3
Core Features in Ngoma Therapy

A straightforward formulation of the subject of this work, based on material presented thus far, would be something like the following: Just as persons or social forces around the sufferer are involved in the cause of affliction (as understood in the proto-Bantu cognate dòg ), so others may help in the diagnosis, decision making of health seeking, and continuing support of the sufferer to achieve well-being. As we have seen, these as well as other notions are embedded in the vocabulary of Central and Southern African languages and constitute a classic institutional form of the quest for therapy and wholeness.

And yet, this formulation does not cover all aspects of ritual healing in Central and Southern Africa. There are dimensions that are not so readily captured in the cultural vocabulary, or that are missing from that vocabulary. For example, I am not aware of a consistent vocabulary for the choreography or spatial layout of healing rituals, although they share a striking consistency of pattern across the ngoma area.

The social sciences, since their inception, have grappled with the relationship of the verbalized self-conscious model of a culture to the analytical account of behavioral and structural features and historical patterns observed by the scholar. On the one hand, there is the phenomenologist's insistence on studying only that which is consciously and verbally identified. On the other, there is the challenge to bring together analytical theories and interpretations to provide a convincing basis for those unnamed structures, contradictions, and historical

changes that identify and clarify a body of disparate information. As Clifford Geertz notes, the ethnologist's task includes a good deal of persuasion in depicting a situation so that the reader finds it uniquely realistic (Geertz 1988:4–5).

Twenty years ago I asked what were the most characteristic features of Central African healing (Janzen 1969). Then I was not concerned with verbal categories, and I did not perceive the widespread existence of historical patterns. Part of the answer to the question of characteristic features has been provided in establishing the widespread use of the process I later called the "quest for therapy," which was spelled out in a book by that title. In Kongo society it appeared that the process of being afflicted and seeking relief for that affliction was driven by a compelling worldview issue, namely, the question of whether or not the affliction is merely matter of fact, or whether "there is something else going on," that is, other persons, spirits, ancestors, or the social setting itself, in the causation of the case. As soon as this determination was made, in cases I studied, significant others around the sufferer became involved in the search for an appropriate solution. These "therapy managers began the process [of] individuals around the sufferer assisting in offering support and diagnostic affirmation as the case came before numerous types of care, be it herbalists, biomedical hospitals and clinics, diviners, and assorted social therapies" (Janzen 1978a ). Seven years later Marc Augé identified a similar process as "therapeutic itinerary" (1985). Many other scholars have reported comparable findings in their work around the continent.

This picture of Central African therapy seeking is still valid, although it is a limited one. Also, the exercise of creating a generalized, synthesized model of a culturally specific institution has its perils. One of these perils is that of using too limited an empirical or ethnographic basis of information, which might lead to a dull, stereotyped, and possibly erroneous depiction of the institution. Another danger is to abstract the core features based on a misunderstanding of the dynamics of the institution. A synthetic picture of an institution is correct not because it reflects the statistical averages of all practices but because it explains the underlying logic. And this may not correspond to any particular local tradition.

This chapter and the next present such a synthetic model of ngoma as a ritual therapeutic institution in terms of core features that include: (1) sickness and therapeutic initiation as a phased rite of passage; (2) identifying the causes of misfortune; (3) associating nosology with

"spirit fields"; (4) the "course through the white," of sickness and transition; (5) a sacrifice that sets in motion a circuit of exchanges; (6) the power of the wounded healer, together with fellow sufferers, that is, transforming sufferer into healer. All of these come together in a final feature presented in the next chapter, the core ritual, "doing ngoma."

Sickness And Therapeutic Initiation As A Phased Rite Of Passage

The first of the core features of the cult of affliction is the choreography of events over time. Throughout the region where ngoma affliction institutions are found, the process of sickness, labeling, healing, searching for answers, becoming well, and emerging as a healer is framed by rites that define the entry into, and exit from, the position of the ngoma sufferer-novice. These formal features and the spatial-temporal structure that results from them are distinguished here from the qualitative transformation of the individual (or group) as identified in another core feature, "the course through the white."

In the historic Lemba cult of the Congo coast and inland along the trade routes, this temporal framing was evident in all regional variants (Janzen 1982). To cite a specific example, after an nganga's initial identification of the sufferers' (in this case, a couple) condition as being Lemba-related, they were put in touch with a senior Lemba healer and, in the first event, purified and given the initial medicine. They were now Lemba novices under the supervision of their priest-healer. After sometimes years of counseling, of dream analysis, of song preparation, and collection of the funds for their final event, they were featured in a "graduation" event, after which they were fully qualified Lemba officiants.

In reading, and later in a comparative fieldwork project, I found that this same structure for the framing of ritual events was widely represented in ngoma settings. The events that open, close, and punctuate the therapeutic initiation are usually of a day-night-day sequence and duration. This was true in Kongo society, in Turner's accounts of Ndembu rites of affliction, in igqira initiations in the Western Cape, in coastal Cameroon rites (in contrast to the Grassfield area in the interior), and in Haitian voodoo, which carries a strong Central African institutional pattern. This pattern is commonly aligned with the preparation for meeting ancestors or spirits and bringing the novice into communication with them. Frequently the sequence of events also spatially

reflects this with a move in the ritual choreography from a profane to a sacred or auspicious place, moves that are announced with transition songs to move the sufferer-novice through an intermediary space (see fig. 8).

These opening, closing, and punctuating events of the ngoma initiation are marked as well by the preparation of medicines, the utilization of color-coded stages and ointments spread on the novice, and by the sacrifice of an animal that is ritually identified with the novice and is then slaughtered, cooked, and eaten as a common meal.

Despite the pervasive presence of the foregoing pattern that structures the ngoma rites, there does not seem to be a set of common verbal cognates that relate to this structure. The events that open, close, and punctuate the process are variously called nkembo (celebration), or mpandulu (initiation to or composition of an nkisi ) in Kongo; nthlombe (celebration, feast) in Nguni languages; ngoma in East Africa. These terms are all used fairly generally to speak of ceremonies, initiations, or rites of all kinds.

As in other widespread cultural patterns, these elements of Central and Southern African therapeutic initiation have been explained by authors utilizing several ethnological principles that go beyond the culturally particular and descriptive. Two major approaches that may be outlined here are those of the "rite of passage" and the "shamanic career." As with all ethnological explanations that are more general than descriptive, these have some value but they also remain problematic.

The elements of initiation in ngoma-type cults of affliction in Ndembu society were explained by Victor Turner as examples of Van Gennepian "rites of passage." They were opened by a rite of separation of the novice-sufferer from a prior social state. This was followed by an intermediary "liminal" or transitional state. The process culminated with a rite of reincorporation by the novice into society, as a full-fledged healer and member of the cult. Later, Max Gluckman argued for a more analytical approach to these rituals, especially in societies moving toward greater complexity and differentiation of roles (1962:1–52). Turner himself came to see the cults-of-affliction rites as more sophisticated and varied examples of ritualization in human society. However, it is useful to understand the African cult of affliction as a culturally specific case of the human rite of passage, on a level similar to that which sees the sick role and the encounter with the medical professional in the West as a "rite of passage." As in any therapeutic course, the outcome is not necessarily assured. Many are those who begin ngoma,

[Full Size]

Figure 8. 
Synthetic configuration of spatial and temporal organization of
events in ngoma-type rites. This chart is to be read as a musical score. The
"time" line moves from left to right. The "space" score also moves from
left to right, but indicates the spaces in which activity is situated through
the rites.

or who participate in it, but who never complete it. For this reason it may not be like a strict rite of passage. The progression through the ritual grid is subject to the inner progress of the novice-sufferer.

Others have identified aspects of African cults of affliction with the "shamanic career" (De Heusch 1971), although this has been held to be problematic and inappropriate by most authors (I. Lewis 1986:78-93), for reasons that will be developed at greater length in chapter 5. In Lewis's analysis of the "career" of spirit possession, a series of stages moves the relationship of the novice to the spirit from a point where it is uncontrolled and involuntary to a point where it reflects greater control, indeed, voluntary interaction with the spirit through mediumship. This "career" is also defined by the subject moving from being a patient or sufferer to mastery over the source of affliction, and becoming a healer of that condition, although, as noted above, the process may stall. Some aspects of the shamanic journey may be seen in the choreographed move from profane to sacred space in every rite (see fig. 8).

And some novices do speak of having journeyed to the bottom of the river, or into the woods or wild bush to rescue a soul, or to commune with a spirit. Thus some of the elements said to universally define shamanism are also in keeping with ngoma ritual. However, rarely is the ngoma graduate or healer involved in classic shamanic journeys following the completion of the therapeutic initiation. Some, such as Botoli Laie in Dar es Salaam, admit freely to having been introduced to ngoma practice as an apprentice.

It thus seems clear that caution must be exercised in applying universalistic ethnological explanations to a phenomenon I have argued is historically and culturally particular, with its own distinctive vocabulary and significant variation within the region where it is found: for example, the day/night/day scheduling of the events; the use of white symbolism as the lengthy inner or middle passage; the role of percussion in setting the stage for passage; the spatial choreography that moves from profane space to sacred space and back.

The Diagnosis Of Misfortune

The student of Central and Southern African collective therapies of the ngoma type is presented in many localities, including the urban centers under particular scrutiny in the present work, with a broad array of afflictions deemed appropriate for treatment. The array is in fact so broad that one can question whether the distinct diagnosis of signs, symptoms, or syndromes is at all a characterizing feature of the ngoma therapeutic mode. And yet, it is analytically important to differentiate problems brought into ngoma-type cults of affliction from other types of problems taken elsewhere.

One way to look for the distinctive arena of problems taken into ngoma healing is to note that, with some exceptions, many of the ngoma orders exist, as it were, beyond kinship. Even those such as Nkita in western Zaire and Angola, although about the reproduction of the matrilineage, are construed as being about how the lineage might transcend the internal dilemmas of factionalism, leadership problems, and ill will that are thought to occasion the sickness and death of infants and the sterility of the mothers. The proto-Bantu cognate verb dòg connotes all the sources of affliction—sorcery, witchcraft, backbiting—that result from relations within the closely knit human community. Ngoma interventions and appeals are made to get beyond the entrapment of dòg (see appendix B, section B.3). Frequently the ancestral shades and

the spirits beyond are believed to call individuals out of their self-consuming destructive tendencies. Ngoma afflictions are thus not noticeably different from non-ngoma afflictions on the surface level of signs and symptoms. The diagnostic interpretation becomes important as a way of reaching beyond the mundane for a way out of an impasse.

An important dimension of all cults of affliction is the analytical and diagnostic function of evaluating sickness and misfortune. A distinction is often drawn, in this connection, between divination, the analysis of a situation, and healing, the attempt to intervene in the situation to change it. This distinction was the basis for Turner's book title Revelation and Divination , that is, the relationship between the cult of affliction and the diagnosis of the problems it addresses. The distinction accounts for some of the diversity of affliction cult types, for where social change is intense, the need increases for cognitive clarity. Thus, in eighteenth-century coastal Kongo, during the decline of the kingdoms with the increase of the trade, including the slave trade, divination cults were in great profusion, particularly those relating to adjudication and conflict resolution. In Southern Africa today, the term ngoma is associated mainly with divination. Closer examination, however, shows that the functions of divination and network-building are complementary.

Divination, or diagnosis, thus always accompanies cults of affliction, either independently of the healing role, or as a part of the specialized techniques and paraphernalia of a particular cult. Divination must be thought of as a continuing query into the "whys," "whos," and "where-fores" begun in the family-therapy management setting but carried through by specialists with expert judgment and training, who may have had their own profound individual dilemmas, who have been recruited to a particular mode of ritual life, and who have been initiated to the spirit world. As a technique, divination may be based on a mechanistic system of signs and interpretations, such as the Southern Savanna Ngombo basket filled with symbolic objects signifying human life, the pengula bone-throwing technique of Southern African Nguni society, or recitation from the Bible or the Koran. Alternatively, and according to some observers increasingly, divination is done by direct recourse to trance, in which the diviner, as medium, speaks the words of the ancestral shade or spirit in answer to a query. Some diviners use a combination of both techniques, or a selection of hierarchically arranged types. In Swaziland, master diviners today train novices in the arts of mechanistic bone-throwing divination as well as the mastery

of trance-divination. In any case, these divined diagnoses, representing a type of analysis or interpretation of daily life, offer the basis for the more synthetic, ritualized follow-through of the cults of affliction. However, there is a contradiction or tension between the specificity and concreteness of the problems channeled into divination and treatment, and the overall hypothesis of spirit causation that brings an individual into a cult of affliction.

The work of Zairian researchers Mabiala (1982) and Byamungu (1982) illustrates this issue. Signs and symptoms (illnesses) accompanying recruitment to Mpombo and Badju are said to include dizziness with hallucinations; throbbing headaches; lack of mental presence (asthenie); skin rashes (algie); lack of appetite; difficulty in breathing; heartburn with anxiety; rapid or arrhythmic heartbeat; fever with chills; sexual impotence; dreams of struggles or of being followed by threatening animals; weight loss or excessive weight, especially if accompanied by spirit visitations; and a variety of gynecological and obstetrical difficulties. These "modes of affliction" are characteristic of most of the other collective therapeutic rites—les grands rites —in Kinshasa.

Byamungu, working in Bukavu, in the mountainous Kivu region of eastern Zaire, is more explicit about the generalized or random character of signs, symptoms, and syndromes in the mode of affliction associated with the five therapeutic rites of this trading and administrative city of about seventy-five thousand people. In the Kakozi rite, of Bashi and Balega ethnic origin, early stages of sickness are manifested by all types of behavioral and physiological problems, reflecting something of the diversity of Kakozi spirits behind these problems. In the Enaama or Mana rites of Bashi origin (a Nilotic, Rwandais group), the characteristic afflictions include behavioral troubles, alienation, and physical sickness such as weight loss. Here again the distinguishing feature of recruitment is not a particular symptom or sign, but affliction by the Enaama nature spirits of Bashi or Rwandais origin. In the Mitumba rite, of Bembe and Balega origin, behavioral and physical troubles of whatever sort may indicate this mode of affliction. The distinguishing feature is, however, evidence in dreams or hallucinations of the presence of "white" or "European" spirits. The same type of symptom-sign randomness exists in the Mulangoyi rite of Lega, Zimba, and Songe origin, in which the major spirits are nature- or river-related, and in the Nyamulemule rite, in which the spirits are of Luba origin, appearing among the Batembo and Bashi peoples. Byamungu emphasizes that al

though signs and symptoms accompanying those afflicted may be identified and even treated with biomedical methods and medicines, the salient point that brings these sufferers to diagnostic entry into the ritual therapies is not so much the sickness but the identification of the spirit force behind the sickness.

Our analytical approach would seem, however, to call for close study of symptoms and signs of affliction or behavioral change in relation to the circumstances of the personal life of the afflicted, on the one hand, and the cultural logic that steers the course of therapy in the direction of identifying the spirits, on the other. It may, of course, be very difficult to correlate the explicit and implicit conditions of distress with the diagnoses and therapies in the Central African setting, where the initial diagnostic work is done apart from therapeutic initiation to a group setting. In Central African cities such as Kinshasa and Bukavu the collective rites are varied by regional and ethnic origin more than by a specialized division of labor. The diagnostic work is often done by kin groups and diviners working separately from the healers. Thus the novices are already identified as being in touch with the appropriate spirits by the time they make their appearance in the ritual communities.

In the Southern African context, that is, in Nguni-speaking societies, the relationship between diagnosis and therapy and the course from diagnosis to therapy may be easier to follow. This is so because here there appears to be a greater concentration of ngoma therapies within a single institutional context, and thus a fusion of divination-diagnosis and therapeutics. In the Nguni context it is easier to see the type of affliction or problem singled out for divination from among the wide array of common afflictions in a populace, as well as those cases further singled out for therapeutic initiation into the ngoma order. Illustrations from Swaziland and the Western Cape, given in chapter 1, demonstrate this process in an ethnically diverse and urban clientele.

The cases seen by Ida Mabuza of Swaziland are said to be both "African" and "non-African"; she has both African and non-African clients, the latter mostly Afrikaaners from South Africa. For Africans, she says, the most common problem presents in vague pains and anxieties and is explained by harm or sorcery (umbelelo to meqo ) resulting from interpersonal tension. This affects both men and women. Next most prevalent is amakubalo , affecting mostly men, in the form of harm or illness resulting from violated social or moral precepts resulting from illicit sex with a protected married woman. Many young people come wishing to divine their fates, desirous of good fortune in job seeking,

examinations, or love. She diagnoses their problems and offers advice accordingly.

Whites also come for these types of problems. Their main concern, however, is fear of poverty. They also come for help in seeking promotions and other job-related matters. Finally, they come with illnesses improperly diagnosed in the hospital or not effectively treated. Stress is a common complaint whose root cause can often be traced, she noted, to tensions or conflicts with domestic workers or subordinates who have retaliated.

The cases that are divined with the more powerful—and expensive—femba (possession) mediumistic method do not vary from those already presented, except perhaps that they are more chronic and less specific, and Western medicine has been unable to produce a perceptible improvement in health.

The foregoing problems typify those brought to ngoma diviners and therapists. Perhaps the most striking feature in this material is the random, vague, and ambiguous character of the connection of sign/ symptom/syndrome to the therapy, both in terms of physiopathologies and psychopathologies. This seems to be the case especially in contemporary urban settings. Nevertheless, in these urban settings there occur quite precise diagnostic readings of life situations that lead to therapy or entry into one or another ngoma therapeutic group. It is not the particular sign, symptom, or syndrome that predicts the therapeutic course. Rather, it is the diagnostic and divinatory judgment that sends the case beyond the confines of a strictly relational interpretation—as found in the concept dòg —to the interpretation that the subject is called by spirits aligned with the ngoma orders.

Nosology And Spirit Fields

Who or what exactly are these spirits? The worldview that inspires cults of affliction includes, as an axiom, the idea that ancestral shades and spirits, ultimately expressions of the power of God, may influence or intervene in human affairs. They are held responsible for visiting their sentiments and forces upon humans through sickness and misfortune. Who they are, why they come, and what to do about them is what cults of affliction are all about.

Scholarship has gone well beyond merely describing accounts of African spirits, to studying their configurations and relationships in society, in geographical space, and over time (Werbner 1977). Our aim here

is to identify some of the common features in spirit constellations across the ngoma region and to grasp the meaning of some of the variations. The spirits or shades may be either direct, identifiable lineal ancestors, or more generic shades. They may include more distant nature spirits, hero spirits, or alien spirits that affect human events in many ways. They may be benign or malign; very generalized or particular; male or female; African or foreign. The lineal ancestors, who are generally beneficent, although sometimes stern, are contrasted to wild malefic spirits or enemy hosts with sinister and strange characteristics. There is a spirit "geography" or "ecology" that widely contrasts spirits of the land from those of the water. The well-known African color triad—red, white, black—often is invoked to characterize the spirits as well. The strings of colored beads or cloth worn around the shoulder and waist designate spirits with which the novice or practitioner has a working relationship. Old as well as new knowledge tends to be related to the shade and spirit forces, as events are interpreted and adversities dealt with. Sometimes the proto-Bantu term zimu or dímu is used to name ngoma spirits, as in the Venda ngoma dza vadzimu , but a range of other names or terms is used as well (see appendix B, section B.16).

We begin this review somewhat arbitrarily with an Nguni group in southern Mozambique, the Kalanga, studied by David Webster (1982). For them the "Ndau," "Ngudi," "Chikwembe," and "Majuta" are the four main groups of spirits. The Ndau, or Vandau, are considered the original ancestors of the Kalanga (a South Shona or Thonga group) and the most powerful spirit group, with a direct interest in the affairs of the living. Because of Henri Junod's work on the Thonga in the '30s, the Vandau have entered the anthropological literature as one of the major examples of "true shamanism" in Africa (De Heusch 1971:273–276). They are mentioned popularly in Tanzania as having inspired the N'anga cult, thus reflecting a thread of Nguni, or Ngoni, influence of early nineteenth-century conquest fame (Zaretsky and Shambaugh 1978). The Ngudi, associated with local affairs, are spirits inflicting traumatic disease who need to be placated to avert human disaster when they become involved. The Chikwembe are the ancestors of isan-goma diviners. The Majuta are Arab spirits.

The distinction between lineal and alien spirits seen here is widespread. In nearby Swaziland the Emenlozi (literally, those one dreams about) are the personal ancestors and are often associated with white symbolism such as clay, white beads, white cloth, or with "mud," that is, the boundaries of water and earth. The Emenzawe and Benguni

spirits are those whom Swazi warriors killed in previous wars. The Emenzawe prefer the red beads, the Benguni the white. The Emenzawe, Benguni, and Dinzunzu possess diviners. The Dinzunzu or Tinzunzu are those spirits of the water who died of drowning; they also are associated with white beads. These fields articulate Swazi culture and consciousness.

In the Xhosa region of South Africa, and in the urban extensions of Xhosa culture, spirits are commonly identified as being those pertaining to the family or clan, those of the water, and those of the land or of the forest. Clan ancestors are important to keep in touch with, but they do not inflict illness. The ngoma practitioners, the amagqira, are called (twasa) by the forest and water spirits, and are represented in igqira costumes by animal skins and colored beads, in medicines by plants and mineral ingredients, and in ngoma songs by mediatory imagery such as the crab, the horse, or birds.

In coastal Tanzania the distinctions of spirit geography always hinge on the land/water dichotomy. Like the Southern African spirit cosmologies, this one, too, identifies spirits of the trees and shrubs, that is, the forest, with the land. In the urban Dar es Salaam setting immigrant waganga (healers) from the interior, predictably, are specialized in "interior" spirits, and in the corresponding ngoma such as Manianga, N'anga, and Mbungi, whereas waganga from coastal areas relate to coastal or water spirits. The vocabulary of the Swahili coast has adopted Arabic terms masbeitani and majini to speak of spirits. The spirits of the interior carry African names such as Mchela, Matimbuna, Mbongoloni, Chenjelu, and Kimbangalugomi (related to Ngoma Mbungi), whereas those of the beach or the water carry such names as Maruhani, Subiyani, and Mzuka (related to Ngoma Msaghiro). The colors red, white, and black also occur on beads worn over the shoulder of the mganga (healer), and in costumes.

The notion of "spirit fields" has been used by a number of authors to describe the organization of African spirits. It is an analogy of the concept "social field" used by Bateson and Turner long ago. Although some hermeneutic scholars such as Lambek (1981) insist that the spirits have little to do with social categories and forces, many other scholars prefer a Durkheimian correspondence theory between religion and society, economy, historical change, and psychological states. This permits scholars at least an opening hypothesis with which to assess such phenomena as the apparent shift in emphasis from lineal ancestors to more distant and alien spirits in recent decades. Thus the decline of cer-

tain local rites and the ascendance of others, or the rise and decline in many historic examples of spirit-possession rituals and cults, may be explained in terms of historical social forces and the changes that have occurred. For example, the breakdown of specific rituals in the late nineteenth and early twentieth centuries has been explained by the opening up of social relations and the expansion of the scale of known spheres of influence (Werbner 1977). Similarly, worldview as reflected in the understanding and treating of affliction reflects, in turn, changing social forces. It explains, for example, the generalization of symptom/ sign and etiology relations in the context of the Lemba rite in western Congo over the period from the seventeenth century to the early twentieth century in the corridor of the great trade between coastal port towns of Cabinda, Loango, and Malemba, and the great market at Mpumbu near today's Kinshasa. Lemba's sufferers are said to have had a random variety of afflictions. Even the spirits behind the afflicting and therapeutic rites varied from region to region, indicative, I think, of the tremendous upheavals of the hour.

A brief review of the manner in which spirit or ancestor forces are aligned with social contextual disorders in contemporary urban settings suggests a trend toward greater reliance upon more broadly based, generalized in theme, symbolic figures, and a waning of local or lineage ancestors. Byamungu's overview of therapeutic rites in Bukavu, in eastern Zaire, shows the association of signs and symptoms to spirit nosology. In the Kakozi rite, present among Bashi (Rwandaise) and Lega inhabitants, all sorts of behavioral and physiological afflictions are attributed to "red" ancestral spirits or shades; red is also the color of the hair and clothing of the adepts. In the Enaama (also Mana) rite, found exclusively among the Bashi, a more open range of symptoms—for example, behavioral disorders, severe alienation, loss of appetite, and physical illness—is attributed to the Enaama nature spirits who frequently drive the afflicted to prolonged periods of wandering in solitude in the bush. In the Mitumba rite, among the Bemba and Lega, a similar wide variety of symptoms and signs is attributed to spirits of Europeans, that is, aliens, originally revealed or manifested in dreams and visions, and whose visitations are accompanied by loss of consciousness during possession; adepts of Mitumba spirits speak Swahili and smoke cigarettes. In the Mulangoyi rite, among the Lega, Zinga, and Songe residents of Bukavu, a variety of symptoms and signs is attributed to nature or water spirits, who in the possessed mediums present themselves in the Kisonge language, painted "white," eating earthworms,

toads, and so on. Finally, in the Nyamulemule rite, among the Batembo and Bashi of the city, Swahili-speaking spirits of Baluba ancestral figures—whose songs are also sung in Lega and Tembo—afflict individuals who are recruited to therapeutic seances.

The Bukavu setting shows an opening up of the "spirit field" to a wide variety of nature, ancestral, and alien spirits in a setting of ethnic pluralism and an expanded scale of social relations characteristic of this eastern Zaire and wider Eastern African setting.

However, the presence of spirits who represent the influences of nature (particularly rivers) and aliens (Baluba, Europeans, Swahili-speakers) in connection with behavioral troubles, alienation, and, in the case of Mitumba, loss of consciousness in visions of Europeans, introduces into the "spirit field" an attempt to deal with behavioral pathology and its contextual causes. Although one may still invoke a Durkheimian correspondence analysis between expanding scale of relations and behavioral pathologies resulting from strained role expectations and fulfillments, the more interesting issue is the possible correspondence between particular types of spirits and an indigenous analysis of psychopathology and the appropriate therapeutic response. Harriet Sibisi's work is suggestive in this regard.

Sibisi (1976) notes that Zulu sangoma, legitimated by a "call" from their direct lineal ancestors, tend to analyze possession by nature spirits and alien spirits as evidence of abnormalities. In their therapeutic interventions in these cases they strive to replace these "spirits of chaos" with a more normative spirit patronage by ancestors, to fill a role in ritual leadership. The sangoma therapist must thus identify the rehabilitative or reintegrative potential of a client before turning that client in the direction of an initiation to the healer role. Indeed, the distinction between pathological possession and ancestral call is made clear in Nguni nosology: mfufunyani possession by chaotic spirits is a sign of madness, whereas ukutwasa is a possession or call that leads to personal strength and leadership in the sangoma or igqira. In the Southern African setting the symbolism of ukutwasa may also frequently be channeled into Christian fulfillment in Zionist prophet-healing churches, or even in mainstream church roles, whereas mfufunyani cases may be taken to a range of Western psychotherapists and African healers for treatment.

Despite the logical elegance of Sibisi's interpretation of Zulu diagnostic categories, and its clear "fit" with the South African ethnographic data from Natal, the Western Cape, and probably other areas

in between, it adheres a little too closely to a sociological correspondence theory to account for all the ethnographic evidence of the entire area across which ngoma-type therapies are found. In particular, it does not seem to explain the cases in which, as in several of the Bukavu rites (Enaama and Mitumba), alien spirits are the principal symbols of therapeutic rehabilitative orders; or, closer to the basis of Sibisi's work, in Swazi ngoma, in which numerous types of alien spirits are used to inspire divination and heating. The Swazi instance is particularly challenging, since Swazi sangoma are considered part of the Nguni-speaking group and share most of the features of the religious and therapeutic tradition with the Zulu and Xhosa.

In addition to patronage by their lineal ancestor shades, the amadloti , Swazi sangoma feature patronage by the Benguni "white" spirits, who are the victims of Swazi wars with Zulu, Tsonga, and Shangani and who inspire divination with bones; the Amanzawe "red" nature spirits who inspire mediumistic divination; and both "red" and "white" Tinzunzu spirits of those who have drowned in rivers. Mediumistic work with Benguni victim spirits seems to be a recent development in Swaziland. Zulu-type sangoma diviner-healers have been in many regions replaced by "red" takoza mediums. In their more powerful forms of divining they put aside their bones and their attentive ear for very dramatic trance-possession dances of Amanzawe and Tinzunzu nature, and Benguni alien spirits.

To take Sibisi's analysis strictly, these Swazi takoza diviner-healers, as well as some of the Bukavu therapists, would be indulging in dissonant, and according to her, "charlatanistic," practice. And yet the appearance of nonancestral spirits and their mediums has become very widespread. Alternatively, to assume that therapists ignore the social forces of alienation and dislocation when they continue to attribute sickness to ancestral calls (particularly those that are considered amenable to integrative leadership roles), is to underestimate the skill of these talented individuals working often in extremely strained social settings such as in the Western Cape.

The anthropological analysis of the relationship of a "spirit field" to the social context of affliction needs one further analytical parameter beyond those of social scale (i.e., localized vs. regional or cosmopolitan) and of normalcy (i.e., role normalcy vs. situations precipitating abnormal response to role expectations) to explain the relevant variables of recourse to spirit nosology. I have in mind the place in affliction diagnosis hinging on the degree of ambiguity versus clarity in the overall

perception of a social situation or a view of reality (Bernstein in Douglas 1970).

Bernstein's major point, taken over by Douglas, is that in the absence of a clear understanding of a phenomenon, in this instance the cause of affliction, one tends to formulate names, configurations, or stereotypes to compensate for the fuzziness. Following this line of thought, an escalation toward alien or chaotic spirit forces would be used in divination or therapy to come to grips with strange and new issues in a social situation. This hypothesis applies directly to the contrasting types of ngoma spirit manifestation in Southern Africa, notably the Western Cape and Swaziland.

In the Western Cape the challenge facing ngoma diviners and therapists is not an analysis of the situation before them; that is clear enough. Families are divided, and there is general anxiety regarding gainful employment. The major challenge is that of building up a cohesive social fabric out of the fragments of families and interpersonal relations. Accordingly, diviner-therapists forge a network of links within their ranks that bring fragmented individual lives and family segments into some more coherent and supportive arrangement. In Swaziland, there is generally greater economic security and much less anxiety about residence, freedom of movement, and one's personal welfare. Swaziland has one of the highest per capita incomes in black Africa—$800 per person in 1982, near that of Gabon and Cameroon. And yet, this very sense of economic development contains the ingredients to dissolve the normative order. Wage labor, urbanization, and education have had a significant impact upon Swazi society, creating enormous opportunities for upward mobility and prestigious jobs, especially for young adults. It has become common for young women to eschew marriage for professional work and to have a child or two out of wedlock along the way. Furthermore, the economic opportunities of Swaziland have inundated the country with outsiders, such as development experts, tourists, teachers, and traders. Thus, although there is not the material insecurity of the Western Cape and the threat to one's domestic living arrangement, economic development has unleashed other threats to the established cognitive order. This may account for the more aggressive spirit field in Swaziland, in which the spirits of bone-throwing divination (the amadloti and Benguni ancestors) have been partially supplanted by the spirits of trance and possession divination (the Emanzawe and Tinzunzu).

It is evident, then, that "spirit fields" provide a set of parameters having to do with worldview, order and chaos, legitimacy, and cultural categories against which to align, and begin to deal with, personal problems.

The Course Through The White

Therapeutic attention to affliction, through ngoma, often entails elements of initiation of the afflicted into membership, resulting in the elevation of the afflicted to the status of priest or healer in the group. Whether or not this actually happens (there are many dropouts) depends on the novice's progress through early states of therapy and counseling, on the novice's or kin's means, and the extent to which the cult is controlled by an elite that restricts access to its basic resources.

Across the ngoma region, whiteness defines the special transitional status of the sufferer-novice in the course toward health. It is expressed by the use of clay or chalk, cloth, beads, rafia, and other material indicators. In Western Bantu this concept is identified by derivations of the proto-Bantu cognate pémbà (see appendix B, section B.13). In Nguni-speaking Southern Africa the term ikota is used for white, but it is not clear how widespread this term is. In any event, whiteness is the color and dominant symbol of the transitional stage that denotes purity, separateness, isolation, the liminal zone between sickness and health, the condition of being sick. Some ngoma-type therapeutic settings become involved with red and black symbolism, often denoting exposure to alien or less familiar nature spirits.

Stages that articulate phases in the "white" may vary from two to as many as eight, each of which may endure from a few days to many years. In the Western Cape, for example, the stages of this progression begin with (1) being diagnosed as twasa, possessed or called by a spirit; (2) becoming a novice (nkwetha ) following the initiation, and joining a sodality under the counseling of a senior diviner-healer; (3) moving through the "course," becoming a senior, a "five-to" (i.e., "five to midnight," almost completed), and being entrusted with aspects of ritual; (4) fully qualified, completing the course, graduating as a sangoma or in Xhosa, igqira.

Clothing and bodily paint indicate the progression through the white. Initially, the novice is fully smeared in white chalk or wears a white cloth. Gradually, over the course of time, colors replace white-

ness; the costume of a fully qualified igqira/sangoma provides a medium for the self-definition and articulation of the new person.

Initially, face and body are smeared in white during the events of initiation. Two single strands of white beads, and the bladder of a goat, are worn around the head; bead strands are also put on each wrist. In time, more strands of white beads are added. The groups of amakwetha (novices) of a diviner-healer may be seen as a uniformed group, together, dancing, singing, in counsel, in isolation. However, as their therapeutic initiation moves to its final stages and they become more self-confident, they may paint only their cheeks and eyes. When fully qualified, only the eyes remain encircled with "white." Colorful beads and other headdress and costume elements, dreamed or creatively thought out, now replace the white. At this later phase, the fur of wild animals such as puma, cheetah, leopard, lion, beaver, and others replaces the strands of goat hair. Colorful blankets replace the white sheets.

The early novitiate phase is also signaled by the ownership of a forked stirring stick, used initially to bring the ubulau (medicine) to a frothy white whenever the ancestors or spirits are called in and the novice is smeared. This stick, which is also used as a dance wand, is replaced after graduation with a colorful, beaded dance wand comparable to the widespread Nguni knobkerrie.

This progression of exterior process of clothing, body paint, and artifact parallels the inner process that the novice has gone through and of social changes that significant others have created around the novice. The course through the white is the framework of the sick role. Initially, there is a marked withdrawal of self and individuality from the social setting, although often the family that assumes responsibility for the patronage of a diviner-healer shows the support and care they have for the individual sufferer. During this phase the sufferer must acquiesce to ancestors and social others. The death of the sacrificial goat symbolizes the death of the sufferer's self, in exchange for a new life and identity. In the early counseling and therapies the sufferer-novice is passive, receiving songs from others. Then, as he or she develops skills in dreaming and handling spirits—that is, channeling the chaotic visions and dreams of the twasa (call) experience—the costume begins to show bits of color, of identity. Accompanying this outward manifestation of dreaming, of messages about the self, is the emergence of the novice's own songs, composed also from dream impulses. The initial passivity

is replaced by a statement of ego strength as the novice composes, sings, and teaches others his or her song.

At the close, as the novice becomes a " five-to, " ready for graduation, she or he should have a strong self-projection, capability in leading others in therapeutic dance and song, as well as a firmer control over his or her own life than before. The final graduation feast, marked by the sacrifice of a cow or bull and the making of a colorful dance wand, indicates the culmination of the course through the white. The next chapter will study the role of the ngoma song in this process.

The efficacy of the therapy, regardless of its specific techniques, is partially assured because all in the community feel shared affliction and support the sufferer, even though not all the community is kin. In most instances of prolonged sickness in African society, diagnosis and decision relating to the course of healing—"the quest for therapy"—are in the hands of a lay kin therapy managing group. In the cases that come into the orbit of cults of affliction, the support community broadens to become that of the cult members. The quality of support shifts from ad hoc kin aid to that of permanent involvement with such a network in the initiate-novice's life, corresponding to the long-term involvement of the individual with the affliction, as a healer-priest.

Some cults of affliction, such as Nkita among the Kongo peoples of western Zaire, are situated within lineages. Nkita responds to the unique circumstances and symptoms of children's sicknesses and barren women amid the stresses and fears of lineage segmentation. The imputation of a cluster of Nkita afflicted within a lineage segment provides the rationale and the setting for the regeneration of lineage organization; members are reaffiliated with the ancestral source of their collective authority.

Most cults of affliction, however, occur outside the kin setting, functioning as an addition to kin relations, and give the individual lifelong ties with others along the lines of the new affliction or occupation-specific community. The various Southern Savanna reproduction-enhancing ngoma—such as Mbombo (Goblet-Vanormelingen 1988); Isoma, Wubwangu, and Kula (Turner 1968); and unnamed mahamba among the Luvale (Spring 1978)—which isolate women with reproductive problems or during pregnancy and childbirth, radicalize the separation of novices from their kin. The rationale for this separation has to do with the danger of the stresses of daily life and family relations upon the fetus. The white symbolism of clothing and the

mpemba chalk applied to the face and other places in the seclusion compound suggest the liminal special quality of this role. It is liminal, also, because the normal conjugal life of the novice has been interrupted to achieve an end that will enhance that conjugal life.

The social dimension of "the white" may vary in its particulars from occurring within lineage and family to removal from it. The common core beneath this is, however, that it usually represents a contrast from the prior state and brings the individual into touch with ritual experts who rebuild the individual's identity around a solution to the problem or affliction.

Sacrifice And Exchange

Sacrifice of mammals and birds is widely present in the ngoma process of personal transformation. Particular illustrations of this are found in Lemba, where the chickens and pigs are killed at appropriate junctures of the novice's course. This sacrifice seems to occur most often at the initial entry into the course through the white and at graduation. It occurs at the close of the day/night/day event. It also occurs in the Southern African setting, as witnessed by the instances in Cape Town described in chapter 1: a goat at the time of an entry, a goat or a sheep at the time of a healer's or novice's purification, and a cow at the time of a graduation.

Why is sacrifice so important in the commemoration of ritual transition, in the moment of sickness? Victor Turner, in explaining the importance of sacrifice in ngoma rituals among the Ndembu, begins with the repetitive, or cyclical, nature of life in the society. Some of these features are due to the agrarian context of the Ndembu, who yearly plant their crops and harvest them. However, most of the drums of affliction are focused on the human life course, in which moments such as birth, adolescence, the varied conflicts and afflictions of adulthood, and death, need to be addressed. Ndembu society is involved in change, but the core of society, its values and patterns and ideals, are thought of as if they should remain constant. The rituals of affliction seek to return to that constancy, by subsuming misfortune to the permanence of the invisible spirit world. Misfortune that results from conflict creates victims. The sacrificial victim, in the abstract sense of a substitutionary victim, lets the conflict, the anger, so to speak "have its blood." To offer a sacrificial animal also purifies the universe in that it restores or regenerates the human community to its ideals. "At the moment when the wheel

has come a full circle, [sacrifice] sets the cycle going again" (Turner 1968:276).

This is basically an "atonement" view of sacrifice, in which the violence in human society can be overcome through its symbolic manifestation in the victim. This is not far from the Judeo-Christian sacrificial tradition in which lambs, or first sons, are offered for the expiation of human sins.

Although the atonement notion of sacrifice is present in ngoma sacrifice, as the account of the case study of initiation in Cape Town related in chapter 1 suggests, there is another dimension to ngoma sacrifices that justifies an exchange concept of sacrifice. Having attended a few sacrifices in Africa, including that described in chapter 1, I can attest that there is a significant ritual economic dimension at work. The "horizontal" dimension of the distribution of food and the common meal that includes the consumption of the sacrificial animal represent as much a "communion" as a religious atonement. Together they bring out a social dimension that Turner does not specifically mention, namely, the beginning or the renewal of exchange relationships between individuals and social units that are at the basis of ongoing social relations. The nature of these networks that are created and celebrated in connection with the sacrifices will be explored more fully in chapter 6. Also, in the sacrifices of sheep, goats, and cattle, the red meat protein is significant in the diet of those attending.

A communal meal following an all-night communal dance applies tremendous energies to the reconstitution of the social whole that is assembled. This leads directly to the final core feature, the belief that misfortune, adversity, and affliction may be transformed into power and wholeness. There is a strong insistence in ngoma theorizing that singing, sacrifice, and communing turn life around and literally bring life out of death.

The Sufferer Becomes The Healer

The lyrics of ngoma songs echo this core feature, as in this song from ancient Lemba on the Kongo coast: "That which was the sickness, has become the path to the priesthood." From the Western Cape: "Let darkness turn to light." The single most characteristic feature of ngoma is this transformation of the negative, disintegrative affliction into positive, integrative wholeness. There are, of course, other religious and therapeutic traditions in which this occurs. One thinks of the Western

self-help orders, even psychiatry, in which the student receives psychoanalysis before being able to practice.

In the several local traditions of ngoma which have been used repeatedly to illustrate the genre in this book, we have seen a variety of applications of this theme of transformation. Barrenness and threatened miscarriages give way to techniques for fertility and child care in the Southern Savanna and Western Bantu settings. In the Kinshasa urban setting, the trapped-wife syndrome is replaced by membership in the Zebola network of the formerly isolated. In western Tanzania, snake-bite is turned into knowledge of the use of venom for immunization. In Southern Africa, the chaos of rapid industrialization feeds the enhancement of divination techniques; the fragmentation of the family household leads to the amplification of networks linking fragile households.

The energy that comes to the afflicted from joining those who have "been there" and who have survived or recovered, or at least learned to cope with the affliction, often is represented in so-called "medicines" of office such as the kici of Western Bantu, or the kobe of Central Africa. Acquiring these medicines, or charms, or techniques, is part of the end stage of "the white," either generated by the novice during the long course of therapeutic training, or assembled at the time of his or her "graduation." Many times these medicines are outright magical or metaphorical mementos of the affliction, mementos that through association with material objects from cosmological, vegetal, animal, or human domains become statements about the condition of the novice. I come back to this process in chapter 5 and interpret it further under the rubric of "metaphors of difficult experience."

Much of the empowerment of ngoma is, however, in the newly acquired ability of the afflicted to meet specific or general problems with resolve. This often entails the adoption of a healer-prescribed prohibition or rule, illustrating the proto-Bantu concept gìdò . As we will see, the personal song, developed through dream study, and through sharing ngoma with one's fellows, also plays an important part. Subsequent chapters will explore this further.

Conclusion

The core features of ngoma presented in this chapter are relatively static features. On their own they are merely aspects of the wider general culture or set of behaviors. Therefore we must stress that they are

not to be taken too literally as the attributes of ngoma everywhere or that where they are identified, they have to do with ngoma. Together they reveal significant characteristics with which to understand how ngoma-type affliction and healing is organized.

The features discussed in this chapter are usually not identified by descriptive indigenous terms, although I have been able in some cases to relate them to proto-Bantu cognates. On balance, much of the most characteristic behavior in ngoma does not correlate strictly with a given vocabulary. It can be described, discussed, and interpreted, but it is not locked into a technical fixed vocabulary. This would suggest that the phenomenologist's constraint of studying only that which is consciously and verbally identified would be ill-advised, and that a culture's reality, while it may be described and shaped by language, is certainly not limited to verbal cognates.

Verbal cognates and these core features come together in a poignant action-term, "doing ngoma," to which we turn next.