes makes the connection between the traditional and the non-rational a matter of definition, and occasionally further assumes that this is not merely a definition but a description of life. The way these assumptions work can be seen from a fascinating sentence in the otherwise excellent article by Edgerton. He (Edgerton, 1979: 93) writes that a particular healer's father had, in the pre-colonial period, been a member of the elite "among whom traditional belief in supernatural causation coexisted with an intense pragmatic belief in natural cause and effect." What Edgerton seems to be saying here is that in the pre-colonial period belief in supernatural causation coexisted with belief in natural cause and effect, but only supernatural causation is to be characterized as traditional! When authors are less thoughtful than Edgerton, the ideal-typical association leads easily to an assumption that traditional medicine is based solely on supernatural causation. This is a position which emphasizes one stream in traditional thought while suppressing our consciousness of a second. The consequences are damaging.

The third definition of "tradition" in African healing is of a residuum-those parts of contemporary therapeutic practice which are left over once other kinds of medicine are accounted for. Different authors disagree over what is to be accounted for before defining the residuum. For Warren on the Bono of Ghana, traditional medicine is everything outside biomedical practice, a straightforward and comprehensible distinction.

The effect of defining the "traditional" as a residuum is to leave the healing traditions of Islam and Christianity in an ambiguous position. Diop (1979: 85) et al. define some kinds of Muslim healing as "traditional." Maclean (1979: 225) includes Apostolic faith healing as part of traditional medicine. The merging of all non-biomedical practice as traditional leads to a loss of clarity. By contrast Ulin (1979: 243) puts "traditional," "clinic and hospital," and Christian "Zionist" into three separate categories. Abdalla (1981b) and Last (1981) have shown that Islamic healing cannot simply be merged with traditional medicine, but exists as an altogether separate alternative, with its own substantial impact on traditional practice.

Fourth, traditional healing can be defined in terms of its twentieth century social context. This point emerges from articles on Yoruba healing by two Nigerian scholars. Ademuwagun (1979: 160) describes traditional healers as living among health consumers and as sharing a common culture with the consumers. This is, of course, true most (but not all) of the time. University- educated African consumers, to take one example, usually live near M.D.'s and share their culture. Asuni (1979: 179) shows how important it is that the traditional healer lacks legal standing, whether to vouch for a patient at work, or to avoid being charged with manslaughter in the care of someone who dies. The legal position of traditional healing is similar in most parts of the continent.

The fifth definition of "traditional healing" is a historical one. For Zeller (1979) traditional Ganda healing is whatever kind of therapy existed in pre- colonial Buganda. Janzen (1979: 211), writing on Kongo healing, offers a more sophisticated version of the same position by outlining a set of core conceptions




which existed in the pre-colonial period, which appear to have been relatively stable through periods of great pre-colonial change, and which have also continued on into the twentieth century. Kikhela et al. (1979: 217) make a distinction between two medical systems in Zaire, "the first imported and the second an extension of autochthonous medical traditions."

African traditional medicine has also been defined by a Regional Expert Committee of the WHO, with much more emphasis on the efficacy of traditional healing than in most contributions to African Therapeutic Systems (WHO, 1976; WHO, 1978). According to the WHO Committee, traditional African medicine is:

the sum total of all the knowledge and practices, whether explicable or not, used in diagnosis, prevention and elimination of physical, mental or social imbalance and relying exclusively on practical experience and observation handed down from generation to generation, whether verbally or in writing.

Traditional medicine might also be considered as a solid amalgamation of dynamic medical know-how and ancestral experience.

Traditional African medicine might also be considered to be the sum total of practices, measures, ingredients and procedures of all kinds, whether material or not, which from time immemorial had enabled the African to guard against disease, to alleviate his sufferings, and to cure himself.

This is far from a colorless committee document, for it takes vigorous debatable positions on a number of difficult issues. It interprets departures from health (and by implication health itself) in broad and inclusive terms as "physical, mental or social imbalance." It recognizes that a range of traditional medical procedures exists, some with an identifiable material basis and some without, some explicable and some not. It places great emphasis on traditional medicine as knowledge inherited from past generations; presumably biomedicine is to be the source of innovation (WHO, 1978: 9). Above all the concern for efficacy comes through: traditional medicine is "dynamic medical know-how," the fruit of "practical experience," which enables the African "to cure himself." This is the language of those who wish to use the services and knowledge of traditional practitioners.

The emphasis on practical efficacy raises difficult questions about the role of popular healing today if one accepts the broad definition of health (as it is accepted in the document quoted). The problem is this: if we accept the proposition that popular medicine is practical, and that it legitimately deals with practical public matters, then popular healers need political influence or power to be effective. The discussion below will show that pre-colonial healers dealt with a whole range of practical matters-including the planning of economic production. But popular healers lost power at the time of colonial conquest and have never regained it, except in rare and unusual circumstances. It is unlikely that authorities with control over government medical policy in most African countries would agree to give popular healers real power to correct social imbalance or to plan irrigation works. The healers are therefore left to treat illness in its private aspects. To the extent that popular healing has been organized on a public scale in recent years, it has (in most cases) either operated underground, away from the view of the authorities, or it has dealt with social problems at a domestic or village level-one which could be construed as private, as opposed to the public realm of politics.




A Gramscian Interpretation of the "Traditional" We need something more than a new definition of tradition. We need a

different style of analysis-one which explores whether healers can use their knowledge to shape the public distribution of health and disease, or alternatively whether they are merely society's sweepers, cleaning up the mess others make. Or still worse, do healers preserve their own privileges at the expense of general health? The question can be stated in more general terms: what has been the relationship between knowledge and power in modern African healing?

Antonio Gramsci (who did not focus his attention on doctors, health, or Africa) was a profound student of the relationship between knowledge and power. Even though class and the social organization of production were at the heart of his analysis, he did not reduce all knowledge to class knowledge, nor did he assume that all intellectuals were bound to take predetermined class positions. He understood that under some circumstances intellectuals could-if they achieved relative autonomy and range of choice-affect the outcome of the core struggles to control the instruments of power.

Gramsci defined intellectuals in an idiosyncratic way. He did not accept the notion that intellectuals are a special set of people who think, because it is clear that all people think. Even the most degraded physical work includes a minimum of creative intellectual activity. All people are intellectuals, but only some have the social function of intellectuals, a function which is directive, organizational, or educative.

Intellectuals play a variable role in the class structure. In some cases a set of intellectuals is indissolubly linked to a class; in others the intellectuals achieve a greater degree of autonomy. The reasons for the variation are historical.

"Every social group, coming into existence on the original terrain of an essential function in the world of economic production, creates together with itself, organically, one or more strata of intellectuals which give it homogeneity and an awareness of its own function not only in the economic but also in the social and political fields" (Gramsci, 1971: 5). The capitalist entrepreneur, for example, "creates alongside himself the industrial technician, the specialist in political economy," and so on.

The system of class and production, having once elaborated a set of intellectuals, does not remain static. It evolves, is transformed, and elaborates new intellectuals. Organic intellectuals, once elaborated in an earlier stage of development, do not necessarily fade from existence with successive transformations of fundamental structures. "Every essential social group which emerges into history out of the preceding economic structure, and as an expression of a development of this structure, has found (at least in all of history up to the present) categories of intellectuals already in existence and which seemed indeed to represent an historical continuity uninterrupted even by the most complicated and radical changes in political and social forms" (Gramsci, 1971: 7). These are the traditional intellectuals, traditional because the content of their knowledge and practice were shaped by a different set of associations from an earlier age. Seen from the point of view of colonial rulers or plantation owners, popular healers are just such a set of traditional intellectuals, inherited from the terrain of production and control in an earlier age.

Traditional intellectuals' knowledge and practice show strong continuities with the earlier era-for example, among academics a concern for humanistic knowledge, or among western European M.D.'s continuities in training practices




or canons of medical ethics. Because categories of intellectuals survive from one era to another, social knowledge exists which has no direct and unmediated utility in either organizing production or defending class interests. The traditional intellectuals never win for themselves a place outside the class structure, but they achieve a degree of autonomy-the capacity to negotiate their own place within the organization of production, political power, and class.

Two alternative strategies present themselves if we are to use Gramsci's analysis in understanding struggles to control African health. One is to take up the formal distinction between organic and traditional intellectuals; this leads to fruitless debate over whether a given occupation at a given moment ought to be categorized one way or another. The second strategy, which promises richer returns, is to study African healing in ways loosely informed by Gramsci's problematic. In particular, it is important to understand the degree to which healers of any kind-biomedical or popular-enjoy autonomy within the structure of power and class. Do healers find their interests linked tightly to those of ruling political elites or dominant classes? To what extent does healing serve the interests of privileged sub-groups within society? How does the place of healers influence the definition of public issues affecting health and illness? Can healers use their authority to help the oppressed win control over their own health?

This analysis merges issues defined by Gramsci in his writings on intellectuals with others drawn from Weberian sociology of the professions-a thoroughly unorthodox strategy. It can succeed if we retain a Weberian concern with autonomy and authority, but reject Talcott Parsons's argument (within this sociological tradition) that autonomous professionals rise above class to serve the whole society. The most important unexamined question is about the place of professionals within the structure of production, power, and class.

Professionalization is a process in which an occupational group establishes control over medical resources, and achieves cultural authority-broad public acceptance of its knowledge and practice.16 Paul Unschuld (1975, 1979), who writes about the professionalization of medicine in Imperial China, defines the range of potential resources broadly. These include exclusive control over medical education, over the right to administer drugs or receive payment for therapy, and over the authority to certify medical conditions (as, for example, when diviners sort out witchcraft accusations, or when physicians give students illness excuses).

The history of professionalization is one of competition. If physicians in Africa today were to win great professional power, they would probably forbid popular healers to administer drugs or herbs of any kind. Victory for one group is defeat for another. The problem is that the contest between competing groups-allopaths and homeopaths, physicians and "traditional healers," or obstetricians and midwives-absorbs attention (as contests tend to do). While scholars are preoccupied with the competitive game, the fundamental issues slide past unobserved. These are issues of the distribution of social costs, of the relationship between production and health, and more generally of the way changes in society generate patterns of health and disease.

Part of the contest, as both Unschuld and Starr show, is aimed at winning general cultural authority. Professional ethics and an ideology of service are important tools in doing this because they demonstrate that the group struggling for monopoly aims at improving the health of the general population and not




just at winning victories of narrow self-interest. There is invariably a contradiction between the unequal distribution of social

costs and the professional ethic of public service. The ethic requires that all people be served; the unequal distribution of social costs requires some people to do without health or health care. The contradiction is rarely visible or widely understood. It is obscured in a range of ways. We have already seen that migrant workers and rural women and children have been defined, at various times, as outside the sphere of concern: not members of the public being served.

The great strength of professional autonomy, of the sort enjoyed by successful categories of traditional intellectuals, is that it leaves open the possibility that professionals will struggle to make allies among categories of people who are deprived of health. In this case doctors retain the ideology of public service but redefine the public being served, making the unseen visible. This happens only rarely. It is, of course, difficult to explain why a particular action does not happen. In some cases it is because of the class origins of those who become professionals, in others because alliances with the powerless would damage the profession's own position. But it is one of the few courses of action with any real promise of improving health. If you are a doctor in South Africa and do not choose to make alliances among the disenfranchised, how do you serve the whole of society ethically?

A single case of medical treatment illuminates the contradiction inherent in narrowly professional practice. Dr. N. McE. Lamont (1973), employed by the Tongaat Group, owners of a large South African sugar plantation in Natal, reported on an acute form of cardiac failure among African field workers. According to Lamont's report the workers all live in compounds. They rise at 4:30 in the morning, eat a slice of bread, drink a cup of coffee, and work in the sun until late in the afternoon, with only sour, watery maize porridge to eat until the evening meal. The plantation supplies large quantities of so-called "kaffir beer" on the weekend. Under these conditions, many workers experienced intractable heart failure, and were therefore sent back to the homeland from which they came. Dr. Lamont introduced a potassium-sparing diuretic, after which "not one case has needed to be repatriated. Furthermore, provided the patients continue on maintenance therapy, it has been found that they can be kept out of failure, doing heavy manual labour, for lengthy periods of follow-up." Dr. Lamont appears to measure his own success as a physician in terms of the capacity of workers to keep working under inhuman conditions. The only possibility for making a significant improvement in Tongaat workers' health is through improved conditions of work. In this particular case it is unacceptable for the physician to define his job in narrowly medical terms. He cannot ignore the distribution of social costs of production if it is so powerful a cause of ill health.

Our own study of professionalization needs to show who in society the profession serves, and who is not served. Then it is possible to address the most fascinating problem: when do healers succeed in influencing the social distribution of ill health? Under what conditions do healers make alliances which empower the unhealthy to improve their own health?




Pre-Colonial Healing A rough sketch of the history of healing over the past century and a half will

show that some categories of popular healers played a basic role in organizing production in the pre-colonial period, and others were linked organically to the holders of political power, or held power themselves. Colonial conquest deprived healers of control over production, and broke most of the links between healing and public authority. This freed some healers from the control exercised by chiefs, kings, or patriarchs, but also left them relatively powerless. This contradictory state of affairs in the colonial period could, perhaps, be defined as one of autonomy without authority.

It is impossible to make a serious review of pre-colonial healing in a few pages. Healing skills were at the core of political, economic, and religious practices in much of pre-colonial Africa. This is one reason for the subject's fascination. But it also means that to understand healing we would need to explore the totality of social and intellectual history over many centuries.17

Merely attempting to define the sphere of healing in pre-colonial society raises difficult questions about medicine as an ethnographic category. If, for example, people in the Tio kingdom during the 1880s interpreted most deaths as being caused by evil (Vansina, 1972), then anyone with the moral authority to assess the causes of evil was a health worker, and conversely every healer of substance was concerned with social order and disorder.

This meant, of course, that the roles of healer, of political authority, and of ritual specialist all overlapped and in many cases fused with one another. In Southeastern Nigeria the Aro oracle was crucial for the maintenance of public order and administration of laws, for the explanation of misfortune, for regulating trade, and as part of the system of ritual. In the kingdom of Bunyoro, important spirit mediums were separate from the king. The king's own fertility and the success of his armies depended in part on his relations with the mediums (Nyakatura, 1973: 82, 85). The network of Nyoro spirit mediums reached, as a healing cult, into local communities. In fact, the evolution of local authority structures revolved in part around the question of whether mbandwa spirits were to be the most important for healing in place of descent group spirits (Berger, 1981).

In the Shambaa kingdom, anyone who controlled medicine for the fertility of the land was seen as exercising sovereignty. Old men recount the story of a famous healer who travelled through the land successfully treating people. The king's advisors brought back reports of the visitor's skill, and also rumors about the power of his rain charms. The advisors called in the mghanga, and promised him an enormous reward if only he would teach the king the secret of his rain medicines. Once the mghanga finished teaching what he knew, the king ordered his execution, because rain charms confer access to sovereignty.

Healing authority and other kinds of authority were institutionalized with a limitless range of gradations. There was no one characteristic African pattern. In some places the king was a healer; in others, the king pitted his political authority against the healing authority of a cult leader. In some localities, the elders of a descent group controlled collective healing rituals; in others, only outsiders could treat illness successfully.

Healing was, in any event, organically bound up with basic political and economic processes. The literature on healing, based largely on twentieth century practices, rarely describes the interconnections clearly. East and Central African




cults of affliction, for example, which have historical roots in the distant past (De Craemer et al., 1976), are described in the classic ethnographic literature as they have functioned in the twentieth century within the constraints of colonial or national sovereignty (V. W. Turner, 1968; Werbner, 1977; Fry, 1976).

A few authors have broken through the restrictions of this narrow interpretation, most notably John Janzen in his book, Lemba, 1650-1930: A Drum of Affliction in Africa and the New World. Janzen situates Lemba (a cult in the area between the Atlantic and what is now Kinshasa) as merely one part of a therapeutic system alongside other drums of affliction, lesser charms, and also secular treatments (1982a; see also Janzen, 1979). Lemba was a form of government and trade organization for its priests, who were wealthy and influential merchants, judges, healers, diviners, and chiefs (1982a: 317). The drum regulated markets and settled disputes. But it was also was a form of therapy for resolving contradictions between the region's egalitarian ethic and the acquisition of wealth and influence through trade. Lemba provided the metaphors and the institutional mechanisms which allowed leaders to grow in influence and wealth without tearing either themselves or local society asunder (1982a: 318).

Schoffeleers (1978) and van Binsbergen (1978, 1981) understand cults as essential parts of the ecological system. Schoffeleers describes a pre-colonial Malawian cult as "a ritually directed eco-system," which regulated practical activities. The cult protected large areas of wilderness from burning. Cult mediums at times compelled people to plant particular crops, and restricted fishing and grazing so as to protect fragile resources. Schoffeleers describes one instance in the 1930s in which the Mbona cult put pressure on part of the population of the Lower Shire Valley to emigrate in order to relieve conditions of overcrowding (1978: 4). Schoffeleers's examples are limited because, remarkably, scholars have not studied cults of affliction as organizations for the maintenance of health (even though this is their avowed purpose). Scholars find it difficult to either accept or reinterpret nineteenth century cultic conceptions of what it is that maintains health. I take Schoffeleers's position to be one which considers the adaptive consequences of an entire pattern of organization rather than of any one isolated feature which may, from an academic point of view, either maintain or damage health.

One weakness of Schoffeleers's idea of a ritually directed ecosystem is that not all health practices in any society were ritually based. Alongside the ritual regulation of well-being were health practices which were either simply customary or instituted on a purely pragmatic basis. No cult was ever the entirety of a system of health.

Ford, writing about an area near the one described by Schoffeleers, describes pragmaic institutions for maintaining health. He argues (1971: 333-35; 1979: 269), following Swynnerton, that the Nguni chief Mzila (on what is now the western border of Mozambique) understood the relationships among tsetse, trypanosomiasis, and vegetation cover. Mzila compelled his followers to "draw near to the king" in concentrated settlements. With normal practices of bush clearing a tsetse free zone would be established. Mzila also ordered the creation of a game reserve outside of which all wild animals were hunted. On the other side of the continent, Ismail Abdalla (1981a, 1981b) describes how the Fulani conquerors of Sokoto, in Northern Nigeria, tried to establish the primacy of prophetic Islamic healing, and at the same time to lay out towns with wide




streets to improve conditions of health. The examples of Mzila and of the Fulani jihad are isolated ones culled from larger works, and not extended analyses of African control over conditions of health. A great many historical works describe preventive health measures briefly in passing, but few explore the subject systematically, despite the wealth of detail in historical sources.

Donna Maier (1979) is another scholar who demonstrates that a strong and coherent argument can be made for pragmatic public control of health in pre- colonial Africa. She writes on Asante, using well-known sources to make an original case for the importance of rational prevention and treatment in the nineteenth century. She describes the carefully maintained latrines of Kumase, down which people poured boiling water each day. (For a comparable case in East Africa see African Water and Sewage, 1982.) Residents burned rubbish behind each house daily. The 'Public Works Department' under the stool of the Akwammofo Akonnwa organized street cleaning and sanitation. The Asante practiced variolation for the control of smallpox (a practice discussed for Africa as a whole by Herbert, 1975). All healers around Kumase were organized under the stool of the Asantehene's doctor, the Nsumankwehene. The British later, in the 1930s, assigned this officer the job of licensing African healers. (See also McCaskie, 1981 on anti-witchcraft movements in the period, which were another form of health control.)

We must be careful not to imagine an idealized picture of merry Africa, where chiefs and healers were in harmony, and where medicine served society as a whole, irrespective of the interests of particular social groups. Healers exercised their authority to defend their own interests, and those of their allies. This paper has already discussed the strongly patriarchal assumptions found in some varieties of African therapeutics. This was certainly true in the nineteenth century. In addition, healers who worked in close association with the holders of power were unlikely to challenge the bases of that power. Nineteenth century control over the conditions of health was part of a larger system of domination, subject to the inequities and contradictions of that system. McLeod (1972) shows that Zande poison oracles were used in ways which divided nonaristocrats against one another, reinforcing the authority of those in power. Retel-Laurentin (1969: 26-28, 38; 1974: 15) argues that slaves were most likely to die of the effects of the poison oracle, commoners next most likely, and members of chiefly household the least. Institutions for the maintenance of health in pre-colonial Africa were not impartial or class-neutral.

Conquest and the Formation of Traditional Healing Because African healing was bound indissolubly with control over production

and with political power, colonial conquest fundamentally changed the basis of healing, by destroying independent African control of politics and of the economy. Colonial policy towards popular healing varied, but in every case the authorities punished healers severely if they showed signs of wanting to exercise sovereign power by deciding on life or death issues, or on questions of public order. At the time of the white conquest of Zimbabwe, for example, the conquerors held consistently to the goal of suppressing Shona spirit mediums. British attacks on the Aro oracle in Nigeria and German persecution of healers in Tanganyika are part of the same picture. Once colonial conquest was secure, healers in most places were denied any hint of public authority, even when (as




under indirect rule) the colonial rulers made alliances with chiefs. In the mines of Northern Rhodesia or the plantations of Tanganyika, European managers often appointed quasi-traditional "tribal headmen" to prevent the emergence of worker unity. But traditionality on plantations never extended to authority over health. A worker's sick excuse would never be accepted if it came from a popular healer.

Once popular healing was removed from power, its system of thought and action was transformed. Popular healers were forced to abandon the expectation that they would control public conditions of health, whether through the killing of witches, through the isolation of individuals in polluted conditions, through large scale communal rites for rain, or through the ecologically sound management of land. At the same time acceptance of the central forms of popular healing weakened. Some of the unquestiond assumptions which underlay the varieties of pre-colonial healing were now challenged by missions, by hospitals, and (in more places than in the pre-colonial period) by Muslim healers.

There are no systematic studies of the actual conquest as it affected health institutions. The simplest way to sketch the problem, given the weakness of the literature, is to outline the effects of conquest on the organization of health for the area in which I have done research, to make the case that loss of African sovereignty transformed healing. The area is in northeastern Tanzania, extending from the Usambara Mountains eastward to a line about fifteen miles inland from the Indian Ocean.

The base line for change is the mid- to late-nineteenth century, when the area was governed by local chiefs who were at times dominated by a central king (Feierman, 1974). Authority for control of health was in the hands of a set of leaders which included chiefs, healers, and local patriarchs. These controlled the conditions of health in several different ways. First, they were responsible for controlling the kinds of deviance which were thought to threaten communal health. Biomedical health authorities would no doubt agree with them on some points and not others. Diviners identified witches, and chiefs eliminated them, either by bringing them under control at chiefly capitals, or by killing them and selling their children into slavery. Patriarchs, on the advice of diviners, drove out or killed polluted individuals whose presence threatened the survival of local kinship groups. These included twins, and also people with smallpox who were sent out of the village into isolation. The authorities also regulated land use to increase chances of survival. The same triad of healing specialists, chiefs, and patriarchs regulated the use of irrigation channels, burial of the dead, placement of villages and the location of sites for human waste. Chiefs maintained treasuries for public crises; patriarchs had control over collective kin funds to be used for the preservation of health and security. Those with authority also organized rites for communal well-being, to prevent famine, epidemics, and damaging wars.

German conquest can be interpreted (and quite probably was at the time) as an assault on health. The king, whose body was a symbol of fertility and public health, was executed. Forced labor interfered with the farming activities-the source of satiety and well-being. Some healers were arrested and beaten. European planters interfered with irrigation ditches and dug up graves. The government controlled the siting of villages. Missionaries cut down bits of ritually dangerous forest, and gave shelter to lepers. At least one chief was killed for his role in administering communal medicines. The communal funds for




survival and reproduction dissolved. This was true not only of the funds administered by chefs, but also of those in the hands of patriarchs. The popular medicine which emerged from this experience was less oriented towards public health than it had been, and more a collection of treatments for individual conditions or, at most, for afflictions which affected localized groups.

The assault on the institutions of health is a subject best described in novels (Achebe, 1958; Armah, 1979). Individual aspects of the assault are described well by historians, as in Isichei's account of the attacks on the Aro oracle (1973), and by anthropologists (see Willis, 1968 on the missionary destruction of Fipa shrines, and through them of divination).

The literature about conquest has included extensive discussion of spirit mediums and medicine in resistance (Gwassa, 1972; Iliffe, 1979; Ranger, 1967; Cobbing, 1977; Isaacman, 1976). This literature is rich in its analysis of the African struggle to retain autonomy and control. Colonialism as a threat to survival has not been strongly emphasized in this literature, nor has the role of mediums in fighting for life and health. The issue often was, in the words of a leader of the women's war at Aba in Nigeria, that "the land is changed, we are all dying" (Nigeria, 1930).

A number of works examine relations between colonial powers and healers in the period after conquest. Beck (1979) argues that the British preferred not to intervene in the affairs of healers, although she is concerned with the period after the first sharp blows of conquest had been delivered and had receded into the past. Schoffeleers (1978: 40-41) discusses the role of colonial bureaucracy in undermining cult leadership. Zeller (1971: 349, 361, 366, 369) gives several examples of the way the British exercised sovereignty over healing in Buganda. Janzen (1982a: chapter 3) gives a fascinating account of how colonial rule undercut the economic base of the Lemba cult's priests, and then completed the process of undermining their authority by either making them chiefs with responsibility for unpopular taxation, or denying them the chiefship and with it any vestige of authority (MacGaffey, 1970).

Colonial Health Services

Early colonial doctors played a role closer to that of pre-colonial healers than to American doctors in private practice. They came to Africa as government employees. They were expected to oppose measures which threatened to undermine government authority. And they played a direct part in supporting production. Their autonomy as European professionals was in conflict with the discipline normally expected of government employees. The cultural distance between the doctors and the African population, along with the racist ideologies of the period, made it less likely that colonial doctors would ally themselves with the subject population.

The most important stream of early colonial biomedicine is one to which scholars have paid relatively little attention: military medicine (Cantlie, 1974). The earliest doctors in most of colonial Africa were military doctors: the West African Medical Service of the British, Egyptian Army doctors in the Sudan, Indian Army doctors in British East Africa, the Schutztruppe in German East Africa, and French military doctors in West Africa. Military doctors served the state. They rarely made common cause with Africans who were deprived of their health by being moved from their villages, or by being made to serve as forced




laborers. Their main responsibility was for the health of soldiers. They then served a widening circle extending to other government employees, then other whites, then Africans employed by the government, and then to African prisoners. Military medicine tended to treat epidemic but not endemic diseases, to ignore the problems of women and children (unless they were military dependents), and to establish encampments inside which there was tight environmental control and outside which there was little responsibility. All of these appear to me to have had an impact on medical practice in Africa, of a sort which scholars have not discussed. The pattern of environmental control in military encampments is one of the roots of urban segregation, ostensibly on health grounds, which Swanson (1977) has discussed as "the sanitation syndrome." Where prevention became important, as in the French colonies, men of the medical services marched through the countryside rounding people up and innoculating them, and then returned to base (see McKelvey, 1973 on Jamot). The military period came to an end sooner or later-in most places by the mid- 1920s-but left its legacies for the future.

Because military doctors were expected to serve the interests of the colonial authorities rather than the interests of the African population, African M.D.'s were unsuitable in the years after conquest. African doctors with degrees from British universities had served the military in the West African coastal colonies of the nineteenth century (Fyfe, 1972; Adeloye, 1977). Adell Patton (1980, 1981) shows that after the colonies expanded beyond their limited coastal spheres, the British decided to exclude African doctors from the normal ranks of the Medical Service, and apparently discouraged the practice of sending Africans for training as M.D.'s. The objections to African doctors were that Africans might achieve positions where they could give orders to white officers, and that the races would mix at the mess table. But there might also have been fears that it would be difficult to challenge African doctors who were able to argue with full technical authority on matters of social and economic policy, for example the health consequences of labor recruitment or of urban segregation. The writings collected by Adeloye (1977) show that this generation of early African physicians had a remarkable awareness of public health issues and of environmental determinants of health. After leaving the colonial service, most of them provided private care for the tiny elite of urban Africans.

The most substantial recent histories of health services in British colonies (Bayoumi, 1979; Schram, 1971; Patterson, 1981) discuss the equity and rationality of distribution in terms of a division between curative and preventative services. They all agree that prevention was underemphasized. Turshen, in her dissertation (1975), argues that curative services were in harmony with capitalism and were therefore favored in colonial health policy. The argument is not a convincng one. The French placed much more emphasis than the British on mobile health teams, hygiene, prevention, and maternal and child health. Lasker describes the French decision that they could achieve more significant results with less money if they emphasized prevention. They wanted to foster population growth in order to preserve the labor supply. Lasker (1977: 283) quotes the Minister of Colonies as writing, in 1923, "The growth of colonial population is, . . . most especially, a question of labor, of preserving the population and the birth rate, to be accomplished by a major program of hygiene, medical assistance, and education."




Most of the pressures favoring prevention would have applied equally to the British. The differences emerged, I think, because Equatorial Africa was the scene of some of the worst population disasters of the early colonial period, and because the French were more sensitive to questions of population in the metropole. The British, throughout the period before World War II, behaved as though the labor force would reproduce itself in the rural regions while paying a wide range of its own social costs.

It is important not to assume that colonial prevention, where it existed, was a panacea. The medical police could be oppressive. Janzen (1979: 209) describes public health measures in the Belgian Congo which were used as punishments for local resistance. Hygienic meaures in the French colonies often amounted to chastising African mothers for abhorrent practices. Africans sometimes saw curative care as a positive contribution (hence its importance to those who wanted to win hearts and minds), and prevention as oppressive. But this point is much too simply put. Fanon (1967: 123), in one of the classic statements on colonial medicine, explored the ambiguities of being treated medically by one's conquerors. "The sudden deaths of Algerians in hospitals," he wrote, "are interpreted as the effects of a murderous and deliberate decision." Mburu (1977) explores some of the same issues in a discussion of what he calls the "Prospero- Caliban Situation."

Questions of prevention as opposed to cure need to be considered within the context of the distribution of social costs in colonial Africa. The continent's rulers provided medical services to the cities but not the countryside, to men but not to women and children, and to the rich but not the poor. The first two of these inequalities were complements of an economic policy which used rural production (either to feed male workers or to provide export crops) without significant investment in rural areas, and which therefore depended on rural social forms, including healing, to bear significant social costs.

Because colonial medical services had as their earliest mission the protection of the health of whites, services tended to concentrate in places where whites lived in large numbers, mostly in cities. This provided a basis for meeting the next urgent need-for health care at places where large numbers of Africans were employed, so that the process of production would go on. In the Ivory Coast, according to Lasker (1977: 288), this meant that in 1952 all of the territory's eleven major hospitals and twenty-six of its thirty-seven medical posts were located in the rich south of the country. In the French colonies hygiene and vaccination campaigns were the medicine of the poor, and curative services were provided for the rich and the city-dwellers (Sankale, 1969). Lasker (1977: 291) quotes an Ivorian doctor as saying that "the villagers saw physicians only when they came (accompanied by soldiers) for vaccinations; but when the local missionary fell ill, they saw that he was sent to Abidjan to be healed." Turshen (1975, 1984) and Nsekela and Nhonoli (1976) describe the urban emphasis in colonial Tanganyika. Turshen also shows that health services were better in plantation districts than in ones devoted to labor reserves. In Southern Rhodesia, as in the British West African colonies, medical services were concentrated at the workplace (Aidoo, 1982; Gelfand, 1976: 100; Fetter, 1981 discusses spatial distribution of services).

Reports from most parts of the continent attest to continuation after independence of the rural-urban inequalities and class inequalities which began in the colonial period. Frankenberg and Leeson (1974) show that the cost of four




beds at the university hospital in Lusaka could finance a health center for 20,000 people. According to Okafor (1982), Bendel State in Nigeria has urban areas with plentiful medical care, and rural ones where the doctor/population ratio is sometimes as high as 1:70,000. In Kenya, according to Mburu (1981: 22) the urban doctor patient ratio is 1:987, while the rural ratio is 1:70,000. This is not all the legacy of colonialism, for it is reinforced by today's class structure, but the inequalities were built in from the start.

A third set of inequalities, alongside urban-rural ones and those of class, were those in health services provided to men as opposed to women and children. This was in part a consequence of the total pattern of paying social costs at the work place while the population reproduced itself in the rural areas. Eddy (1980: 37), who recalls long service in Sierra Leone, reports that women patients outnumbered men in the hospitals only after 1950. Acording to Schram (1971: 307), the average Nigerian government hospital had no children's ward before 1950. Services for women in the British colonies were most often the work of women doctors or midwives who worked for mission societies. (For a similar case in South Africa see Gaitskell, 1983. In all the British colonies maternal and child health services were more the exception than the rule, although the Sudan had a successful midwife school from 1921, thanks to the individual efforts of one determined woman (Squires, 1958: 61ff.). In the Gold Coast women physicians (who tended to provide services for women and children) were denied entry in the regular medical service, and could join only on inferior terms (Patterson, 1981: 14-15). In Southern Rhodesia the first small number of African maternity assistants were trained only after World War II (Gelfand, 1978). Along with the neglect of women and children went an underemphasis on nutrition before about 1945 (Turshen, 1984, 1977). Men who went to work in mines or on plantations were fed once they got there. The exceptions, in the Belgian Congo for example, were places where labor had been stabilized, where whole families came to the workplace, and where their welfare therefore became the concern of colonial or corporate physicians (Janssens, 1980; Sabben-Clare, Bradley, and Kirkwood, 1980: 226-27).

Were physicians unaware of inequalities in the distribution of ill-health and of health care? Researchers have not explored this question seriously. A report in the Tanzania National Archives, written in the 1950s by a plantation doctor, complains that all his patients are men, even though the women and children of the area were in desperate medical need. The perception must have been common, but what was a doctor to do about the situation? In most countries professionals who wish to improve social services develop constituencies among those who would benefit. But in the colonial situation the ultimate arbiter was in the metropole; employers of labor were well-represented there, but the potential African beneficiaries of improved health care had little influence in the colonial mother-country. What influence they had emerged from the politics of resistance and of nationalism-revolving around an ideology and a form of political activity European doctors in colonial service were unlikely to find sympathethic.

Some doctors in the British colonies expected mission medicine to fill in the services for women and children. Where government concentrated on cities, plantations, mines, and railroads, the missions established hospitals in isolated rural areas. However, total facilities (of both government and missions) rarely reached more than 20 percent of the population, according to estimates of the Christian Medical Commission; the figure cannot be precise, but the point is




clearly valid (Akerele, Tabibzadeh, and McGilvray, 1976; McGilvray, 1975/76; see also Pro Mundi Vita, Bulletin 21, 1967, Medical Activities by the Church in Africa).

Even these inadequate services were a result of great improvements in the years after World War II, when rural services and services for women and children expanded. The improvements appear, however, to have stemmed less from pressure exerted by doctors than from basic changes in labor policy. In East Africa, for example, British changes in economic policy led to changes in the distribution of social costs. The British wanted to expand exports of primary products and to substitute African manufactures for imports from outside the sterling zone. Their decision, therefore, was to stabilize labor, and to expand the payment of social costs. The labor advisor to the Secretary of State for the Colonies wrote about stabilization at the time (Orde Browne, 1946: para. 58): "This will of course entail heavier charges on industry for the social services at present lacking or superfluous; but the great economy effected, and the resultant increase in efficiency should counterbalance this expense, and eliminate any increase in the cost of production which might handicap competition with trade rivals in other countries." The increased social services, in this case, included health, housing, and education.

The great expansion of health services in most African countries came after independence when a permanently urbanized labor force grew rapidly, and when new African governments needed to show that self-rule would improve the lives of citizens. Tanzania, for example, had 875 health centers and dispensaries in the year of its independence and 1443 ten years later (van Etten, 1976: 39). In Nigeria, a period of very rapid expansion came in the late 1970s, in an attempt to overcome rural-urban imbalances (Onokerhoraye, 1984: 110-16). In the great majority of African countries health services grew in the post-independence period much more rapidly than before. The typical pattern, however, was one in which demand continued to outrun supply. Within each country regions which benefited from expansion expected still more, and impoverished regions demanded to catch up.

Biomedicine Reaching Out It should be clear by this point that building a bigger health service does not

necessarily lead to better health care or to improved states of health. We saw in an earlier section that major efforts to improve health would require planners to take control of the social costs of production. This is enormously difficult because the distribution of social costs emerges from the most basic struggles of groups competing for dominance over society's direction. No narrow bureaucratic decree can impose a solution concerning these issues. Health care, like health, is resistant to directed change from above. This essay has shown that decisions on the management of illness are usually made by relatives and neighbors of the sick person-not by doctors, and not by popular healers (unless they are related to the patient). For health planners to take control of treatment they would need to intervene effectively in the workings of therapy managing networks. This is not easy.

The experiments at integrating so-called "traditional practitioners" into government health services are attempts to solve several recalcitrant problems at the same time. National governments which cannot meet the exploding demand




for health services can, with the stroke of a pen, declare thousands of popular healers to be part of the health service. Planners hope that the popular healers can serve, after added training and careful integration into the health services, as a large corps of useful workers (WHO, 1976). Recognizing popular healers is also a way of affirming the value of African technology and culture, denied by white rulers during their years of control. National medical establishments have been eager, also, to build their own pharmaceutical industries using herbs drawn from popular healing. Institutes in a number of countries are testing herbal therapies for this purpose.

Actual attempts to integrate popular healers have faced difficult problems almost everywhere. In most places the healers themselves have been eager for government recognition, although it is doubtful that they would welcome losing their freedom to practice as they wish and becoming instead carefully controlled junior health workers within the government system.

In the system as it has evolved since the time of colonial conquest, popular healers are free of any control. There is therefore no way of regulating their practice or of distinguishing between useful and damaging therapies without creating a whole new system of regulation, of a kind which would likely destroy popular healing as it is known today. Ministries of health review the credentials of physicians, but have no practical way of judging whether someone is qualified to practice in any of the sub-specialties of popular healing. Popular healing is illegal in most African countries. French law, which is still on the books in most of the nations which had been French colonies, makes it illegal for anyone but a licensed physician to perform "medical acts." British law makes alternative medical practice legal so long as the practitioner does not take measures "implying that he is registered," but so-called "anti-witchcraft laws" in most British colonies were so broad as to prohibit almost any kind of popular medical practice (Stepan, 1983: 297, 311). In all colonies, popular healing was able to continue because the laws were rarely enforced.

Healers' organizations have only limited utility as an alternative source of regulation. Some, like the Yoruba associations Oyebola (1981) describes, have histories going back into the pre-colonial period. Other associations are relatively new. Their number has grown rapidly in recent years. They have proven valuable in defending the interests of popular healers. But they cannot function as regulatory bodies without full government backing, of a kind they are unlikely to win. This is not to say that popular healing is totally without standards. For example, Ifa diviners among the Yoruba go through long and careful training careers, with periodic checks on competence after the end of training (Abimbola, 1968, 1976). In Central Africa, one normally becomes a healer in an established drum of affliction through a long process of treatment in which sufferer learns cultic knowledge. Witchfinding movements are methods of mobilizing popular sentiment to prevent some practitioners from healing in the future (Parkin, 1968). But in general there are few practical ways of preventing people from declaring themselves to be healers.

The central problem can be seen in Chavunduka's discussion of the Witchcraft Suppression Act of 1899, which was a part of Zimbabwe's law when the country achieved majority rule. The law made many normal medical practices illegal, and forbade the levelling of witchcraft accusations, but it did not prohibit witchcraft itself. The fundamental problem, according to Chavunduka (1980: 134), is that "when an individual in this society accuses




another of witchcraft he may well be right." In Chavunduka's view, the aim of the law should be to prohibit malignant practices while permitting legitimate practitioners to carry on. But it is virtually impossible for secular African nation-states to go far enough in accepting the theoretical bases of popular healing to make this distinction in any meaningful way. Bibeau (1982: 1846) acknowledges "the impossibility for modern legal codes as they now stand to incorporate the fundamental conceptions . .. which underly traditional therapy," but argues nevertheless that African governments should license popular healers after collecting systematic information about the record of each healer's practice. Whether or not this is theoretically possible, it has not in fact been done systematically anywhere. Thorough-going acceptance of popular healing would mean thorough-going regulation, so that full integration would, I believe, destroy popular therapeutics as it is known today.

Even modest experiments at a local level involve many of the same issues: how are biomedical practitioners to use the services of popular healers if they do not accept the basic premises of popular practice? The history of integrating biomedical and popular healing in Africa has been much stronger on international and national pronouncements than on actual practice, except in midwifery and the study of herbal knowledge (Ampofo, 1977; Bannerman, 1977, 1981). Barbara Pillsbury (1982) reviews the actual use of popular healers in biomedical health services all around the world. Midwives have received formal training in sixteen African countries, of which only two have created national programs, as opposed to small scale pilot projects. Pillsbury lists two African countries (Nigeria and Zimbabwe) in which popular healers are used in their healing roles within the national health system. Neither of these cases bears closer examination: in Zimbabwe the official council which was to regulate integration under the Traditional Medical Practitioners' Act of 1981 has never met; the majority of members of the Nigerian Medical Association, in discussions of actual integration, argued that recognizing traditional healers was like licensing killers (Oyebola, 1983; Chavunduka, 1983).

A few rare imaginative efforts show that some form of modest cooperation is possible. Dennis Warren and his associates developed a successful hospital-based training scheme for practitioners in Techiman, in Ghana (Warren et al., 1982). The distinguishing features of the program show, however, that even though it might deserve imitation, it is not likely to serve as a model for national programs. Warren had worked with Techiman-Bono healers and had developed rapport with them over a ten-year period before starting the training program. Preparatory discussions lasted for some months. The large training team worked with eight to ten healers at a time, for training periods of fourteen weeks. And the learning goals were modest-the healers felt that they benefited most from sessions on preparation, storage, and preservation of herbs, although they were also instructed on diarrhea and rehydration, nutrition, and a range of other subjects. The probability is small that national programs would go about their business with the same patience and respect.

The general picture at this time is that experiments at integrating popular healers into biomedical services have not made great contributions to African health. Training programs for midwives have been significant, as have reviews of herbal therapies. But popular healers at this time seem far more effective working under their own rules than as adjuncts in national health services.




A second possible strategy for biomedicine reaching out to influence the social conditions of health is through work to change the distribution of the social costs of production. The problem here is that the health workers need to find a political constituency-to form a coalition of those who would work for change and have the power to win it.

Shula Marks and Neil Andersson (1983) explore the fascinating history of the South African National Health Commission of 1944, on which a very special set of physicians proposed the creation of a health service for South Africans of all races. The Commission's proposal had the possibility of winning support within the political-economic context of South Africa at that time. The proposal came at a time when the capacity of the reserves to support their African population was declining, and when manufacturing was growing very rapidly. Manufacturers needed stable semi-skilled labor. Manufacturers who invested in training labor wanted to get the best return for their investment-not to sustain losses because of ill health. In the mining sector, the African workers, who struck in 1946, made demands which showed a similar desire for stabilization. From the government's point of view, a well-planned health service offered to keep the work force healthy for less cost than an expansion of existing hospital-based care. Labor stabilization and improvements to health services tend to go together. The Health Commission's plan, which would have provided benefits for poor whites as well as for Africans, had the possibility of helping to create a political coalition for fundamental change in the South African economy, uniting all the groups in favor of labor stabilization. As it happened, of course, the possibility of stabilization, which was real in the years after the war, was turned back by the coalition of forces which won the election of 1948 and codified apartheid. Apartheid brought the disastrous health policies which have been described in the section on the social costs of production.

Today the Industrial Health Research Group at the University of Cape Town is working on a much more modest scale to make common cause with some of those who have been deprived of health. The group provides advice on health and safety issues to worker organizations. It provides expert testimony under the unenforced occupational health laws, advice to unions on how to preserve the health of members, and on how to use the laws to maximum advantage. The Group's research helps to uncover systematic but well-hidden ties between the stresses at the work place and ill-health among the workers, as for example in the case of the high prevalence of hypertension among stevedores, or the way African workers are deprived of compensation for poor lung function by a system of valuation which sets different standards for each ethnic group (Myers, 1984; Myers, White, and Cornell, 1982; Myers and Steinberg, 1983; Cornell and Kooy, 1981).

University-based health workers, like those of the Industrial Health Research Group, have greater autonomy-a greater capacity to establish links with those deprived of health-than do employees of national health services. There is a peculiar trap here. Medical ethics and the norms of public employment require that state health workers serve the entire population, and that they do so within the pre-existing political framework. It is therefore difficult to make alliances with those who have been deprived. To mobilize the deprived is to challenge the established order of political priorities. Even progressive political regimes do not welcome challenges to their own authority. I doubt, for example, that the Tanzanian government would welcome a workers' movement, in alliance with




doctors, to change the basic organization of work in breweries, or textile factories. The Mozambique government tried to create a framework within which health workers would make alliances across narrow professional lines, by requesting that they form councils-for example, within hospital wards for collective managment (Walt and Melamed, 1983). But these councils appear not to have been expected to reach out into factories or farms, to improve the health conditions of work. When the councils had functioned for a period, the President gave a speech stressing that councils ought not to challenge doctors' authority, and that party leaders must be treated in separate wards (Walt and Melamed, 1983: 37). The Mozambique health services were, nevertheless, far more determined to extend health care to the underserved majority in the countryside than most other African governments, and more successful also at controlling priorities by limiting the freedom of action of international drug companies.

Cases of health workers reaching out effectively to the deprived are rare. The most common form of political action, within national health services, is professionally ethical and politically populist. The professional and political tendencies both are expressed as a need to serve the entire public, to redistribute resources so as to serve those who need help most, but to do so without challenging the establishd framework of power and class relations. This is usually the best we can hope for. It is the approach which has come to be known as Primary Health Care (International Conference on Primary Health Care, 1978). The approach leads to valuable work which must be respected, but which is necessarily limited by the contradictions of the society in which it is implemented.

Ronald Frankenberg and Joyce Leeson, in an article about "Health Dilemmas in the Post-colonial World" (1974), discuss the contradictions in their own work in Zambia. They quote Morley: "Three quarters of our population are rural, yet three quarters of our medical resources are spent in the towns where three quarters of our doctors live. Three quarters of the people die from diseases which could be prevented at low cost and yet three quarters of medical budgets are spent on curative services" (for this general position see King, 1966). Frankenberg and Leeson (1974: 265) then begin their lament: "The Republic of Zambia was committed to many of these [improvements to health services] by the First National Development Plan, yet they proved difficult to put into practice. Despite King, as Professor of Social Medicine, one of us (R.F.) as a leading administrator, a sympathetic Vice-Chancellor and Deputy Vice- Chancellor, a friendly Ministry of Health, and WHO backing, it was not possible to establish medical education on an appropriate basis." The description of good intentions frustrated is repeated over and over in the literature on current health services.

The IDS team which evaluated health services in Ghana describes rural-urban inequities, and then writes: "The Government has committed itself in the Plan to correcting the situation; the Government ... will emphasise ... the expansion of the primary care segment by extending the distribution of health centres and health posts to as many settlements as possible and by developing the promotive and preventive services; Government will seek the involvement of the local communities in efforts to satisfy their own simple needs" (IDS Health Group, 1981: 411). A bit later the evaluation team comments (1981: 411): "The future looks particularly bleak when even the Plan's commitment to reallocate resources was contradicted by the actual allocations within the Plan." The reasons for this




state of affairs are clear: those with the power to make decisions live in the cities, and those with the capacity to offer effective resistance to the government also live in the cities. Furthermore, the entire pattern of past investment in transportation and infrastructure favors the cities, so that rural investments must be much larger to achieve an identical effect.

Despite these gloomy truths, populist and ethical primary health care, inspired by nationalist politics, has made great achievements. In Tanganyika in the 1920s, at a time when the British had decided to provide health services only at work places and administrative centers, peasants demanded rural dispensaries and paid through their local government units to create them. Later, local nationalist politicians of the 1940s and 1950s made passionate demands for improved health services. In Lushoto District the nationalists in 1947 listed the failings of their Paramount Chief: "Usambara has no midwifery. Shambaa women give birth in miserable circumstances. Many Shambaa women die in childbirth. Why are there no maternity clinics?" The popularly-built clinics of the 1920s were the first core of the rural health system. The demands of the 1940s and 1950s helped to push along improvements being made by a government ready to begin stabilizing labor. Then, with independence, health services were among the contributions the new government could make to the well-being of its citizens. The enormous expansion in health services and piped water led to rapid declines in infant mortality and to increases in life expectancy. The achievements are real, just as the inequalities are.18 But the health services have stumbled once again, more recently, against the realities of social costs-this time in their international distribution. Dispensaries are there, but often with no drugs, for lack of foreign exchange. Water pipes are there, but with no spare parts for the valves. Maternal and child health teams are there, but without vaccines, or if they have vaccines then without refrigeration, or without transport for mobile units. Populism wins its victories, but cannot push beyond the limits on social costs.

What we are left with is pessimism of the intellect and optimism of the will. It is possible to hope for very great accomplishments and for very modest ones. Either kind is difficult. The very great accomplishments would be changes in the distribution of social costs. If the labor movement and the political movements in South Africa, for example, were able to win victories which would undermine the migrant system and the rigidity of the homelands, this would lead inevitably to enormous improvements in health. Elsewhere in Africa, where political movements have won (or might win) greater power for the rural population, improvements in health and care are likely to result. The peasant nationalism of Tanzania had this effect. If political movements were capable of winning power for rural women, this would have profound positive effects. None of these are easy to accomplish, and none can be written in any simple way into a national health plan. But it is only broadly based political or economic change which can alter the entire system of constraints on the way people live and die. In the absence of broad power change, planners must of course continue to work for the redistribution of health resources, for better rural health, for improved water supplies, for health education, for maternal and child health, and so on. It would be disastrous if recognition of the major constraints destroyed the will to act constructively.

And then there are the initiatives which make no great claim to changing entire health systems, which are modest in scale, but hopeful. They share one




characteristic: they are built on the assumption that those who suffer ill health are the ones most likely to create effective techniques for changing health conditions, and those who deliver primary care are the most likely to reform it, given the appropriate opportunity. Health decrees from on high have not had a good record in modern Africa. A few brief examples will illustrate the case.

Two examples from a project in Niger show how a concern for primary care can either increase or undermine the power of those most likely to help with primary health (Belloncle and Fournier, 1975). A project in the 1960s aimed at improving the health of new-born infants by changing birthing practices. The first trainees were older women who were serving already as midwives in their own villages. The older women did not easily accept the advice they were given-they resisted, and quarreled with their trainers; they accepted some advice and rejected some. The project organizers then tried to change the basis of recruitment, to recruit young girls out of school rather than older midwives. These would much more easily accept the judgments of their trainers. But village women trust the judgment, in these matters, only of older women who have experienced motherhood. In the event, older midwives continued to supervise most births. Evading their control was merely a way of escaping the realities of lay health practices.

The second example from the same project, in the same period, trusted those who do the work to assess the situation. Nurses at local dispensaries were instructed to cooperate with and to supervise an emerging stratum of local-level health workers. They resisted. They were then given the job of assessing the spatial distribution of health care in their own district. Each nurse recorded the home villages of patients who attended dispensary. The nurses found that dispensaries drew patients from very small areas, and that large parts of the district never sent patients to any dispensary. The nurses then sought out the help of local-level workers because their own analysis had shown that this was necessary.

One project for training health care workers in Tanzania, and another for improving health in villages near a hospital in Sierra Leone, both use related approaches. In the Tanzanian project, based on the idea of participant research (Swantz, 1982), a small team of health care workers must report in detail on what the people of a single village see as their most important health problems, what they now do about those problems, why it is that some difficulties remain, and what they expect health workers to do (von Troil, 1982). The same approach is carried further in the Sierra Leone village projct, where villagers are expected to do collective research to construct a picture of morbidity and mortality among children (Edwards and Lyon, 1983). The village health committee can then ask for the help of hospital workers at dealing with the most urgent problems. The villagers themselves make the crucial decisions. The hospital workers serve as technical consultants. The villagers then intervene to make improvements, and continue to study child health to learn whether these are working.

At this modest level, and at the level of national health plans, the lesson is similar. Health professionals do not make the most important health-giving decisions. At the local level, networks of people caring for their own health within their home communities decide on the crucial changes. At the national level, political movements and political factions of all kinds struggle to influence the shape of the polity and of the economy, and through them of health. In either case the professionals, the government workers, and the "experts" make




important but marginal contributions. They are most constructive if they recognize that their own role is secondary, and that they must try to identify and to serve that special sub-group of laypeople who are already at work to improve health.

NOTES

Special thanks are due to Allen Isaacman, John Janzen, and Gerda Lerner for encouragement and helpful criticism. Thanks should also go to Arthur Kleinman, Jean Comaroff, Antony Klouda, Michael MacDonald, Jan Vansina, and Ellen Brickwedde for comments on drafts.

1. This essay uses the terms popular medicine and biomedicine instead of the more common terms "traditional" and "western." Traditional intellectuals, in the terms defined below, exist in Europe as well as in Africa. "Tradition" cannot therefore be used on only the African side of the equation. Nor is biomedicine solely "western." African physicians have practiced medicine since the nineteenth century.

2. For a later review of these issues see Iliffe (1979); on eastern Zambia see Vail (1977). 3. On infertility in contemporary Africa, see also Adadevoh (1974). 4. The demographic consequences of the slave trade have been the subject of a large and

impressive literature: Curtin, 1968; Curtin, 1969; Austen, 1979; Fage, 1975; Manning, 1981; Inikori, 1976; Curtin, Anstey, and Inikori, 1976.

5. Bayoumi (1979: 203) on the importance of an integrated approach to water projects see Coumbaras 1977b; Hunter, Rey, and Scott, 1982.

6. See the contributions on the cost of schistosomiasis in Abdallah (1978). 7. See Patterson (1981: 37) on forest clearing for cocoa; Roberts (1974: 306) on the

enormous expansion of malaria in western Kenya. 8. On the nature of this process, the fullest and most interesting interpretation is by

Chapin and Wasserstrom (n.d.). 9. For an early exception see Richards (1939) and Wisner (1983).

10. Haswell (1953, 1963, 1975, 1981); for a similar process in Ghana see Bukh (1979). 11. For the fascinating but grim story of the way tuberculosis spread from the mines to the

rural population, see Packard (1983). 12. On the issue of rationality see Bryan Wilson (1970) and Young (1981). 13. Shweder (1981) cites literature on this subject. 14. For further discussion of natural causation in Evans-Pritchard, see Janzen (1981). 15. Private communication 22 October 1984. 16. The International African Institute sponsored an important conference on the

professionalization of medicine in Africa, held at Gaborone in September, 1983. Murray Last and Gordon Chavunduka are editing a volume of the conference proceedings.

17. For one of the rare attempts at reconstructing pre-colonial healing see Waite (1981). 18. For the inequalities see Klouda (1983).

REFERENCES Abbreviations: IJHS International Journal of Health Services JAH Journal of African History SSM Social Science and Medicine Abdalla, Ismail Hussein. 1981 a. "Medicine, Authority and Kinship: The Case of the Sokoto

Jihad." Paper delivered to African Studies Association Meeting, Bloomington, Indiana. . 1981b. "Islamic Medicine and its Influence on Traditional Hausa Practitioners in

Northern Nigeria." Ph.D. Dissertation, University of Wisconsin-Madison.




Abdallah, Ahmed (ed.). 1978. Proceedings of the International Conference on Schistosomiasis, Cairo, October 18-25, 1975. Vol. 1. Cairo: Ministry of Health, Egypt.

Abimbola, W. 1968. "Ifa as a Body of Knowledge and as an Academic Discipline." Lagos Notes and Records 2/1:30-40.

- . 1976. Ifa: An Exposition of Ifa Literary Corpus. Ibadan: Ibadan University Press. Achebe, Chinua. 1958. Things Fall Apart. London: Heinemann. Adadevoh, B. Kwaku (ed.). 1974. Sub-Fertility and Infertility in Africa. Ibadan: The Caxton

Press, Ltd. Adeloye, Adelola (ed.). 1977. Nigerian Pioneers of Modern Medicine: Selected Writings.

Ibadan: Ibadan University Press. Ademuwagun, Z. A. 1979. "Problem and Prospect of Legitimizing and Integrating Aspects

of Traditional Health Care Systems and Methods with Modern Medical Therapy: the Igbo-Ora Experience," pp. 158-64 in Ademuwagun, et al. (eds.) African Therapeutic Systems.

Ademuwagun, Z. A., J. A. A. Ayoade, I. E. Harrison, and D. M. Warren (eds.). 1979. African Therapeutic Systems. Waltham, Mass.: Crossroads Press.

Adler, Alfred, and Andras Zempleni. 1972. Le baton de l'aveugle: divination maladie et pouvoir chez les Moundang du Tchad. Paris: Hermann, Collection Savoir.

African Water and Sewage. 1982. "Lamu-A Simple Blueprint That Others Might Follow." pp. 14-16.

Agarwal, Anil. 1978. "Malaria Makes a Comeback," New Scientist. 77, February 2: 274-77. . 1979. "Pesticide Resistance on the Increase, says UNEP." Nature, 279, May 24:

280. Aidoo, Thomas Akwasi. 1982. "Rural Health under Colonialism and Neocolonialsim: A

Survey of the Ghanaian Experience." IJHS 12/4: 637-57. Ajaegbu, H. I. 1977. "The Demographic Situations in Pre-Colonial and Early Colonial

Periods in West Africa: An Assessment of the Usefulness of Non-Conventional Data Sources," pp. 169-94 in University of Edinburgh, Centre of African Studies African Historical Demography.

Akerele, O., I. Tabibzadeh, and J. McGilvray. 1976. "A New Role for Medical Missionaries in Africa." WHO Chronicle 30: 175-80.

Al-Naqar, 'Umar. 1972. The Pilgrimage Tradition in West Africa. Khartoum: Khartoum University Press.

Ampofo, Oku. 1977. "Plants that Heal." November: 26-30. Armah, Ayi Kwei. 1979. The Healers. London and Ibadan: Heinemann. Asuni, Tolani. 1979. "Modern Medicine and Traditional Medicine," pp. 176-81 in

Ademuwagun et al. (eds.) African Therapeutic Systems. Austen, Ralph. 1979. "The Trans-Saharan Slave Trade: A Tentative Census," in Henry A.

Gemery and Jan S. Hogendorn (eds.) The Uncommon Market: Essays in the Economic History of the Atlantic Slave Trade. New York: Academic Press.

Azevedo, Mario Joaquim. 1978. "Epidemic Disease Among the Sara of Southern Chad, 1890-1940," pp. 118-52 in Hartwig and Patterson Disease in African History. An Introductory Survey and Case Studies.

Bannerman, R. H. 1977. "WHO's Programme in Traditional Medicine." WHO Chronicle 31: 427-28.

1981. "The Virtues of Traditional Therapy." West Africa. 65, 18 May: 1083-87. Bayliss-Smith, Tim. 1981. "Seasonality and Labour in the Rural Energy Balance," pp. 30-

38 in Chambers, Longhurst, and Pacey (eds.) Seasonal Dimensions to Rural Poverty. Bayoumi, Ahmed. 1979. The History of Sudan Health Services. Nairobi: Kenya Literature

Bureau. Beck, Ann. 1970. A History of the British Medical Administration of East Africa.

Cambridge, Mass.: Harvard University Press. 1977. "Medicine and Society in Tanganyika 1890-1930: A Historical Inquiry."

Transactions of the American Philosophical Society 67/3.




.. 1979. "The Traditional Healer in Tanzania." Issue 9/3: 2-5.

- . 1981. Medicine, Tradition and Development in Kenya and Tanzania, 1920-1970. Waltham, Massachusetts: Crossroads Press.

Belloncle, Guy, and Georges Fournier. 1975. Sante et developpement en milieu rural Africain. Paris: Editions economie et humanisme.

Belsey, Mark A. 1976. "The Epidemiology of Infertility. A Review with Special Reference to Sub-Saharan Africa." Bulletin of the WHO 54: 319-42.

Benyoussef, A., J. L. Cutler, A. Levine, P. Mansourian, T. Phan-Tan, R. Baylet, H. Collomb, S. Diop, B. Lacombe, J. Ravel, J. Vaugelade, and G. Diebold. 1974. "Health Effects of Urban-Rural Migration in Developing Countries-Senegal." SSM 8: 243-54.

Berger, Iris. 1981. Religion and Resistance: East African Kingdoms in the Precolonial Period. Annales, no. 105. Tervuren: Musee Royal de l'Afrique Centrale.

Bibeau, Gilles. 1982. "New Legal Rules for an Old Art of Healing: The Case of the Zairian Healers' Associations." SSM 16/21: 1843-49.

"Biological Control of Insect Vectors of Disease." 1978. B WHO. 56/3: 377-78. Brader, L. 1979. "Integrated Pest Control in the Developing World," Annual Review of

Entomology 24: 225-54. Brown, A. W. A., J. Haworth, and A. R. Zahar. 1976. "Review Article: Malaria Eradication

and Control from a Global Standpoint." Journal of Medical Entomology 13/1: 1-25. Bruce-Chwatt, L. J. 1974. "Resurgence of Malaria and Its Control." Journal of Tropical

Medicine and Hygiene 77/4 suppl.: 62-66. and Julian de Zulueta. 1980. The Rise and Fall of Malaria in Europe. Oxford:

Oxford University Press. Bruijning, C. F. 1980. "Control of Schistosomiasis in the People's Republic of China," pp.

151-56 in Health Policies in Developing Countries, edited by Clive Wood and Yvonne Rue. Royal Society of Medicine, International Congress and Symposium Series, no. 24. London.

Bukh, Jette. 1979. The Village Woman in Ghana. Uppsala: Scandinavian Institute of African Studies.

Busvine, James R. 1978. "Current Problems in the Control of Mosquitoes." Nature 273, June 22: 604-07.

Buxton, Jean. 1973. Religion and Healing in Mandari. Oxford at the Clarendon Press. Caldwell, John C. 1975a. "Introduction: Some Important Issues," pp. 3-28 in Caldwell

(ed.), with Addo, et al. Population Growth and Socio-economic Change in West Africa. 1975b. "Fertility Control," pp. 58-97 in Caldwell (ed.), with Addo, et al.

Population Growth and Socio-economic Change in West Africa. 1. 975c. The Sahelian Drought and its Demographic Implications. Washington, D.C.:

American Council on Education, Overseas Liaison Committee, Paper No. 8. . 1976. "Toward a Restatement of the Demographic Transition Theory." Population

and Development Review 2: 321-66. 1977. "Major Questions in African Demographic History," pp. 7-22 in University

of Edinburgh, Centre of African Studies African Historical Demography. . 1978. "A Theory of Fertility: From High Plateau to Destabilization." Population

and Development Review 4: 553-77. . 1985. "The Social Repercussions of Colonial Rule-Demographic Aspects,"

chapter 18, in A. Adu Boahen (ed.) Africa Under Colonial Domination, 1880-1935. Volume 7, UNESCO General History of Africa. London and Berkeley: Heinemann and University of California Press.

. (ed.), with N. O. Addo, S. K. Gaisie, A. Igun, and P. O. Olusanya. 1975. Population Growth and Socio-economic Change in West Africa. New York and London: Columbia University Press.

, and Pat Caldwell. 1977. "The Role of Marital Sexual Abstinence in Determining Fertility: a Study of the Yoruba in Nigeria." Population Studies 31/2: 193-213.

Caldwell, Pat, and J. C. Caldwell. 1981. "The Function of Child-Spacing in Traditional Societies and the Direction of Change," pp. 73-92 in Hilary J. Page and Ron




Lesthaeghe (eds.) Child-Spacing in Tropical Africa: Traditions and Change. London: Academic Press.

Cantlie, Sir Neil. 1974. A History of the Army Medical Department. vol. 2, Edinburgh and London: Churchill Livingstone.

Cantrelle, Pierre. 1975. "Mortality: Levels, Patterns, and Trends," pp. 98-118 in Caldwell, J. (ed.) with Addo, et al. Population Growth and Socio-economic Change in West Africa.

, B. Ferry, and J. Mondot. 1978. "Relationships between Fertility and Mortality in Africa," pp. 181-205 in Samuel H. Preston (ed.) The Effects of Infant and Child Mortality on Fertility. New York, San Francisco and London: Academic Press.

Castleman, Barry I. 1979. "The Export of Hazardous Factories to Developing Nations." IJHS 9/4:

. 1981. "Response to Levenstein-Eller Critique." IJHS 11/2: 311-13. 1983. "The Double Standard in Industrial Relations." IJHS 13/1: 5-14.

Chambers, Robert, R. Longhurst, A. Pacey (eds.). 1981. Seasonal Dimensions to Rural Poverty. London: Frances Pinter.

Chapin, Georganne, and Robert Wasserstrom. n.d. (ca. 1981) "Agricultural Production and Malaria Resurgence in Central America and India." mimeographed. Columbia University.

Chauveau, J. P., J. P. Dozon, and J. Richard. 1981. "Histoires de riz, histoires d'ignames: le cas de la moyenne Cote-d'Ivoire." Africa 51/2: 621-58.

Chavunduka, G. L. 1978. Traditional Healers and the Shona Patient. Gwelo: Mambo Press. 1980. "Witchcraft and the Law in Zimbabwe." Zambezia 8,2: 129-47.

•-. 1983. "Professionalisation of Traditional African Medicine in Zimbabwe." Paper

presented at the International African Seminar, Professionalisation of African Medicine, Gaborone.

Chayanov, A. V. 1966. The Theory of the Peasant Economy. Homewood, Ill.: Richard D. Irwin, Inc.

Cissoko, Sekene, Mody. 1968. "Famines et epidemies a Tombouctou et dans la Boucle du Niger du XVIe au XVIIe siecle." Bulletin de 1'.F.A.N. 30/3, series B: 806-21.

Cleaver, Harry. 1977. "Malaria, the Politics of Public Health and the International Crisis." Review of Radical Political Economics 9: 81-103.

Cobbing, Julian. 1977. "The Absent Priesthood: Another Look at the Rhodesian Risings of 1896-1897." JAH 18/1: 61-84.

Comaroff, Jean. 1982. "Medicine: Symbol and Ideology," pp. 49-68 in P. Wright and A. Treacher (eds.) The Problem of Medical Knowledge. Edinburgh: Edinburgh University Press.

Conco, W. Z. 1979. "The African Bantu Traditional Practice of Medicine: Some Preliminary Observations," pp. 58-80 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

Coquery-Vidrovitch, Catherine. 1972. Le Congo au temps des grandes compagnies concessionaires, 1898-1930. Paris.

. 1977. "Population et demographie en Afrique Equatoriale Francaise dans le premier tiers du XXe siecle." pp. 331-51 in University of Edinburgh, Centre of African Studies African Historical Demography.

Cordell, Dennis D. 1983. "Low Fertility in Africa: An Evaluation of the Biomedical Approach of Anne Retel-Laurentin." Canadian Association of African Studies. Quebec, Quebec. May 16, 1983.

Corin, Ellen. 1979. "A Possession Psychotherapy in an Urban Setting: Zebola in Kinshasa." SSM 13B/4: 327-38.

Cornell, Jud, and Alide Kooy. 1981. "Wiehahn Part 5 and the White Paper." South African Labour Bulletin 7/13: 51-74.

Coumbaras, A. 1977a. "Travaux hydrauliques et problemes de sante dans les pays en voie de developpement." Acta Tropica. 34: 229-48.

1977b. "Sante et irrigation," pp. 385-94 in Worthington 1977.




Cousins, Norman. 1981. Anatomy of an Illness. New York: Bantam Books. Curtin, Philip D. 1968. "Epidemiology and the Slave Trade," Political Science Quarterly.

83,2: 190-216. . 1969. The Atlantic Slave Trade: A Census. Madison: University of Wisconsin Press.

~- , Roger Anstey, and J. E. Inikori. 1976. "Discussion: Measuring the Atlantic Slave Trade." JAH 17,4: 595-627.

Davidson, G., and A. R. Zahar. 1973. "The Practical Implications of Resistance of Malaria Vectors to Insecticides." B WHO. 49: 475-83.

Davis-Roberts, Christopher. 1981. "Kutambuwa Ugonjuwa: Concepts of Illness and Transformation among the Tabwa of Zaire." SSM 15B,3: 309-16.

Dawson, Marc H. 1978. "Disease in Nineteenth-Century Kenya." Historical Association of Kenya Annual Conference.

1979. "Smallpox in Kenya, 1880-1920." SSM 13B/4: 245-50.

•-. 1981. "Disease and Population Decline of the Kikuyu of Kenya, 1890-1925," pp.

121-38 in University of Edinburgh Centre of African Studies African Historical Demography, Vol. II.

. 1983. "Social and Epidemiological Change in Kenya: 1880-1925." Ph.D. Dissertation, University of Wisconsin.

De Craemer, Willy, Jan Vansina, and Renee C. Fox. 1976. "Religious Movements in Central Africa: A Theoretical Study." Comparative Studies in Society and History 18/4: 458-75.

Desowitz, Robert S. 1976. "How the Wise Men Brought Malaria to Africa." Natural History 85/2: 36-44.

Dias, Jill R. 1981. "Famine and Disease in the History of Angola c. 1830-1930." JAH 22/3: 349-378.

Diop, Moussa, Paul Martino, and Andras Zempleni. 1979. "Les techniques therapeutiques traditionelles des maladies mentales au Senegal," pp. 85-86 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

Douglas, Mary. 1970. "Introduction: Thirty Years after Witchcraft, Oracles and Magic," pp. xiii to xxxviii in Mary Douglas (ed.) Witchcraft Accusations and Confessions, ASA Monograph No. 9. London: Tavistock.

Doyal, Leslie, with Imogen Pennell. 1979. The Political Economy of Health. Boston: South End Press.

Dunn, Frederick L. 1968. "Epidemiological Factors: Health and Disease in Hunter- Gatherers." pp. 221-28 in Richard B. Lee and Irven De Vore (eds.), Man the Hunter. Chicago: Aldine.

Dupre, Georges. 1982. Un ordre et sa destruction: economie, politique et histoire chez les Nzabi de la Republique populaire du Congo. Paris: ORSTOM.

Dutta, Hiran M., and Ashok K. Dutt. 1978. "Malaria Ecology: A Global Perspective." SSM 12: 69-84.

Eddy, T. P. 1980. "Food Shortage as a Health Catastrophe," pp. 37-41 in Sabben-Clare, Bradley, and Kirkwood 1980.

Edgerton, R. G. 1979. "A Traditional African Psychiatrist," pp. 87-94 in Ademuwagun et al. 1979.

Edinburgh, University of, Centre of African Studies. 1977. African Historical Demography. Proceedings of a Seminar held in the Centre of African Studies, University of Edinburgh, 29th and 30th April 1977.

. 1981. African Historical Demography, Vol. II. Proceedings of a Seminar Held in the Centre of African Studies, University of Edinburgh, 24th and 25th April 1981.

Edwards, Nancy, and Mary Lyon. 1983. "Community Assessment: A Tool for Motivation and Evaluation in Primary Health Care in Sierra Leone," pp. 101-13 in David Morley, Jon Rohde, and Glen Williams (eds.) Practicing Health for All. Oxford: Oxford University Press.

Evans-Pritchard, E. E. 1937. Witchcraft, Oracles and Magic among the Azande. Oxford at the Clarendon Press.




Fabrega, H. Jr. 1977. "The Scope of Ethnomedical Science." Culture, Medicine and Psychiatry. 1: 201-28.

Fage, J. D. 1975. "The Effect of the Export Slave Trade on African Populations," pp. 15- 23 in R. P. Moss and R. J. A. R. Rathbone (eds.) The Population Factor in African Studies.

Fanon, Frantz. 1967. "Medicine and Colonialism," pp. 121-45 in A Dying Colonialism. New York: Grove Press.

Feierman, Steven. 1974. The Shambaa Kingdom: A History. Madison: University of Wisconsin Press.

1979. "Change in African Therapeutic Systems." SSM 13B/4: 277-84. . 1981. "Therapy as a System-in-Action in Northeastern Tanzania." SSM 15B/3:

353-60. Fetter, Bruce. 1981. "People and Resources in Zambia under Colonial Rule," pp. 183-229

in University of Edinburgh African Historical Demography, Vol. II. Field, Margaret Joyce. 1960. Search for Security: An Ethnopsychiatric Study of Rural

Ghana. London: Faber and Faber. Flynn, Laurie. 1982. "South Africa Blacks out Blue Asbestos Risk." New Scientist 22 April:

237-39. Ford, John. 1971. The Role of Trypanosomiases in African Ecology: A Study of the Tsetse

Fly Problem. Oxford: Clarendon Press. - . 1979. "Ideas Which Have Influenced Attempts to Solve the Problems of African

Trypanosomiasis." SSM 13B/4: 269-75. Foucault, Michel. 1973. The Birth of the Clinic. London: Tavistock.

.. 1980. Power/Knowledge.: Selected Interviews and Other Writings, 1972-1977. New

York. Fox, Renbe. 1974. Experiment Perilous. Philadelphia: University of Pennsylvania Press.

, and Judith Swazey. 1974. The Courage to Fail. Chicago: University of Chicago Press.

Frake, C. 1961. "The Diagnosis of Disease among the Subanum of Mindanao." American Anthropologist 63: 113-32.

Frank, Jerome. 1974. Persuasion and Healing. Rev. ed. New York: Schocken Books. Frankenberg, Ronald. 1980. "Medical Anthropology and Development: A Theoretical

Perspective." SSM 14B: 197-207. , and Joyce Leeson. 1974. "The Sociology of Health Dilemmas in the Post-colonial

World: Intermediate Technology and Medical Care in Zambia, Zaire, and China," pp. 255-78 in de Kadt, Emanuel and Gavin Williams (eds.) Sociology and Development. Explorations in Sociology, 4. A series under the auspices of The British Sociological Association. London: Tavistock Publications, Ltd.

Frishman, Alan. 1977. "The Population Growth of Kano, Nigeria," pp. 212-50 in University of Edinburgh, Centre of African Studies African Historical Demography.

Fry, Peter. 1976. Spirits of Protest: Spirit-mediums and the Articulation of Consensus among the Zezuru of Southern Rhodesia (Zimbabwe). Cambridge Studies in Social Anthropology, 14. Cambridge: Cambridge University Press.

Fyfe, Christopher. 1972. Africanus Horton 1835-1883: West African Scientist and Patriot. New York: Oxford University Press.

Gaitskell, Deborah. 1983. "Women, Religion and Medicine in Johannesburg between the Wars." Seminar paper. School of Oriental and African Studies, May 18.

Ganon, M. F. 1975. "The Nomads of Niger," pp. 694-700 in Caldwell et al., Population Growth and Socio-economic Change in West Africa.

Gelfand, Michael. 1974. Medicine and Custom in Africa. Edinburgh and London: E. & S. Livingstone, Ltd.

. 1976. A Service to the Sick: A History of the Health Services for Africans in Southern Rhodesia (1890-1953). Gwelo: Mambo Press.

1978. Midwifery in Tropical Africa: The Growth of Maternity Services in Rhodesia. Salisbury: Supplement to Zambezia, The Journal of the University of Rhodesia.




. 1979. "Psychiatric Disorders as Recognized by the Shona," pp. 8-14 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

Gillham, F. E. M. 1972. "The Relationship between Insect Pests and Cotton Production in Central Africa," pp. 407-22 in M. T. Farvar and J. P. Milton (eds.) The Careless Technology. Garden City, New York: The Natural History Press.

Gillies, Eva. 1976. "Causal Criteria in African Classifications of Disease," pp. 358-95 in J. B. Loudon (ed.) Social Anthropology and Medicine. Association of Social Anthropologists of the Commonwealth, Monograph 13. London: Academic Press.

Good, Byron. 1977. "The Heart of What's the Matter. The Semantics of Illness in Iran." Culture, Medicine and Psychiatry 1/1: 25-58.

Good, Charles M. 1978. "Man, Milieu, and the Disease Factor: Tick-Borne Relapsing Fever in East Africa," pp. 46-87 in Hartwig and Patterson 1978.

. 1980. "A Comparison of Rural and Urban Ethnomedicine among the Kamba of Kenya," pp. 13-56 in P. Ulin and M. Segall (eds.) Traditional Health Care Delivery in Contemporary Africa.

Gramsci, A. 1971. Selections from the Prison Notebooks. Edited and translated by Quintin Hoare and Geoffrey Nowell Smith.

Green, Phillipa, and Shirley Miller. 1980. "The Commission of Enquiry on Occupational Health," pp. 145-68 in F. Wilson and G. Westcott, Economics of Health in South Africa, vol.2: Hunger, Work, and Health. Johannesburg: Raven Press.

Gregory, Joel W., and Victor Piche. 1982. "African Population: Reproduction for Whom?" Daedalus. 111,2: 179-209.

Grollig, Francis X. and Harold B. Haley (eds.). 1976. Medical Anthropology. A part of the proceedings of the IXth International Congress of Anthropological and Ethnological Sciences. The Hague: Mouton Publishers.

Guttmacher, Sally. 1979. "Whole in Body, Mind & Spirit: Holistic Health and the Limits of Medicine." The Hastings Center Report 9/2: 15-21.

Gwassa, G. C. K. 1972. "Kinjikitile and the Ideology of Maji Maji," pp. 202-17 in T. O. Ranger and I. Kimambo (eds.) The Historical Study of African Religion. London: Heinemann.

Hahn, Robert A. 1982. " 'Treat the Patient, Not the Lab': Internal Medicine and the Concept of 'Person.' " Culture, Medicine and Psychiatry 6/3:; 219-36.

Harms, Robert. 1975. "The End of Red Rubber: A Reassessment." JAH 16/1: 73-88. 1981. River of Wealth, River of Sorrow. New Haven and London: Yale University

Press. Hartwig, Gerald W. 1976. The Art of Survival in East Africa: The Kerebe and Long-Distance

Trade 1800-1895. New York and London: Africana. 1979. "Demographic Considerations in East Africa During the Nineteenth

Century." International Journal of African Historical Studies 12: 653-72. , and K. David Patterson. 1978. Disease in African History: An Introductory Survey

and Case Studies. Durham, North Carolina: Duke University Press. Haswell, Margaret. 1953. Economics of Agriculture in a Savannah Village: Report on Three

Years' Study of Genieri Village and its Lands, The Gambia. London: Colonial Office. 1963. The Changing Pattern of Economic Activity in a Gambia Village. Department

of Technical Cooperation, Overseas Research Publication No. 2. London: HMSO.

-. 1975. The Nature of Poverty. London: Macmillan Press. 1981. "Food Consumption in Relation to Labour Output," pp. 38-41 in Chambers,

Longhurst, and Pacey (eds.) Seasonal Dimensions to Rural Poverty. Hedman, P., J. Brohult, J. Forslund, V. Sirleaf, and E. Bengtssom. 1979. "A Pocket of

Controlled Malaria in a Holoendemic Region of West Africa." Annals of Tropical Medicine and Parasitology 73/4: 317-25.

Henderson, W. 0. 1965. "German East Africa, 1884-1918," pp. 123-62 in Vincent Harlow and E. M. Chilver (eds.) History of East Africa, Vol. 2. Oxford: Clarendon Press.

Herbert, Eugenia W. 1975. "Smallpox Inoculation in Africa." JAH 16/4: 539-59.




Hill, M. N., J. A. Chandler, and R. B. Highton. 1977. "A Comparison of Mosquito Populations in Irrigated and non-Irrigated Areas of the Kano Plains, Nyanza Province, Kenya," pp. 307-15 in E. Barton Worthington (ed.) Arid Land Irrigation in Developing Countries. Environmental Problems and Effects. Oxford: Pergamon Press.

Hill, Polly. 1977. Population, Prosperity and Poverty: Rural Kano 1900 and 1970. Cambridge: Cambridge University Press.

Horton, Robin. 1967. "African Traditional Thought and Western Science." 4Africa 37: 50- 71, 155-87.

Howell, Nancy. 1979. Demography of the Dobe !Kung. New York: Academic Press. Hughes, Charles C., and John M. Hunter. 1970. "Disease and 'Development' in Africa."

SSM 3: 443-93. Hull, Harry F., Pap John Williams, and Fred Oldfield. 1983. "Measles Mortality and

Vaccine Efficacy in Rural West Africa." Lancet 8331: 972-75. Hunter, John M., Luis Rey and David Scott. 1982. "Man-Made Lakes and Man-Made

Diseases: Towards a Policy Resolution." SSM 16: 1127-45. IDS Health Group. 1981. Health Needs and Health Services in Rural Ghana. SSM 15A/4:

397-517. Iliffe, John. 1979. A Modern History of Tanganyika. London: Cambridge University Press. ILO. 1970. Conditions of Work of Women and Young Workers on Plantations. Committee

on Work on Plantations, Sixth Session. Geneva. ILO. 1977. Migrant Workers, Occupational Safety and Health. Occupational Safety and

Health Series, No. 34. Geneva. Imperato, P. J. 1979. "Traditional Medical Practitioners among the Bambara of Mali and

their Role in the Modern Health Care Delivery System," pp. 202-7 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

Imperato, P. J. and D. Traore. 1979. "Traditional Beliefs about Smallpox and Its Treatment in the Republic of Mali," pp. 15-18 in Z. A. Ademuwagun et al. (eds.) .frican Therapeutic Systems.

Ingman, Stanley R., and Anthony E. Thomas. 1973. Topias and Utopias in Health: Policy Studies. Selected papers from the International Congress of Anthropological and

Ethnological Sciences, 9th, Chicago, 1973. The Hague: Mouton Publishers. Inikori, J. E. 1976. "Measuring the Atlantic Slave Trade: An Assessment of Curtin and

Anstey." JA4H 17/2: 197-223. 1981. "Under-Population in 19th Century West Africa: The Role of the Export

Slave Trade," pp. 283-314 in University of Edinburgh African Historical Demography, Vol. II.

International Conference on Primary Health Care. 1978. Primary Health Care. Geneva and New York: WHO and United Nations Children's Fund.

Isaacman, Allen F. 1976. The Tradition of Resistance in Mozambique. Anti-Colonial Activity in the Zambesi Valley 1850-1921. Berkeley and Los Angeles: University of California Press.

Isichei, Elizabeth. 1973. The Ibo People and the Europeans: The Genesis of a Relationship-to 1906. New York: St. Martin's Press.

Jakobsen, Oddvar. 1978. "Economic and Geographical Factors Influencing Child Malnutrition: A Study from the Southern Highlands, Tanzania." BRALUP Research Paper No. 52. Dar es Salaam: University of Dar es Salaam.

Janssens, P. G. 1980. "Comparative Aspects: in the Belgian Congo," pp. 209-26 in Sabben-Clare, Bradley, and Kirkwood 1980.

Janzen, John M. 1978. The Quest for Therapy in Lower Zaire. Berkeley and London: University of California Press.

1979. "Pluralistic Legitimation of Therapy Systems in Contemporary Zaire." pp. 208-16 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

1981. "The Need for a Taxonomy of Health in the Study of African Therapeutics." SSM 15B/3: 185-94.




. 1982a. Lemba, 1650-1930: A Drum of Affliction in Africa and the New World. New York and London: Garland Press.

. 1982b. "Indicators and Concepts of Health in Anthropology: The Case for a 'Social Reproduction' Analysis of Health." Xeroxed paper.

. 1983. "On the Comparative Study of Medical Systems: Ngoma, A Collective Therapy Mode in Central and Southern Africa." Xeroxed paper.

, and Steven Feierman (eds.). 1979. The Social History of Disease and Medicine in Africa. Special issue of SSM 13B/4.

, and Gwyn Prins (eds.). 1981. Causality and Classification in African Medicine and Health. Special issue of SSM 15B/3.

Johnny, Michael, John Karimu, and Paul Richards. 1981. "Upland and Swamp Rice Farming Systems in Sierra Leone: The Social Context of Technological Change." Africa 51/2: 596-620.

Johnson, Terence J. 1972. Professions and Power. London: Macmillan. Kapp, K. William. 1950. The Social Costs of Private Enterprise. Cambridge, Mass.:

Harvard University Press. Kikhela, N., G. Bibeau, and E. Corin. 1979. "Steps Toward a New System of Public Health

in Zaire," pp. 217-24 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems. King, Maurice. 1966. Medical Care in Developing Countries. Nairobi: Oxford University

Press. Kjekshus, Helge. 1977. Ecology Control and Economic Development in East African History:

The Case of Tanganyika 1850-1950. Berkeley and Los Angeles: University of California Press.

Kleinman, Arthur. 1980. Patients and Healers in the Context of Culture.: An Exploration of the Borderland between Anthropology, Medicine, and Psychiatry. Berkeley: University of California Press.

. 1981. "On Illness Meanings and Clinical Interpretation: Not 'Rational Man', but a Rational Approach to Man the Sufferer/Man the Healer." Culture, Medicine and

Psy'chiatry 5/4: 373-77. Kloos, Helmut, Giuseppe Desole, and Aklilu Lemma. 1981. "Intestinal Parasitism in

Seminomadic Pastoralists and Subsistence Farmers in and around Irrigation Schemes in the Awash Valley, Ethiopia, with Special Emphasis on Ecological and Cultural Associations." SSM 15B: 457-69.

Kloos, Helmut, and Aklilu Lemma. 1977. "Schistosomiasis in Irrigation Schemes in the Awash Valley, Ethiopia." The American Journal of Tropical Medicine and Hygiene 26/5: 899-908.

Kloos, Helmut, and K. Thompson. 1979. "Schistosomiasis in Africa: An Ecological Perspective." J. Trop. Geog. 48: 31-46.

Klouda, Antony. 1983. " 'Prevention' is more costly than 'cure': health problems for Tanzania, 1971-81," pp. 49-63 in D. Morley, J. Rohde, and G. Williams (eds.), Practicing Health for All. Oxford: Oxford University Press.

Knutzen, V. K., and D. E. Bourne. 1977. "The Reproductive Efficiency of the Xhosa." SAMJ 51: 392-94.

Kooy, Alide. 1980. "Notes on Mine Accidents," pp. 169-77 in Westcott and Wilson (1980). Kramer, Joyce and Anthony Thomas. 1982. "The Modes of Maintaining Health in

Ukambani, Kenya," pp. 159-98 in P. S. Yoder (ed.) African Health and Healing Systems.: Proceedings of a Symposium.

Lamont, N. McE. 1973. "A Possible Form of Acute Cardiomyopathy as Encountered in Sugar Cane Field Workers in Tongaat, Natal." SAMJ 24 February: 311-15.

Lasker, Judith N. 1977. "The Role of Health Services in Colonial Rule: The Case of the Ivory Coast." Culture, Medicine and Psychiatry 1/3: 277-97.

1981. "Choosing Among Therapies: Illness Behavior in the Ivory Coast." SSM 15A/2: 157-68.

Last, Murray. 1979. "Strategies against Time." Sociology of Health and Illness 1/3: 306-17.




. 1981. "The Importance of Knowing about not Knowing." SSM 15B/3: 387-92. Laurell, Asa Cristina. 1981. "Mortality and Working Conditions in Agriculture in

Underdeveloped Countries." IJHS 11/1: 3-19. Levenstein, Charles, and Stanley Eller. 1981. "Occupational Safety and Health Regulation:

An International Perspective." IJHS 11/2: 303-9. Levi-Strauss, Claude. 1963. "The Effectiveness of Symbols," pp. 186-205 in Structural

Anthropology, translated by Claire Jacobson and B. G. Schoepf. New York: Basic Books. . 1969. The Raw and the Cooked. Translated by J. and D. Weightman. New York:

Harper & Row. Lewis, I. M. 1971. Ecstatic Religion. Harmondsworth: Penguin Books. Linares, Olga F. 1981. "From Tidal Swamp to Inland Valley: On the Social Organization of

Wet Rice Cultivation among the Diola of Senegal." Africa 51/2: 557-95. Lipsey, Richard G., and Peter O. Steiner. 1969. Economics. Second Ed. New York: Harper

& Row. Livingstone, Frank B. 1967. "The Origin of the Sickle-Cell Gene," pp. 139-66 in Creighton

Gabel and Norman R. Bennett (eds.) Reconstructing African Culture History. Boston: Boston University Press.

Lux, Andre. 1976. "Le probleme de la sterilite en Afrique et ses implications de politique demographique: a propos deux ouvrages recents." CJAS 10: 143-55.

MacGaffey, Wyatt. 1970. "The Religious Commissions of the Bakongo." Man n.s. 5: 27- 38.

. 1983. Modern Kongo Prophets: Religion in a Plural Society. Bloomington: Indiana

University Press. Maclean, Una. 1971. Magical Medicine: A Nigerian Case-Study. Harmondsworth: Penguin.

. 1979. "Traditional Healers and their Female Clients," pp. 225-34 in Z. A. Ademuwagun et al. (eds.) African Therpeutic Systems.

Maier, D. 1979. "Nineteenth-Century Asante Medical Practices." Comparative Studies in Society and History 21: 63-81.

Manning, Patrick. 1981. "The Enslavement of Africans: A Demographic Model." CJ4S 15/3: 499-526.

Marks, Shula, and Neil Andersson. 1983. "Industrialization, Rural Health and the 1944 National Health Services Commission in South Africa." Paper presented to the African Studies Association. Boston.

Mata, Leonardo J. 1978. The Children of Santa Maria Cauque: A Prospective Study of Health and Growth. Cambridge: The MIT Press.

Mburu, F. M. 1977. "The Duality of Traditional and Western Medicine in Africa: Mystics, Myths and Reality," pp. 158-85 in P. Singer (ed.) Traditional Healing: New Science or New Colonialism? (Essays in Critique of Medical Anthropology).

1981. "Implications of the Ideology and Implementation of Health Policy in a Developing Country." SSM 15A/1: 17-24.

McCaskie, T. C. 1981. "Anti-Witchcraft Cults in Asante: An Essay in the Social History of an African People." History in Africa 8: 125-54.

McGilvray, James C. 1975/76. "The Future of Medical Missions in World Perspective." World Mission 26: 51-55.

McGregor, I. A. 1964. "Measles and Child Mortality in the Gambia." West African Medical Journal 13: 251-56.

Mckelvey, John J., Jr. 1973. Man Against Tsetse: Struggle for Africa. Ithaca and London: Cornell University Press.

McLeod, Malcolm. 1972. "Oracles and Accusations among the Azande," pp. 158-78 in A. Singer and B. V. Street (eds.) Zande Themes. Oxford: Blackwell.

Mechanic, David. 1973. "Apartheid Medicine." Society March/April: 36-44. Meillassoux, Claude. 1975. Femmes, greniers et capitaux. Paris: Librairie Maspero. Miller, Joseph C. 1982. "The Significance of Drought, Disease and Famine in the

Agriculturally Marginal Zones of West-Central Africa." JAH 23/1: 17-61.




Monsted, Mette, and Parveen Walji. 1978. A Demographic Analysis of East Africa: A Sociological Interpretation. Uppsala: Scandinavian Institute of African Studies.

Moss, R. P. and R. J. A. R. Rathbone (eds.). 1975. The Population Factor in African Studies. London: University of London Press.

Murray, Colin. 1979. "The Work of Men, Women and the Ancestors: Social Reproduction in the Periphery of Southern Africa," pp. 337-63 in S. Wallman (ed.) Social Anthropology of Work. A. S. A. Monograph 19. London.

. 1981. Families Divided: The Impact of Migrant Labour in Lesotho. Cambridge: Cambridge University Press.

Myers, Jonny. 1981. "The Social Context of Occupational Disease: Asbestos and South Africa." IJHS 11/2: 227-45.

. 1984. "Differential Ethnic Standard for Lung Functions, or One Standard for All?" SAMJ 65, 12 May: 768-72.

, and Malcolm Steinberg. 1983. "Health and Safety Organisation: A Perspective on the Machinery and Occupational Safety Act." South African Labour Bulletin 8/8 and 9/1 (September-October): 79-95.

, N. White, and Judith E. Cornell. 1982. "Prevalence of Hypertension in Semi- Skilled Manual Workers." SAMJ 62 (4 December): 894-98.

Navarro, Vicente. 1976. Medicine Under Capitalism. New York: Prodist. Nayenga, Peter F. B. 1979. "Busoga in the Era of Catastrophes," pp. 153-78 in B. A. Ogot

(ed.) Ecology and History in East Africa, Hadith 7. Nairobi: Kenya Literature Bureau. Ndagala, D. K. 1981. "Pastoralists and Cultivators in Bagamoyo District," pp. 186-92 in

R. Chambers, R. Longhurst, and A. Pacey (eds.) Seasonal Dimensions to Rural Poverty. Ngubane, Harriet. 1977. Body and Mind in Zulu Medicine: An Ethnography of Health and

Disease in Nyuswa-Zulu Thought and Practice. London: Academic Press. . 1981. "Aspects of Clinical Practice and Traditional Organization of Indigenous

Healers in South Africa." SSM 15B/3: 361-66. Nigeria. 1930. Notes of Evidence Taken by the Commision of Inquiry to Inquire into the

Disturbances in the Calabar and Owerri Provinces, December, 1929. Lagos. Nsekela, Amon J. and Aloysius M. Nhonoli. 1976. The Development of Health Services in

Mainland Tanzania. Kampala, Nairobi, Dar es Salaam: East African Literature Bureau. Nyakatura, J. W. 1973. Anatomy of an African Kingdom: A History of Bunyoro-Kitara.

Garden City, New York: Anchor Press. Okafor, S. I. 1982. "Policy and Practice: The Case of Medical Facilities in Nigeria." SSM

16/22: 1971-77. Omoyajowo, J. Akinyele. 1982. Cherubim and Seraphim: The History of an African

Independent Church. New York, London, Lagos, Enugu: Nok Publishers International. Onchere, Simeon R., and R. Slooff. 1981. "Nutrition and Disease in Machakos District,

Kenya," pp. 41-45 in R. Chambers, R. Longhurst, and A. Pacey (eds.) Seasonal Dimensions to Rural Poverty.

Onokerhoraye, Andrew G. 1984. Social Services in Nigeria. London: Kegan Paul. Orde Browne, Major G.St.J. 1946. Labour Conditions in East Africa. Colonial Office, No.

193. London: H.M.S.O. Orubuloye, I. 0., and J. C. Caldwell. 1975. "The Impact of Public Health Services on

Mortality: A Study of Mortality Differentials in a Rural Area in Nigeria." Population Studies 29/2: 259-72.

Oyebola, D. D. 0. 1981. "Professional Associations, Ethics and Discipline among Yoruba Traditional Healers of Nigeria." SSM 15B/2: 87-92.

. 1983. "Recent Attempts in Nigeria to Utilize Traditional Healers in Health Care Delivery." Paper presented to the International African Seminar, Professionalisation of African Medicine, September. Gaborone.

Packard, Randall. 1983. "Industrialization, Rural Poverty, and Tuberculosis in Southern Africa." Paper presented to the African Studies Association, Boston.

PAHO Advisory Committee on Medical Research. 1972. Vector Control and the Recrudescence of Vector-Borne Diseases. Proceedings of a Symposium held during the




Tenth Meeting of the PAHO Advisory Committee on Medical Research, 15 June 1971. Scientific Publication No. 238. Washington, D.C.

Parkin, David J. 1968. "Medicines and Men of Influence." Man 3: 424-39. Parsons, Talcott. 1947. "Introduction," pp. 3-86 in Max Weber, The Theory of Social and

Economic Organization, edited by Talcott Parsons. New York: The Free Press of Glencoe.

1951. The Social System. Glencoe, Illinois: The Free Press. 1968. "Professions," pp. 536-47 in International Encyclopedia of the Social

Sciences, Vol. 12. The Macmillan Company and The Free Press. Patterson, K. David. 1975. "The Vanishing Mpongwe: European Contact and Demographic

Change in the Gabon River." JAH 16/2: 217-38. 1977. "The Impact of Modern Medicine on Population Growth in Twentieth-

Century Ghana: A Tentative Assessment," pp. 437-52 in Edinburgh 1977. ~ . 1979. "Health in Urban Ghana: the case of Accra." SSM 13B/4: 251-68.

1981. Health in Colonial Ghana: Disease, Medicine, and Socio-Economic Change, 1900-1955. Waltham, Mass.: Crossroads Press.

- , and Gerald W. Hartwig. 1978. "The Disease Factor: An Introductory Overview," pp. 3-24 in G. W. Hartwig and K. D. Patterson Disease in African History: An Introductory Survey and Studies.

- , and Gerald F. Pyle. 1983. "The Diffusion of Influenza in Sub-Saharan Africa during the 1918-1919 Pandemic." SSM 17: 1299-1307.

Patton, Adell, Jr. 1980. "E. Mayfield Boyle: 1902 Howard University Medical School Graduate's Challenge to British Medical Policy in West Africa." Private paper.

. 1982. "Howard University and Meharry Medical Schools in the Training of African Physicians, 1868-1978," pp. 142-62 in Joseph E. Harris (ed.) Global Dimensions of the African Diaspora. Washington, D.C.: Howard University Press.

Pearce, Tola Olu. 1983. "Professional Interests and the Creation of Medical Knowledge: A View on the Competition between Western Orthodox and Indigenous forms of Medical Information in Nigeria." Paper presented to the International African Seminar on the Professionalization of African Medicine, Gaborone.

Peel, J. D. Y. 1968. Aladura: A Religious Movement among the Yoruba. London: Oxford University Press for the International African Institute.

Perrot, Claude-Helene. 1981. "Traditions orales et demographie historique: a propos de la population du Ndenye au XVIIe et XIXe siecles," pp. 433-55 in University of Edinburgh Centre of African Studies African Historical Demography Vol II.

Pillsbury, Barbara. 1982. "Policy Evaluation Perspectives on Traditional Health Practitioners in National Health Care Systems." SSM 16/21: 1825-34.

Piven, Frances Fox, and Richard A. Cloward. 1971. Regulating the Poor: The Functions of Public Welfare. New York: Vintage Books.

Pool, Ian D. 1977. "A Framework for the Analysis of West African Historical Demography," pp. 45-61 in University of Edinburgh, Centre of African Studies African Historical Demography.

Popkin, B. M. and F. S. Solon. 1975. "Income, Time, the Working Mother and Child Nutriture." Journal of Tropical Paediatrics and Environmental Child Health 22/4: 156- 67.

Press, I. 1980. "Problems in the Definition and Classification of Medical Systems." SSM 14B: 45-57.

Prins, Gwyn. 1979. "Disease at the Crossroads: Towards a History of Therapeutics in Bulozi since 1876." SSM 13B/4: 285-316.

Pro Mundi Vita. 1967. "Medical Activities by the Church in Africa." Bulletin 21. Brussels. Ranger, T. 0. 1967. Revolt in Southern Rhodesia 1896-7. London: Heinemann.

1981. "Godly Medicine: The Ambiguities of Medical Mission in Southeast Tanzania, 1900-1945." SSM 15B/3: 261-77.

Raphael, Dana (ed.). 1980. Breastfeeding and Food Policy in a Hungry World. New York: Academic Press.




Raynault, Claude. 1976. "Transformation du systeme de production et inegalite economique: le cas d'une village haoussa (Niger)." Canadian Journal of African Studies 10: 279-306.

. 1977. "Circulation monetaire et evolution des structures socio-economiques chez les haoussa du Niger." Africa 47/2: 160-71.

Reboul, Claude. 1982. Barrages contre le developpement? Contribution a l'etude des projets d'amenagement de la vallee du fleuve Senegal. Institut National de la Recherche Agronomique. Serie Economie et Sociologie Rurales.

Redmayne, Alison. 1970. "Chikanga: An African Diviner with an International Reputation," pp. 103-28 in Mary Douglas (ed.) Witchcraft Confessions and Accusations. ASA Monographs No. 9. London: Tavistock Press.

Retel-Laurentin, Anne. 1969. Oracles et Ordalies chez les Nzakara. Paris and the Hague: Mouton.

1974a. Sorcellerie et ordalies: l'epreuve du poison en Afrique Noire, essai sur le concept de negritude. Paris: editions anthropos.

1974b. Infecondite en Afrique Noire: Maladies et consequences sociales. Paris: Masson et Cie.

1979a. Cause de l'infecondite dans la Volta Noire. Institut national d'etudes demographiques avec le concours du C.N.R.S. Travaux et Documents. Cahier no. 87. Paris: Presses Universitaires de France.

. 1979b. Un pays a la derive: Une societe en regression demographique. Les Nzakara de l'est centrafricain. Paris: Jean-Pierre Delarge.

Richards, Audrey. 1939. Land Labour and Diet in Northern Rhodesia. Oxford: Oxford University Press.

Roberts, J. M. D. 1974. "Malaria," pp. 305-17 in L. C. Vogel, A. S. Muller, R. S. Odingo, Z. Onyango, and A. deGeuys (eds.), Health and Disease in Kenya. Nairobi: East African Literature Bureau.

Rogers, Barbara. 1980. The Domestication of Women: Discrimination in Developing Societies. London: Kegan Paul.

Romaniuk, Anatole. 1967. La Fecondite des populations congolaises. Paris and The Hague: Mouton.

1980. "Increase in Natural Fertility during the Early Stages of Modernization: Evidence from an African Case Study, Zaire." Population Studies 34: 293-310.

Rowland, M. G. M., et al. 1981. "Seasonality and the Growth of Infants in a Gambian Village," pp. 164-75 in R. Chambers, R. Longhurst, and A. Pacey (eds.) Seasonal Dimensions to Rural Poverty.

Ruesink, William G,. 1978. "Status of the Systems Approach to Pest Management." Annual Review of Entomology 21: 27-44.

Sabben-Clare, E. E., D. J. Bradley, and K. Kirkwood (eds.). 1980. Health in Tropical Africa during the Colonial Period, Based on the proceedings of a symposium held at New College, Oxford, 21-23 March 1977. Oxford: Clarendon Press.

Samuelson, Paul A. 1970. Economics. Eighth ed. New York: McGraw Hill. Sanders, David. 1982. "Nutrition and the Use of Food as a Weapon in Zimbabwe and

Southern Africa." IJHS 12/2: 201-13. Sankale, M. 1969. Medicins et action sanitaire en Afrique noire. Paris: Presence Africaine. Sargent, Carolyn Fishel. 1982. The Cultural Context of Therapeutic Choice: Obstetrical Care

Decisions among the Bariba of Benin. Dordrecht: Reidel. Savage, Michael. 1979. "The Political Economy of Health in South Africa," pp. 140-160 in

Westcott and Wilson 1979. Schoenmaeckers, R., et al. 1981. "The Child-Spacing Tradition and the Post-Partum Taboo

in Tropical Africa: Anthropological Evidence," pp. 25-71 in J. Page and R. Lesthaeghe (eds.) Child-Spacing in Tropical Africa: Traditions and Change. London: Academic Press.

Schoffeleers, J. M. (ed.). 1978. Guardians of the Land: Essays on Central African Territorial Cults. Gwelo, Zimbabwe: Mambo Press.




Schofield, Susan. 1974. "Seasonal Factors." Journal of Development Studies 11/1: 22-40. Schram, Ralph. 1971. A History of the Nigerian Health Services. Ibadan: Ibadan University

Press. Scudder, Thayer. 1973. "The Human Ecology of Big Projects: River Basin Development

and Resettlement." Annual Review of Anthropology 2: 45-61. Shweder, Richard A. 1981. "Rationality Goes Without Saying." Culture, Medicine and

Psychiatry 5/4: 348-58. Simkins, Charles. 1979. "The Spatial Distribution of Mortality and its Relation to

Socioeconomic Indicators in South Africa in 1970," pp. 95-104 in Westcott and Wilson Economics of Health in South Africa Volume I: Perspectives on the Health System.

Sindzingre, Nicole, and Andras Zempleni. 1981. "Modeles et pragmatique, activation et repetition: reflexions sur la causalite de la maladie chez les Senoufo de Cote d'Ivoire." SSM 15B/3: 279-93.

Singer, Philip (ed.). 1977. Traditional Healing: New Science or New Colonialism? (Essays in Critique of Medical Anthropology). Owerri, New York, London: Conch Magazine Ltd. (Publishers).

Soforowa, A. 1982. Medicinal Plants and Traditional Medicine in Africa. Chichester: Wiley. Spring, Anita. 1978. "Epidemiology of Spirit Possession among the Luvale of Zambia," pp.

165-90 in J. Hoch-Smith and A. Spring (eds.) Women in Ritual and Symbolic Roles. New York and London: Plenum Press.

. 1980. "Traditional and Biomedical Health Care Systems in Northwest Zambia: A Case Study of the Luvale," pp. 57-79 in Ulin and Segall (eds.) Traditional Health Care Delivery in Contemporary Africa.

Squires, H. C. 1958. The Sudan Medical Service: an Experiment in Social Medicine. London: William Heinemann.

Starr, Paul. 1982. The Social Transformation of American Medicine. New York: Basic Books.

Stepan, Jan. 1983. "Patterns of Legislation Concerning Traditional Medicine," pp. 290-313 in R. B. Bannerman and Ch'en Wen Chieh (eds.) Traditional Medicine and Health Care Coverage. Geneva: WHO.

Stockard, J. L. 1978. "Economic Justification for Schistosomiasis Control," pp. 3-11 in A. Abdallah (ed.) Proceedings of the International Conference on Schistosomiasis, Cairo, October 18-25, 1975. Vol. I.

Swanson, Maynard W. 1977. "The Sanitation Syndrome: Bubonic Plague and Urban Native Policy in the Cape Colony 1900-1909." JAH 18/3: 387-410.

Swantz, Marja-Liisa. 1982. "Human Inquiry-New Ways of Doing Research with People." Paper presented to the Annual Meeting of the British Association for the Advancement of Science. 6-10 September. Liverpool.

Swindell, Kenneth. 1981. "Domestic Production, Labour Mobility and Population Change in West Africa, 1900-1980," pp. 655-90 in University of Edinburgh, Centre of African Studies African Historical Demography, Vol. II.

Tabutin, Dominique. 1984. "La fecondite et la mortalite dans les recensements africains des 25 dernieres annees." Population 39/2: 295-312.

Taussig, Michael T. 1980. "Reification and the Consciousness of the Patient." SSM 14B/1: 3-13.

Thomas, G. C. 1981. "The Social Background of Childhood Nutrition in the Ciskei." SSM 15A: 551-55.

Thomas, Jacqueline M. 1963. Les Ngbaka de la Lobaye: Le depeuplement rural chez une population forestiere de la Republique Centrafricaine. Paris and The Hague: Mouton & Co.

Thornton, John. 1977. "Demography and History in the Kingdom of Kongo." JAH 18/4: 507-30.

Tobias, Phillip V. 1975. "Anthropometry among Disadvantaged People: Studies in Southern Africa," pp. 287-305 in F. S. Watts, F. E. Johnson, and G. W. Lasker (eds.) Biosocial Interrelations in Population Adaptation. The Hague: Mouton.




Tompkins, Andrew. 1981. "Nutritional Status and Severity of Diarrhoea among Pre-School Children in Rural Nigeria." Lancet 8225: 860-62.

Turner, Victor W. 1968. The Drums of Affliction: A Study of Religious Processes among the Ndembu of Zambia. Oxford: Clarendon Press and the International African Institute.

1. 975. Revelation and Divination in Ndembu Ritual. Ithaca: Cornell University Press.

Turshen, Meredith. 1975. "The Political Economy of Health: With a Case Study of Tanzania." Ph.D. dissertation, University of Sussex.

. 1977. "The Impact of Colonialism on Health and Health Services in Tanzania." IJHS. 7/1: 7-35. -. 1984. The Political Ecology of Disease in Tanzania. New Brunswick, New Jersey: Rutgers University Press.

Twumasi, Patrick A. 1975. Medical Systems in Ghana: A Study in Medical Sociology. Tema: Ghana Publishing.

Ulin, Priscilla R. 1979. "The Traditional Healer of Botswana in a Changing Society." pp. 243-46 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

, and Marshall H. Segall (eds.). 1980. Traditional Health Care Delivery in Contemporary Africa. Foreign and Comparative Studies/African Series XXXV. Syracuse: Maxwell School.

Unschuld, Paul U. 1975. "Medico-Cultural Conflicts in Asian Settings: An Explanatory Theory." SSM 9/6: 303-12.

1979. Medical Ethics in Imperial China. A Study in Historical Anthropology. Berkeley: University of California Press.

Unterhalter, Beryl. 1982. "Inequalities in Health and Disease: The Case of Mortality Rates for the City of Johannesburg, South Africa, 1910-1979." IJHS 12/4: 612-36.

Vail, Leroy. 1977. "Ecology and History: the Example of Eastern Zambia." Journal of Southern African Studies. 3: 129-55.

Vallin, J. 1976. "La mortalite infantile dans le monde. Evolution depuis 1950." Population 31/4-5: 801-38.

Van Binsbergen, Wim M.J. 1978. "Explorations into the History and Sociology of Territorial Cults in Zambia," pp. 47-88 in J. M. Schoffeleers (ed.) Guardians of the Land: Essays on Central African Territorial Cults. Gwel., Zimbabwe: Mambo Press.

1981. Religious Change in Zambia: Exploratory Studies. London and Boston: Kegan Paul International.

Van Onselen, Charles. 1976. Chibaro. London: Pluto Press. Vansina, Jan. 1972. The Tio Kingdom of the Middle Congo, 1880-1892. London: Oxford

University Press for the International African Institute. Von Troil, Margaretha. 1982. Village Participation in the Healthworkers' Training

Programme. Helsinki: University of Helsinki and Ministry of Health, Tanzania. Waite, Gloria Martha. 1981. "The Indigenous Medical System in East-Central African

History." Ph.D. dissertation in History, U.C.L.A. Walt, Gillian, and Angela Melamed (eds.). 1983. Mozambique: Towards a People's Health

Service. London: Zed Books. Ware, Helen. 1977. "Women's Work and Fertility in Africa." pp. 1-34 in The Fertility of

Working Women, edited by S. Kupinsky. New York: Praeger. 1978. Population and Development in Africa South of the Sahara: A Review of the

Literature, 1970-1978. International Review Group of Social Science Research on Population and Development, Appendix 7A. Mexico City: IRG, El Colegio de Mexico.

Warren, Dennis M. 1974. "Disease, Medicine, and Religion among the Techiman-Bono of Ghana: A Study in Culture Change." Ph.D. dissertation in Anthropology, Indiana University.

. 1979. "The Interpretation of Change in a Ghanaian Ethnomedical Study," pp. 247-50 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

-~ , G. Steven Bova, Mary Ann Tregoning, and Mark Kliewer. 1982. "Ghanaian National Policy Towards Indigenous Healers." SSM 16,21: 1873-81.




Watt, John Mitchell, and Maria Gerdina Breyer-Brandwijk. 1962. The Medicinal and Poisonous Plants of Southern and Eastern Africa. Edinburgh and London: E. & S. Livingstone Ltd.

Watts, Michael. 1983. Silent Violence: Food, Famine & Peasantry in Northern Nigeria. Berkeley: University of California Press.

Weisbrod, Burton A., Ralph L. Andreano, et al. 1973. Disease and Economic Development: The Impact of Parasitic Diseases in St. Lucia. Madison: University of Wisconsin Press.

Werbner, R. P. (ed.) 1977. Regional Cults. Anthropological Studies Association of the Commonwealth, Monograph No. 16. London, New York, San Francisco: Academic Press.

Westcott, Gill. 1979. "Indicators of Health Status," pp. 107-115 in Westcott and Wilson 1979.

, and R. A. P. Stott. 1977. "The Extent and Causes of Malnutrition in Children in the Tsolo District of Transkei." SAMJ 52: 963-68.

- , and Francis Wilson. 1979. Economics of Health in South Africa. Volume 1: Perspectives on the Health System. Johannesburg: Ravan Press for the South African Labour and Development Research Unit.

. 1980. Economics of Health in South Africa. Volume 2: Hunger, Work and Health. Johannesburg: Ravan Press for the South African Labour and Development Research Unit.

White, Gilbert F., David J. Bradley, and Anne U. White. 1972. Drawers of Water: Domestic Water Use in East Africa. Chicago and London: University of Chicago Press.

White, Neil. 1980. "The Nutritional Status of Children in Crossroads and Nqutu," pp. 1-5 in G. Westcott and F. Wilson Economics of Health in South Africa. Volume 2: Hunger, Work and Health.

WHO. 1976. African Traditional Medicine. Regional Office for Africa. Technical Report Series, No. 1. Brazzaville.

WHO. 1978. The Promotion and Development of Traditional Medicine. Technical Report Series, No. 622.

WHO. 1983. Apartheid and Health. Geneva. WHO Committee on Insecticides. 1975. Ecology and Control of Vectors in Public Health.

Twenty-first Report of the WHO Expert Committee on Insecticides. Technical Report Series No. 561. Geneva.

WHO Expert Committee on Insecticides. 1976. Resistance of Vectors and Reservoirs of Disease to Pesticides. Twenty-Second Report of the WHO Expert Committee on Insecticides. Technical Reports Series, No. 585. Geneva.

Whyte, Susan Reynolds. 1982. "Penicillin, Battery Acid and Sacrifice: Cures and Causes in Nyole Medicine." SSM 16/23: 2055-64.

Willis, Roy G. 1968. "Changes in Mystical Concepts and Practices among the Fipa." Ethnology 7: 139-57.

Wilks. Ivor. 1975. Asante in the Nineteenth Century: The Structure and Evolution of a Political Order. London: Cambridge University Press.

Wilson, Bryan (ed.). 1970. Rationality. Oxford: Basil Blackwell. Wisner, Ben. 1983. "African Hunger: 20 Years' Attempt at Explanation." Paper presented

to the African Studies Association. Boston. Wolpe, Harold. 1972. "Capitalism and Cheap Labour-Power in South Africa: From

Segregation to Apartheid." Economy and Society 1: 425-56. Worthington, E. Barton. 1977. Arid Lands Irrigation: Environmental Problems and Effects.

Oxford: Pergamon Press. Wrigley, C. C. 1979. "Population in African History." JAH 20,1: 127-131. Wyndham, C. H., and L. M. Irwig. 1979. "A Comparison of the Mortality Rates of Various

Population Groups in the Republic of South Africa." SAMJ 55/20: 796-802. Yoder, P. Stanley (ed.). 1982. African Health and Healing Systems: Proceedings of a

Symposium. Los Angeles: Crossroads Press.




Young, Allan. 1979. "The Practical Logic of Amhara Traditional Medicine," pp. 132-37 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

. 1980. "The Discourse on Stress and the Reproduction of Conventional Knowledge." SSM 14B: 133-146.

. 1981. "When Rational Men Fall Sick: An Inquiry into some Assumptions made by Medical Anthropologists." Culture, Medicine and Psychiatry. 5/4: 317-35.

. 1982a. "The Amhara Medical System," pp. 21-41 in P. S. Yoder (ed.) African Health and Healing Systems: Proceedings of a Symposium. Los Angeles: Crossroads Press.

*. 1982b. "The Anthropologies of Illness and Sickness." Annual Review of Anthropology 11: 257-85.

Zeller, Diane Leinwand. 1971. "The Establishment of Western Medicine in Buganda." Ph.D. dissertation in Political Science, Columbia University.

. 1979. "Traditional and Western Medicine in Buganda: Co-Existence and Complement," pp. 251-56 in Z. A. Ademuwagun et al. (eds.) African Therapeutic Systems.

