Conclusion

A major goal of this work has been to explore the basis for the institution variously known as the "ritual of affliction," "cult of affliction," or "drum of affliction," the latter term being derived directly from the widespread notion ngoma, a Bantu language cognate. Utilizing a variety of historical, linguistic, archaeological, and comparative sources, the case was made that ngoma may have emerged as part of the classical Bantu expansion over two millennia ago, although some evidence points to a more recent Eastern Bantu origin.

Nineteenth-century culture historians might have said that we had perhaps identified a health and healing "complex," a set of interrelated traits and practices, and unique characteristics. Others might have sought to explain it in terms of its origins. One reader of this manuscript noted that if ngoma was as old as suggested, it must have responded to a distinctive need during the original Bantu expansion.

Unfortunately, to unearth such an original purpose is beyond our reach. Perhaps one day very sophisticated sociolinguistic history or semiotic analysis of bodily motion will divulge the prehistory of ngoma. This work does not rest its case for ngoma's existence solely on a rigorous cultural-historical approach to ritual. Rather, utilizing a broad survey approach, I have demonstrated certain underlying common properties—linguistic, behavioral, and structural—that seem to make it appropriate to speak of ngoma as an institution, in the tradition of Durkheim, Weber, and Marx.

The formal properties of this institution were quite readily identified: the phased rite of passage, an approach to classifying misfortune which emphasizes the place of the social context—of people; the role of the shades and spirits in the outcome of dealing with that misfortune; the release of the sufferer from normal social obligations, and the progression of the sufferer-novice through the "white" liminal sick role and initiation to full membership in the order; the pervasiveness of sacrifice, which sets in motion an exchange between the living and the supernatural, as well as creating the bonds and networks among the living through the cooking and distribution of the sacrificial animal; finally, the empowerment of the initiate as qualified healer.

However, these features do not hold together as an institution until they are identified in their performance context, that of "doing ngoma." In this setting, a far more dynamic set of features is at work. The actions, described and analyzed in chapter 4, bring together every level of participant—novice, trainee, senior qualified person, as well as whoever else is present—to share of themselves and to respond to one another. Doing ngoma sets the stage for each and every individual in the order to work on self-transformation through song, as if before a constant mirror of others also seeking the same goal. Together, they seek renewed self-definition, in personalized song statements that resonate with the common terms of the ritual. Western institutional analysis has all too often crushed and distorted ngoma-type phenomena by forcing them into its own familiar slots and categories. In the present work I have tried to interpret the uniqueness of the institution in a number of other ways.

Ritual analysis offers one approach. Ritual may be defined as the amplification of layers upon layers of meaning, of levels or mediums of expression, or as the addition of more lines of communication to those normally used between individuals. Ritualization may occur because of "clogged channels" or contradictions in the individual's life, or in the household unit or society around the individual. Or it may result from unique ruptures within the social environment or from chronic suffering of sectors of society, such as women with children or working men alienated from their families. It may result from constant segmentation of the lineage basis of society.

Our task would have been facilitated had there been a conscious awareness in Central and Southern Africa of an overarching institutional presence. In my survey work in Kinshasa, Dar es Salaam, Swaziland, and Cape Town I encountered little or no evidence of this among

patients and healers, nor among scholars. Healers did have an awareness of regional networks, of common ties. In western Zaire, there was awareness of the comparability of Nkita, Bilumbu, Zebola, and the other grands rites , and in Tanzania, of the existence of other ngoma than the ones practiced, including those from other regions. In Southern Africa, there was a greater degree of mixing and cross-fertilization of ngoma or the work of the amagqira than anywhere else. In some ways, scholars were more provincial than the healers, in that they often limited the focus of their research to tribal and national entities, for example, the Zulu diviner, the Zigua rite of affliction, the Ndembu cult of affliction.

This absence of the awareness of a larger institution and its character called for a further ontological query of whether the overlapping levels of evidence—the verbal cognates, the behaviors—could provide us with a clear picture of what ngoma was all about. Healers and scholars alike were intrigued by the possibility that their local institution might be part of a larger entity. They were able to point out recognizable common practices or features in photographs and musical recordings, as well as differences between their own practices and those of others. Some asserted that the version from afar was "wrong" in terms of the way they did it. Nevertheless, there emerged in practitioners' accounts of "how ngoma worked" a more or less common theory that words and spirit etiologies—often as expressed in song—provided the core working principle of ngoma. This aspect of ngoma needs extensive further research, as very little case material exists on the way in which individuals reach creative self-expression through the crucible of affliction.

But we are getting ahead of ourselves. What finally is going to be the name of this newly identified, widespread, apparently ancient institution? The name is no trifling matter, for on the choice of name will hang much of the identity of the institution. "Cult of affliction," "possession cult," "divination," "rite of affliction," "drum of affliction," and ngoma have been used more or less interchangeably. The first several are recognized analytical names; the last, an indigenous term that has multiple meanings and is therefore ambiguous. The advantage of utilizing an analytical term is that we can say it exists even if the locals give it another name. However, the advantage of utilizing the African name, ngoma , or another variant, ngoma za kutibu ("therapeutic ngoma," in the words of E. K. Makala in Tanzania's Ministry of Culture), or "ngoma-type ritual" or therapy, to cite Anita Spring, is that

we tie into the conscious level of awareness of it. This offsets certain scholarly excesses that have distorted the identity and definition of the institution in the literature.

This perspective helps us overcome particularly the Western preoccupation with "trance" and "possession" as the central definers of the institution, which the term cult denotes. Fixation upon natives with upturned eyes grunting unintelligible spirit talk and dancing about in a frenzy has seduced scholars from the important task of looking at the context, structure, history, intention, and change of an ancient instituted process. As has become clear in the course of this study, trance behavior is but an occasional corollary of the etiology that attributes misfortune to ancestors and spirits. The spirit possession etiology is a type of ideology or hypothesis, usually invoked by diviners, to interpret a range of usually vague symptoms or signs, although at other times it is used in very clear-cut expressions of distress. The spirit hypothesis, as I like to call it, rarely blurs the practical exigencies of day-to-day life. It is rather a framework that sets up and legitimates an institution that may bring together many kinds of perspectives and theories. To limit ngoma expression to possession trance is particularly unfortunate because its outcome is the exact opposite of possession, namely, creative self-expression. Trance, when it does make an appearance, is usually an aspect of the performance ritual and needs to be seen in the context of dialogical music, confessions, and divinatory sessions in which problems are being worked out. Only occasionally do trance and possession join as a shamanic-type journey of the analyst.

For these reasons, this work has stressed the centrality of discourse in ngoma interaction and knowledge. The "doing ngoma" process is a format for the identification of sources of misfortune, for bringing out and articulating, energizing, and transforming individual identity and purpose, and for gathering others who are in a position to vicariously help with these tasks. It is a process that may address any type of situation in which the form is servant to the content. It may accommodate a great latitude of types of knowledge: common sense, the hypotheses embedded in spirit fields; technical knowledge of the natural world, scientific knowledge based on propositions, social principles in kin and non-kin society, and other types of "survival knowledge," to cite Lihamba (1986). Gregory Bateson long ago emphasized that one of the functions of ritual was the preservation and utilization of knowledge.

Readers will have recognized the similarity between ngoma-type

healing orders in Africa and Western self-help institutions such as Alcoholics Anonymous, Weight Watchers, Parents Anonymous, networks of organ transplant subjects, parents of sudden infant death syndrome victims, cardiac rehabilitation groups, and many more. These, in the West, have often had the benefit of biomedical consultants or liaisons, although in their organizational character they have been independent of direct medical institutional control. Part of their success in the lives of members is the transformation of the self in the very area of prior weakness or failing.

Out of this process emerges a further agenda of this work, namely, to define health, to test the institution's impact on health, and to determine its efficacy in either preventing deterioration of health or in restoring health where it has deteriorated. There is much variability in the criteria by which health inventories are judged. The theoretical terrain of this discussion is fraught with philosophical dilemmas about the identification of sets of variables used in defining health. In this work, criteria that have been brought forward have had to do with words and consciousness, labeling of diseases, and the predisposition of the sufferer; with the role of metaphor and symbol and the amplification of these expressions in ritual performance; with the creation of communities of the afflicted and the importance of sharing one's problems with others.

Because of the prevalence of a trance and possession definition of ngoma in much of the scholarly literature and popular culture, or the implication that ngoma's therapeutic orientation is psychological only, there has been a certain resistance on the part of scholars and policy-makers to consider ngoma's role as it addresses epidemiological and demographic definitions of disease. Yet as we have seen, there appears to be significant evidence that ngoma rites such as the Tanzanian snake handling or the Southern Savanna fertility-enhancement procedures for mothers and children at risk may directly address these issues.

Given the promise of significant ill-health risk reduction in some particular ngoma institutions, what are the prospects that they can be engaged more intentionally in the agendas of health ministries and other agencies? One issue that has emerged in this regard is the struggle over the organizational matrices of ngoma entities. For the most part, ngoma organizations represent an "uncaptured" power base whose leaders and resources, decentralized as they are, resist co-optation by the nation-state. Where there has been centralized organization, as in coastal

Tanzania, struggle for the control of ngoma resources has remained a problem.

Ngoma resources have been effectively integrated into national institutions in Swaziland, where the king has personally endorsed, and underwritten the care and training of, skilled heaters such as Ida Mabuza. It is possible, but difficult, to generalize this model of integration in the modern state, where bureaucratic rules and decisions are prevalent. It might work to the extent that leadership remains personalized within the mantle of bureaucratic rationality. However, in such a patrimonial system there is enormous competition for the legitimation that such personal support entails, and it hardly provides a basis for consistent and equitable distribution of resources to a needy populace.

It appears most appropriate to seek integration of ngoma-type institutions, and their overall enhancement, at a less bureaucratic level, such as has been sketched in Zaire. There, the state health service has recommended general hygienic instruction for all healers to upgrade their contribution to health without eroding their power base. Public health and hygiene instruction of healers in Ghana and Sierra Leone has produced significant reduction of infant mortality rates.

As the Southern African ngoma networks show, there are transnational channels of information at work that tend to distribute new knowledge. Ngubane writes that the shifting cell composition at successive sessions of the sangoma uniquely serves to spread insight about healing techniques and solutions to problems. In theory, it should be possible to introduce knowledge of hygiene, public health, and other areas of public concern at any point in such a network and see it spread throughout the system. Of course, ngoma knowledge, as we have seen, has its own assumptions and structures. Ancestral or spirit legitimation, rather than individual legitimation, must remain prior. This is Africa's way of stressing the moral basis of useful knowledge in the institutional process. Yet the distinction between common and personal songs allows for the introduction of new personal knowledge. To the extent that personal knowledge, or messages about personal transformation, are generalized and taken on by others, and then become common knowledge, ngoma settings may serve as channels of new outside knowledge and information.

Ngoma, an apparently classical institution of Central and Southern Africa, deserves our consideration, whether we are interested in its

dance and discourse format, its therapeutic instrumentalities, or its applicability to the resolution of societal problems. We do well to consider ngoma as a uniquely ngoma instituted combination of processes and attributes that the West puts together in other ways, or leaves undone.