Regional Cults

[Front Matter]

[Synopsis]

From the introduction:

“Regional cults are found in many parts of the world. They are cults of the middle range — more far-reaching than any parochial cult of the little community, yet less inclusive in belief and membership than a world religion in its most universal form. Their central places are shrines in townsand villages, by cross-roads or even in the wild, apart from human habitation, where great populations from various communities or their representatives come to supplicate, sacrifice or simply make pilgrimages.”

This book (which is derived from the 1976 ASA Conference) is the first attempt by social anthropologists to create a theoretical framework for the comparative analysis of these cults. It gives the study of religious organizations a new perspective on major historical transformation from the pre-colonial past to the neo-colonial present, in North and South-central Africa, South America and the Middle East. It advances a basic debateabout opposed theories of religion and society inspired by Robertson Smith, Durkheim and Marx. The focus is on the problems of change in trans-cultural symbolism and relationships which cross political, economic or ethnic boundaries. Most of the case studies are from South-central Africa, allowing a comprehensive view of cults in a core area over a long period of time. The overall coverage is also inclusive of Islamic marabouts, Christian and non-Christian prophets, local territorial mediums, High God oracles. Numerous maps, diagrams and charts illustrate the distribution and regional patterns of the various cults.

The book introduces new concepts and a rich body of data of interest to social anthropologists, sociologists, students of comparative religion, history, politics, economic development and cultural evolution.

Association of Social Anthropologists: A Series of Monographs

13. J.B. Loudon, Social Anthropology and Medicine — 1976

14. I. Hamnett, Social Anthropology and Law — 1977

15. J. Blacking, The Anthropology of the Body. In preparation

16. R.P. Werbner, Regional Cults — 1977

[Title Page]

[Copyright]

[Dedication]

Contents

Introduction

RICHARD P. WERBNER

Beyond Naive Holism

Level Specific Analysis

Turner’s Exclusive and Inclusive Domains

The ‘Correspondence’ Theory

Elitist versus Egalitarian Regional Cults

Hierarchy and the Waxing and Waning of Cults

The Interpenetration of Cults

Future Research

References

Part One: Islamic Cult Regions: Historical Interaction or Symbolic Congruence

1. Ideological Change and Regional Cults: Maraboutism and Ties of ‘Closeness’ in Western Morocco

DALE F. EICKELMAN

Ideologies and the Social Order

Maraboutism as a Part-Ideology

The Sherqawi Zawya

Maraboutism and “Closeness”

Maraboutic Descent

Ideology and Practice

Sacrifices and Social Groups

Conclusion

References

2. Communal and Individual Pilgrimage: The Region of Saints’ Tombs in South Sinai

EMANUEL MARX

Introduction: The Region and the Wider Universe in the Cult

Three Tribes and their Setting in the Peninsula

‘Tribal Organization’ as Ideology

Annual Migration and Pilgrimage

The Symbolism of Saints’Tombs

References

Part Two: Regional and Non-Regional Alternatives: The Waxing and Waning of Cults

3. Disparate Regional Cults and a Unitary Field in Zimbabwe KINGSLEY GARBETT

Introduction: The Interpenetration of Cults

The Region of Mutota

The Land-Shrine Community

Lineal and Territorial Aspects

Change in Ritual Territories

Expansion and the Regional Medium

The Decline of Chthonic Cults

Kupara’s Recruitment

The Linking of Local and Territorial Cults

Conclusion

References

4. Bush Negro Regional Cults: A Materialist Explanation

BONNO THODEN VAN VELZEN

Introduction: Radical Economic Change and Cults

Heydays

The Spread of a Regional Cult and its Centralization

Theology

Allocation of Resources and Internal Distribution

Gaan Tata’s Bulwark on the Tapanahoni

Gaan Tata’s Decline

A New Regional Cult?

Summary

References

5. A Continuing Dialogue: Prophets and Local Shrines among the Tonga of Zambia

ELIZABETH COLSON

Cults and their Publics

The Setting and the Cults

Shrine Custodians

Prophets and the Greater Community

The Diversification of Society

Continuity and Change in the Cults

The Gwembe Resettlement: A Test Case

Prophets and Custodians

References

6. Regional and Non-Regional Cults of Affliction in Western Zambia

WIM M.J. VAN BINSBERGEN

Introduction: Change and Types of Cult

Non-Rcgional Cults of Affliction

Regional Cults of Affliction in Western Zambia:

General Characteristics

The Development of Nzila and Bituma as Regional Cults

Conclusion

References

Part Three: Regional Instability: Cult Policy and Competition for Resources

7. Continuity and Policy in Southern Africa’s High God Cult

RICHARD P. WERBNER

Introduction: The Definition of Regions

Change in Congregations

The Development of a Region and the Limits of its Expansion

The Priesthood and its Managerial Problems

Hierarchy and the Making of Oracular Policy

References

8. Cult Idioms and the Dialectics of a Region

J.M. SCHOFFELEERS

Introduction

The Formal Organisation of the Cult

The Chiefs and the Shrine

The Spirit Medium and the Informal Operation of the Cult

The Medium’s Entourage

The Mbona Medium and the Chiefs

The Encounter between Medium and Principals

Rumour and Possession

The Conceptualization of a Regional Community

References

Notes on Contributors

Author Index

Subject Index

Maps

Introduction

1 Regional Cults in South-Central Africa

Chapter 1

1 Spheres of Sherqawi Influence in Western Morocco

Chapter 2

1 Holy Tombs in South Sinai

Chapter 3

1 Region of Mutota: Spirit Provinces and Chthonic Cult Areas, Dande (1963) Zimbabwe

2 The Ritual Field of the Region of Mutota (1963–69), Zimbabwe

Chapter 4

1 Gaan Tata Cult and Christian Mission Posts (c. 1900), Surinam

Chapter 5

1 Prophets and Local Shrines: Zambian Tonga and their Neighbours in Zimbabwe

Chapter 6

1 Ethnic Groups and Localities in Western Zambia

2 Origin and Spread of the Nzila and Bituma Regional Cults

Chapter 7

1 Congregations and Shrines of the High God Cult, Northeastern Botswana facing

2 Territorial Divisions in Northeastern Botswana

3 Interlocal Shrine and Band of Adepts, Northeastern Botswana

4 Congregations of Southwestern Region, Botswana and Zimbabwe

Chapter 8

1 Mbona Shrine and Surrounding Chiefdoms (1975), Malawi

Diagrams

Chapter 7

1. Regions and their Surrounding Vicinities

2. Cult Domain and Ecological Zones: Distribution of Oracles

Genealogies

Chapter 3

1 Perpetual Relations among Korekore, Chthonic and Chikunda Regional Cults

2 Mutota’s Descendants

3 Succession to Mediumships

Chapter 7

1 Genealogy of a Band of Adepts

Tables

Chapter 2

1 Bedouin Tribes of South Sinai

Chapter 3

1 Index of Spirit Provinces shown in Map I

Chapter 7

1 Succession of Adepts

2 Recruitment of Junior Adepts

3 Recruitment of Senior Adepts

Charts

Chapter 7

1 Grades of Shrine

2 Western Regions of the High God Cult

Chapter 8

1 Resume of Origin Myths of Two Principals

2 Mbona Mediums in the Twentieth Century

Introduction

by Richard P. Werbner

Beyond Naive Holism

Essential relations with a wider context get stripped away when a small group, little community or tribe is studied as an isolated whole. This idea is virtually a commonplace, and it is now conventional, also, to reject the method as a fault of earlier studies. Yet relatively little advance has been made beyond naive holism — indeed, it may be winning new advocates (Barth 1975; Douglas 1970). The challenge remains one that Max Gluckman raised in his early work (1940, 1942): to analyse change in fields of relations that cross political, economic or ethnic boundaries.

These essays on regional cults, which derive mainly from papers presented at the 1976 A.S.A. Conference in Manchester, take up that challenge.

Regional cults are found in many parts of the world. They are cults of the middle range — more far-reaching than any parochial cult of the little community, yet less inclusive in belief and membership than a world religion in its most universal form. Their central places are shrines in towns and villages, by cross-roads or even in the wild, apart from human habitation, where great populations from various communities or their representatives, come to supplicate, sacrifice, or simply make pilgrimage. They are cults which have a topography of their own, conceptually defined by the people themselves and marked apart from other features of cultural landscapes by ritual activities (on ritual topography see Turner 1974 : 183).

Diverse as the cases in this book are — cults of Islamic marabouts, Christian and non-Christian prophets, local and territorial mediums, High God oracles — they represent much less than the full range of regional cults. Their forms in different parts of the world are far too many to be covered in a single book. Hence this selection is concentrated in two ways. First, along with cases from North Africa, the Middle East and South America, there is a fairly representative coverage in one part of the world, South-Central Africa (see Map 1). Second, besides such ethnographic concentration, there is a theoretical focus: our primary concern is to explain change in the ideology and organization of cults which are based on a central place and its relations with its hinterland or smaller dependent centres. In Carol A. Smith’s terms these cults are ‘nodal forms of organization’ (Smith 1976b : 9), and though our focus is somewhat like hers, the differences in our theoretical interests need to be seen from the outset.

Level Specific Analysis

The major collection of essays edited by Smith (1976a, 1976b) develops a ‘level-specific regional analysis’, to give the approach a label of its own. Concerned primarily with complex societies, this approach relies on a model of levels — local, regional, national — and focuses specifically on the level (this may be a set, rather than a single level) that is always intermediate between the other levels in the model, namely the regional. Analytically, the limitation to a specific level has considerably force, at least for the comparative purposes which Smith and her contributors have in mind. Their cross-societal generalizations apply to standard units; they avoid the misplaced comparison of unlike orders of relations. But the gain in comparability is costly. Indeed, without radical modification it is too costly for the approach to be applied in order to make sense of change in regional cult organization and ideology.

In empirical terms, the cost arises because the approach makes no room for certain activities and phenomena which are crucial in regional cults. As Smith stresses, regional analysis aims to show how

‘flows of goods, services, information, and people through [a] network of centres ... structure and maintain internodal relations’ (Smith 1976b : 8, Smith’s italics).

In regional cults such flows run across major political or ethnic boundaries. Hence the characteristic direction of the flows, along with the cult’s distinctive topography, is of specific theoretical interest. However, the level-specific approach ignores activities overriding and not congruent with the boundaries and subdivisions of a nation. Similarly neglected are the phenomena which define international or non-national regions. Theoretically, the approach introduces a bias towards nesting relations and it does not allow for cultural variability, i.e., the variable importance of different conceptions held by the people studied themselves of their fields of relations. To overcome these and other limitations calls for quite a different approach to regional analysis, without the three tiered model or its level-specific assumptions. Hence, our analysis of regional cults takes as problematic and subject to explanation what level-specific regional analysis neglects or takes for granted.

Our theoretical interests lead us to pay most attention to the historical transformation of cult regions. How do the people themselves define and organize their cult regions over time, through radical economic or political change and also in more stable periods? How do religious ideas and practices alter as cult regions become more or less inclusive? What impact do regional and non-regional cults have on each other, and how do they wax and wane in relation to each other? Some of these are problems which Victor Turner, too, considers in a discussion of ‘Pilgrimages as Social Processes’ (1974 : 166–230). Turner’s suggestions differ somewhat from ours, and it is fruitful to clarify some of the differences, particularly because a good part of our work builds on this and other discussions of his (Turner 1957, 1967, 1968, 1975).

Turner’s Exclusive and Inclusive Domains

Parallel dichotomies are at the heart of Turner’s argument: inclusiveness versus exclusiveness, peripherality versus centrality, generic versus particularistic relationships, egalitarianism or homogeneity versus nonegalitarianism or differentiation. For our purposes, the most crucial is a dichotomy

between politico-jural rituals focused on localized subsystems of social groups and structural positions, and rituals concerned with the unity and continuity of wider, more diffuse communities’ (Turner 1974 : 197).

Only a few of his conclusions need concern us here, although it is through these dichotomies that he arrives at a much larger body of conclusions, along with a general perspective on the very process of pilgrimage itself. Three conclusions are relevant here: first, about opposed types of cult; second, about how far cult ideologies represent or correspond with various politico-jural relations; and finally, about the dialectical way that one type of cult becomes its opposite.

The conclusions have to do with opposed types of cult in two kinds of society, once more a dichotomous division. On the one hand, Turner refers, in ‘certain traditional African societies’, to earth and fertility cults (most of which are regional cults in our terms) versus ancestral and political cults (apparently non-regional). On the other hand, he refers, in ‘complex large-scale societies and historical religions’, to pilgrimages and the organization of pilgrimage centres (again, mainly regional) versus localized religious activities focused on local shrines, ‘which are themselves parts of bounded social fields, and which may constitute units in hierarchical or segmentary politico-ritual structures’ (Turner 1974 : 186) (again, apparently non-regional). This is his suggestion about the African cults:

... ancestral and political cults and their local embodiments tend to represent crucial power divisions and classificatory distinctions within and among politically discrete groups, while earth and fertility cults represent ritual bond between those groups, and even, as in the case of the Tallensi tendencies toward still wider bonding. The first type stresses exclusiveness, the second inclusiveness. The first emphasizes selfish and sectional interests and conflict over them; the second, disinterestedness and shared values’ (Turner 1974 : 185).

Similarly, he sees a parallel in the contrasting kind of society:

‘The networks formed by intersecting pilgrimage routes ... represent at the level of pre-industrial, politically centralized, high-agricultural cultures, the homologues of those reticulating the earth shrines and shrines of non-ancestral spirits and deities found in many stateless societies in sub-Saharan Africa with simple agricultural technology.

Here the peripherality of pilgrimage centers distinguishes them from the centrality of state and provincial capitals and other politico-economic units. It further distinguishes them from centers of ecclesiastical structure such as the sees or diocesan centers of arch-bishops and bishops’ (1974 : 197).

The great utility of this contrast is its illumination of extreme cases. Like other ideal typologies, however, it tends to represent as mutually exclusive alternatives — the halves of parallel dichotomies — what are, in fact, aspects which combine in a surprising variety of ways within a range of actual cases. In this book the chapters on Islamic and African cult regions provide such cases not covered by the typology. The point is perhaps best illustrated first by Schoffeleer’s chapter, because he views an earth or territorial cult — the Mbona cult of Malawi and Mozambique — in terms of dichotomies that are comparable to Turner’s Moreover, Schoffeleers resolves a paradox which is important for Turner’s discussion. As Schoffeleers puts it for the Mbona cult:

‘The very organization which asserts social difference and privilege becomes the vehicle through which the denial of difference and privilege is affirmed’.

Boundedness and political divisions are represented by the Mbona cult’s localized congregations and its principals. As chiefs, the latter come from the autochthonous ethnic group, the Mang’anja, with a legendary, or rather, ‘traditional’ realm, i.e., the current bounds of the cult’s region, and in legendary times, all their chiefdoms were under the centralized rule of a single paramount chief. Not that the cult region merely replicates some political organization which exists apart from the cult’s activities: the political communities which it covers are in different countries, and they do not now exist as one body under any paramount ruler or even a single government. Instead, the cult region has its own nodal organization, and its own centre alone provides the focus for the region, quite distinct from any other political or economic centres. Its spirit medium, usually an immigrant rather than an autochthon, but never a chief or cult principal, makes his pronouncements on the public morality of chiefs and criticizes their relations with their people. But he does so only during seances at the region’s innermost shrine, while he is possessed by the spirit of Mbona. For the whole region and from its centre he speaks with an authority that goes beyond that of any central government and is independent of it; thus, he readily becomes a focus of resistance to unpopular measures imposed by the government.

The centre of this territorial cult is the extra-territorial sanctuary, an area beyond the territorial jurisdiction of any government and thus free of its taxes and rules of land tenure. Within it, and well apart from the seat of any legendary paramount is the cult’s innermost shrine. Within the sanctuary also, the cult hero, Mbona, was murdered, according to myth, at the first paramount’s command, after Mbona became a rival for political influence because of his powers as a rainmaker. It is according to the location of events associated with Mbona’s martyrdom that the sanctuary is centralised and set apart in cult topography. Hence the myth about the sanctuary emphasizes more than separateness, externality or even centralization: it ‘projects an image of a region which is unified on the basis of common concerns and a common morality’, but it also presents a paradigm for power struggles and an opposition between chiefly and non-chiefly ritual authority.

Whereas Turner suggests that types of cult fit a ‘polar distinction between cultural domains of exclusivity and inclusivity’ (1974 : 186), Schoffeleers demonstrates that both such domains are polarised within the Mbona cult. Each donjain is emphasized by different protagonists who oppose each other in the cult. From one viewpoint, the Mbona cult appears as an unstable combination of both of Turner’s types, and it is unstable just as the domains themselves are, in relation to each other. Schoffeleers shows that there is a persistent struggle between the representatives of political hierarchy — the cult principals as chiefs — who assert the legitimacy of social differentiation, and the representative of their people — the medium — who puts forward ‘the conception of a non-differentiated society’ [what Turner might call ‘ideological communitas’ (1974 : 1697]. Furthermore, the dialectic in this struggle is such that the representative of the people comes to appeal increasingly to more inclusive conceptions and idioms:

Mbona is now said to be the “Son of God” (Mwana wa Mulungu) and the “black Jesus” (Yesu wakuda), who is the guardian spirit of all ethnic groups, and not just of the Mang’anja.’

Here, to carry the argument a step further, we must turn to Colson’s analysis of cults and their publics among Tonga in different ecological zones of Zambia. In some respects the Tonga prophet cult, compared with the Mbona cult, represents an extreme at the very margin of the range of regional cults. Insofar as a region emerges in the prophet cult, it is characterized by diffuseness and ephemeral, weak, even uncertain centralization, instead of clear ritual demarcation, recognised boundedness, and a firmly anchored nodal organization for the whole region. Hence the Tonga case illuminates the grey, marginal area of relations between cult exclusiveness or inclusiveness and continuities or discontinuities in local communities.

Colson concentrates her analysis on the changing interplay between the local and the prophet cult. The local cult always has a well-defined constituency, unlike the prophetic cult. In the local cult there are numerous ’district’ or neighbourhood land shrines which are the centres of small, discrete ritual areas whose shrine custodians are independent of each other. Neither the custodians nor their land shrines form any hierarchy or more inclusive body. Colson establishes that the local cult and its communal ritual give concrete expression to the continuity and exclusiveness that a particular, highly localized agricultural community may have. The cult is stable in its land shrines and ritual — indeed, more stable than Colson herself understood on the basis of her earliest fieldwork (Colson 1948, reprinted in 1962) In the light of her continuing observation from 1946 to 1973, she now considers that ‘local shrines are remarkably viable’: they survive despite resettlement and the relocation of neighbourhoods. Over the years of her research among Plateau and Gwembe Tonga, she heard of no new local shrines. Her long-term evidence points to a great measure of cult autonomy, rather than immediate responsiveness to local fluctuations and the movements of shifting cultivators. Important as the vagaries of local shifts in villages and their populations are, they do not directly bring about the rise and fall of local shrines: land shrines have a life of their own.

The relation between the local and the prophet cults is, in the main, complementary, though not unproblematic or unchanging. The prophet serves a variable clientele and, while possessed by spirits identified with no known social community (spirits of the wild, of long dead prophets, foreign spirits and most recently, angels and God), provides warnings about personal or communal troubles and guidance about ritual for healing or rain. Most are minor prophets and play their part within an immediate vicinity roughly the same as, or little more than their own local shrine community. Colson observes:

Though they claim a power that stems from beyond the community, and is independent of it, they are subject to the community controls. They link their messages to the local order ... These prophets are, therefore, among the principal upholders of community, the land shrines and the existing custodianships.’

For both prophet and local cults in the narrowest order of relations — the local order — it can be said, in Turner’s terms, that the main emphasis is on the exclusive domain, boundedness and stability.

Beyond the narrowest order, the prophet cult is more flexible. It alters its emphasis, whereas the more rigid local cult does not, and remains virtually contained within the local order. Moreover, the flexibility of the prophet cult threatens to become a danger to the local cult, rather than merely a complement to its rigidity. Recently, prophets have proliferated along with waves of new cults of possession by angels, God, and alien spirits (see chapter six where van Binsbergen explains the importance of this phenomenon in other parts of Zambia, also). Through’ its major prophets the prophet cult addresses itself nowadays to ‘the expanding universe’ of the country at large, and thus more to the changing inclusive domain. These prophets still build shrines where offerings are made for rain and the public good by communal delegations from near and far neighbourhoods. But now clients’ personal needs bring them to major prophets as distant healers. Such prophets appeal to a widespread, heterogeneous clientele from diverse ethnic groups and urban as well as rural areas. It is here, in the wider order of relations, that the prophet cult is most strikingly flexible in response to the economic and ethnic diversification in the country at large.

The Tonga case has telling implications for Turner’s dichotomous typology. Whether a single cult (such as the Tonga prophet cult) gives greater emphasis, during a period of its history, to one domain or another may depend on the order of relations involved: the narrower the order, the greater the emphasis on the exclusive domain, and conversely. Nowadays, the inclusive domain concerns major prophets perhaps more than ever before, and certainly more than it does the run of minor prophets. Clearly, Turner’s typology does not adequately allow for such internal relativity: that is, the variation within a cult according to different orders of relations. Yet, as the Tonga case proves, such internal relativity may be crucially important.

The Tonga case raises a further point about the central places of a cult and the folk conception of regionalism. This point leads aside somewhat from our discussion of Turner’s conclusions. But it must be considered here, because it is important for regional analysis in general, and our perspective in the Tonga case in particular. Colson’s evidence is that, in one sense, regionalism is still in its infancy as a folk conception among Tonga:

‘Even the idea of a social unit of all Tonga is a recent creation and is still likely to be invoked principally in the national political arena though the continued importance of the shrine of Monze may have political overtones of which lam unaware (my italics)’.

In the part I stress Colson allows for something problematic about the most famous Tonga shrine, Monze’s. It is a matter of historical record that Monze’s shrine was the central place of a very great, perhaps the greatest, Tonga prophet of the nineteenth century: he received delegations from Tonga, Ila and Sala neighbourhoods as much as 100 miles away. No longer does any prophet reside by Monze’s shrine, yet it continues to serve a very wide vicinity, and not merely a single local community. Delegations seeking rain still come to Monze’s shrine from near and far and, significantly, from distant neighbourhoods in another ecological zone where rain is even more scarce. Why so? Or rather, what does this imply about the central places of a cult among Tonga?

The clues to the answer are in Colson’s data on prophets. Although my interpretation differs somewhat from hers, it also takes account of the apparent tenuousness of formal regional relations. From time to time, prophets of a distinct kind have arisen — regional prophets, in my view — and they have arisen in the same central area, as shown on Map I of chapter five. This central area is on the escarpment edge in the high rainfall zone. In the nineteenth century Monze was one regional prophet, and close by Chibwe was another, from the 1940’s until her death in the 1960’s when her shrine ceased to be a centre for distant delegations. Colson records, ‘By 1950 various less-noted prophets were attempting to share in her prestige by claiming to have received her authorization to serve as local deputies’ (1958 : 34).

Formally, the regional organisation in this cult is rudimentary, without an extensive chain of formal linkages, elaborate devices for control of delegates, or an effective hierarchy of command. Fluctuations in the reach of particular major prophets or shrines are apparently meteoric. Diffuseness, rather than boundedness, is a characteristic of their catchment areas for the visiting communal delegations or minor prophets, and the cult thus lacks any culturally defined regions like the de jure territory of the Mbona cult. Indeed, the differences in culture between Tonga and their Shona neighbours in another higher rainfall zone represent no barrier to Tonga use of Shona spirit-mediums as prophets: some Tonga delegations from the lower rainfall zone supplicate with the Shona mediums for rain.

What the prophet cult has is a region (or regions) of the de facto kind. It is characterized by the flow of cult traffic, along with dependence for ritual services: a vague hinterland relies on the cardinal points in a central area. These cardinal points are represented by the prophets or shrines that become the focus of the widest communal traffic in the cult. Significantly, the delegations go from the lower rainfall zone’s communities to the cardinal points in a higher rainfall zone, never downwards to the lower rainfall zone’s shrine or prophet. It is a regularity observed in certain other cults also, such as the High God cult which I discuss in chapter seven. The directional cult traffic, flowing along with other trends of migration, carries forward the interdependence between people of different climatic or ecological zones, and a ritual region overlaps or cuts across these zones. Of course, not all cult regions are cross-zonal, as Marx argues for South Sinai in chapter two. However, I must make one further comment first, before I can discuss his argument.

The present sites of major prophets are central places more accessible for a widespread, heterogeneous clientele. Nowadays, Colson reports, ‘the most powerful set up their headquarters on the railway and bus routes rather than in the high rainfall zone’. In that zone, for the time being therefore, it is a prophetless shrine, Monze’s, which serves the communal traffic as a cardinal focus. I infer that despite the diversification in the country at large, cross-zonal interests do continue to underpin ritual centrality and stabilize the prevailing direction of the communal traffic. Even more, the patterns of cult traffic themselves have a resilience which makes for continuity through change.

The ‘Correspondence’ Theory

So far this discussion has explored problems which arise in studies of African cults. Others that Turner would regard as associated with ‘historical religions and complex large-scale societies’ must be considered now. There is a major disagreement in approach to religious organization and ideologies, and some contributors to this book are closer to Turner’s views than others. The disagreement appears most sharply, perhaps, in the studies of Islamic pilgrimage and marabouts’ cults by Marx and Eickelman. Not surprisingly, at the very nub of it lies a classical theory, Robertson Smith’s ‘correspondence theory’, and a long anthropological tradition stemming from him via Durkheim. The theory asserts a correspondence, in Eickelman’s terms, ‘between symbolic representations of the social world and patterns of conduct’. By developing Robertson Smith’s fine insight into the ‘community of the god’, Marx supports the theory. Eickelman rejects it, on the basis of Moroccan data, in order to get a better grasp of the ideology and practice of maraboutism in different historical contexts. Marx concentrates on the interests represented in Bedouin pilgrimages and symbolised by their marabouts and tombs. The issues are critical for the analysis of regional cults elsewhere. Hence Thoden van Velzen, van Binsbergen and I explicitly debate them also. However, here the disagreement is best appreciated against the background of a single historical religion and in two studies which advance our immediate discussion of Turner’s conclusions.

The ‘community of the god’ is a conception of boundedness, within the exclusive domain in Turner’s terms. Membership in the ritual congregation is on a closed basis; its members share rights in resources, recognition of their vital interdependence, or a concern with the maintenance of a group. The correspondence theory suggests that their cult’s spiritual subjects (i.e. marabouts or patron saints) and material objects (tombs, shrine buildings, shrouds and other ‘paraphernalia) stand for something else: there is a symbolic representation — Marx would go so far as to say ‘a very precise image’ — of particular interests (in the case in point, territorial interests). It fits Turner’s understanding of the exclusive domain, also, that the ‘community of the god’ has its own shrines, the foci of segments, in central locations valued for communal control of crucial resources. After all, the contrary — the peripherality of shrines — would express and foster the tendency towards universalism, openness, and emanicipation from the community (Turner 1974 : 193f).

Thus Marx observes the following about the social unit mainly involved in communal pilgrimage among the South Sinai Bedouin: ‘The tribe claims not so much to control a clearly bounded territory, but defined points in it and paths leading through it’. Accordingly, he finds the communal pilgrimage shrines on routes, by water points and at cross-roads that are highly valued for their strategic importance in trade and pastoralism. His evidence is best, and thus he concentrates, on one of two clusters of shrines. It is the one confined to the very heart of South Sinai, with its shrines on or near the main east-west passage through the mountains. These shrines, sited close to the edge of what a tribe claims as its own, are centralized at points where tribes meet or divide, rather than right by the more inaccessible settlements recently built as fastnesses in the mountains. Each shrine, except the peninsula’s most senior one (Nebi Saleh which is near Mount Moses, that worldwide pilgrimage centre for Moslems, Christians and, nowadays, Jews) serves a single tribe and its sections as a central meeting place for communal ritual. Annually, it is a distinct social unit, the section or tribe, that sponsors its own occasion: members share their animal sacrifices in the name of the one, universal God (the marabout is an intercessor, not a lesser god) and create commensal bonds with each other and their guests at the shrine. When the five main tribes of the mountains all make communal pilgrimages to the same shrine, Nebi Saleh, they do so in turn, on separate occasions. For some time, however, even this shrine has had a narrower reach. Communal pilgrimages are now held there only by the largest tribe, the Muzeneh, which of course asserts claims to the peninsula’s most central points and the senior shrine itself. Apparently, the change took place before the Israeli occupation ended smuggling and thus altered the value of passages and intertribal interests.

Marx surmises

‘that the location turned this tomb into the central meeting place of the Bedouin of the mountains; it is situated on the cross-roads of the major east-west and north-south mountain passes, and close to water. In view of the great importance of smuggling in those days (pre-1965), each tribe attached major significance to its rights of passage through the mountains, as well as to co-ordination of activities and maintaining peaceful relations with other tribes’.

Having gone with Marx on pilgrimage to the top of Mount Moses, I am tempted to pursue a wider concern. Cult regionalism is not tribalism writ large. Marx is himself very frankly puzzled by some aspects of the shrines and their significance for Bedouin — in my view, rightly so, because these aspects challenge him to break the frame of his analysis. Furthermore, they go right to the crux of our discussion of the exclusive and inclusive domains in relation to regional cults.

One shrine with its own marabout, shroud, and tomb is enough, at least for the purposes of a single section or tribe as a ‘community of the god’, and from the viewpoint of a correspondence theory. More than one is too many, and several differentiated ones for a certain tribe alone and none of the others does not fit at all. Admittedly, Marx’s understanding of marabouts and tombs goes beyond Robertson Smith’s insight into the religious representation of boundedness. Marx perceives that an enshrined marabout has an inclusive aspect also, ‘a second face’ so to speak, just as ‘tribesmen reserve certain rights in territory for themselves, but ... other rights are expressly opened up to all Tawara [literally, ‘of the Mount’, i.e. Sinai Bedouin collectively] on a mutual basis, i.e. rights to pasture, drinking water and, without prejudice to the tribe’s interests, passage also. However, while this perception of correspondence reaches the common denominator in the general run of instances, it stops our understanding short. Left more of a puzzle is the overall configuration of these shrines and marabouts, which is not a correlate of the pragmatic interests of Bedouin qua tribesmen. What is the significance of the differentiation of cult centres and their more universalistic aspects? How does this relate to their distribution in space, to cult topography? And what does all this imply about the cult itself?

Marx leads us a good part of the way towards the answer, despite the burden of the correspondence theory. He unpacks the symbolic significance of the three marabouts, including the cult’s senior, whose shrines at the geographical heart of the peninsula, so strategic in its mountains and passages, form an innermost cluster around the worldwide pilgrimage centre on Mount Moses (Jebel Musa) and the Christian monastery of Saint Catherine. His interpretation leads me to suggest that this innermost cluster lays down a central configuration in the cult’s topography, and it is defined by the cult’s cardinal orientation. Nowadays, with the largest tribe as hosts, communal pilgrims go from Faranja to Nebi Saleh to Nebi Harun (see Map I in chapter two). Each shrine, as a station in the pilgrimage circuit, has a positional significance, but all together convey the configuration’s pattern. In space it is a movement towards Mount Moses (the peninsula’s holiest place for individual pilgrims, whose holiness the Bedouin compare to Mecca’s) and the monastery (by which the shrine of Harun, the priest Aaron, is fittingly sited). Symbolically, there is an increase in inclusiveness, moving from the most particularistic and tribal marabout, the host pilgrims’ tribal ancestor Faraj, to the senior marabout of the region, the pre-Islamic Arabian prophet (some say, another tribe’s ancestor) Saleh, and last to the most universal figure, Moses’s brother the prophet Harun. Nowhere else is there a repetition of this pattern, just as nowhere else is there a focus for the traffic of individual pilgrims that is comparable in its holiness to the region’s centrally sited holy place, on top of Mount Moses. To have considered individual pilgrimage at length along with the communal would have taken Marx well beyond the limits set for a single essay. But allowing for this limitation, the point here applies to both kinds of pilgrimage. On the whole the pilgrim shrines are central, rather than being instances of ‘sacred peripherality’, in Turner’s terms (1974 : 195); yet neither in organization nor ideology does the cult replicate the rest of what might be called ‘Bedouin society’. The cult has its own cardinal orientation, and this is towards the greatest inclusiveness at its very centre. Moreover, without any permanent staff and for both the individual and communal traffic, a nodal organisation is sustained overall through ritual activities. Here it is ‘sacred centrality’ which expresses and fosters the tendency towards universalism and openness (cf. Turner loc.cit., on Mecca).

Whereas Marx casts light on enshrined marabouts among Bedouin, with their characteristic egalitarianism or rather, subtle, unformalized differentiation, Eickelman illuminates the enshrined along with the living marabouts based in a Moroccan town, Boujad, where an elite is culturally recognized and, even more, legitimized through maraboutism and the notion of baraka or grace. Like Marx, Eickelman recognizes aspects of maraboutism which, taken in isolation, would support a correspondence theory’s one-to-one correlations. However, Eickelman’s data compel him to take account also of disparities which do not fit a correspondence theory whether it be, as he demonstrates, the version of Robertson Smith, Gellner, or Shils. Eickelman observes:

‘Maraboutism is especially linked to two key concepts through which Moroccans make sense of existing social arrangements: obligation (haqq) and closeness (qaraba)’.

With these concepts, some’Moroccans draw an analogy, man is to God as man is to man, which makes sense of both relationships in particularistic terms. There is a flow of benefits in each, and it is contingent upon specific bonds: the greater the obligation and the closer the bond, the greater the flow of benefits ’ (and conversely also, of course). Such reasoning, implicit in belief in marabouts (with the benefits of grace or baraka contingent upon closeness to God), is virtually the correspondence theory in a folk version. It is a great strength in such a folk version, and a crippling weakness in analysis, to absorb differences and give them a semblance of similarity, irrespective of varying historical contexts. Eickelman’s analysis grasps the differences and contextualizes transformations recognized by the people along with others they do not, and perhaps cannot, recognize.

As he shows, change takes place at different rates and with more radical consequences in certain parts of the Sherqawi lodge’s region. Away from town and the regional pilgrimage centre, in tribal sections or rural local communities believers in marabouts continue to seek dyadic bonds with members of a town-based elite. Partly through local feasts and commensality, the aim still is to put under obligation the living marabouts or marabouts’ descendants who come on visits to their clients’ communities. Moreover, some tribal sections are still able to insist, despite opposition from townsfolk, that their annual pilgrimage to the town centre must be an occasion for animal sacrifice, public feasting, commensality and all that this implies in the creation of specific bonds of obligation and the maintenance of covenants. Against such continuity runs a counter tendency which is most apparent at the centre in town, though it is of consequence elsewhere also. Offerings are converted into cash by Sherqawi visitors and the caretakers of major Sherqawi shrines. Moreover, whereas in the past marabouts and the economic and polical elite were virtually one and the same, living marabouts are now few, and not of the Sherqawi elite that has contemporary political and economic importance. Furthermore, many of this Sherqawi elite now dissociate themselves from maraboutic enterprises and ‘seek to redefine the activities of some of their kinsmen and ancestors in the light of the formal doctrines of reformist, modernist Islam’. Thus along with a tendency still most evident in the hinterland of the region, Eickelman points to a counter-tendency which is foremost at the centre, though not restricted to it. The counter-tendency is to favour mediation or the relaying of grace through an enshrined marabout, rather than any living person, and even further, to turn away from the personification of grace towards formal Islam which is universalistic and thus ‘requires no direct analogy with the nature of the local social order’.

Elitist versus Egalitarian Regional Cults

At the risk of being taken to task for a lack of consistency — in the light of my general criticism of dichotomous typologies and Turner’s in particular — lam tempted to make use of one for comparative purposes here: elitist versus egalitarian regional cults. Ralph Waldo Emerson’s homely adage, from his celebrated essay on Self-Reliance, urges me on: ‘A foolish consistency is the hobgoblin of little minds’. I find the typology useful to clarify certain patterns and to show how their permutations are regular in the four cases mainly considered so far. The first kind of cult, the elitist, is characterized by processes which bring an elite and its public into dialectical opposition over the relative emphasis given in the cult to the exclusive or inclusive domains. Our examples of the Mbona cult and the Sherqawi maraboutic lodge illustrate this, and together represent an elementary permutation: the pattern in one appears in the other in reverse.

Briefly, to begin once again with the Mbona cult, the public appeals through its representative — so centrally placed in his extra-territorial sanctuary — to a conception of an undifferentiated society. It turns to universalism without ethnic bounds in the image of the black Jesus, whereas the elite of cult principals asserts its tribal authority and is particularistic in its appeal. By contrast in the case of the Sherqawi lodge, it is the public in the rural local community, the tribal section and elsewhere which appeals to greater particularism and continues to seek ‘closeness’ to a living person with grace or baraka as well as commensal bonds through shared sacrifice at an enshrined marabout’s holyplace. The centrally placed Sherqawi elite, on the other hand, now turn to universalism in the form of reformist Islam or they seek to free the offering of donations from commensality and its association with boundedness and the most particularistic oneness.

Underlying these inverse patterns is one or perhaps several regularities. In the elitist cults the emphasis on the inclusive domain is increased from the cult centre, rather than from elsewhere. Who fosters this increase depends on who is in control of the centre. On the one hand, the public or its representative fosters it where the centre is an extra-territorial one, under a religious authority and beyond the elite’s. Here the Mbona cult is the case in point. On the other hand, the increase comes to be fostered by members of the elite where the religious centre is the political one under the elite’s own control. Further studies of elitist regional cults are needed to clarify what are the limiting conditions for this regularity. But the point that must be stressed is that this is a regularity of cult organization and ideology; it is not a mere reflection or replication of some other social order. Religious change from particularism to greater universalism has its own momentum rather than being a mere correlate or after-effect of other transformations.

Constant in both these cases of elitist regional cults is the stability of the centre. In both this remains in a fixed location and is somewhat like a stationary capital, providing an established point of reference around which cult relations refocus. There is an alternative in which the cult centre is still the cardinal point for refocusing but itself shifts in location and is somewhat like the mobile headquarters of an army on the march, ever ready to be pushed forward with each major advance of the cult. This alternative, especially important under certain conditions of cult expansion, still correlates central control or cult centrality with the greater emphasis on the inclusive domain, as shown by Evans-Pritchard (1949: 70 f) in his classic account of the Libyan based Sanusi Order. Emrys Peters is working on a detailed reanalysis of the Order’s development, which will illuminate our discussion further. Here the comparison between elitist and egalitarian cults has to be explored on the basis of our four cases in order to complete a view of these as one whole set.

Our cases of egalitarian regional cults are the Tonga prophet cult and the South Sinai maraboutic cult. Each is characteristically free of any political elite with special ritual roles in the cult. The South Sinai case is a straightforward instance of a familiar regularity in that the cult’s stable centre is the focus of an increasing inclusiveness. The difference here is a lack of any dialectical opposition between recognised protagonists of the different domains. Perhaps less straightforward and thus more revealing is the Tonga case. Not only is the location of a given center variable within the more enduring focal area of the higher rainfall zone but cult centrality itself is also intermittent, somewhat ambiguous and problematic. Nowadays formerly peripheral points attract prophets and they scatter away from established shrines. Seen in relation to the other cases, it suggests the hypothesis that the greater the central instability, the more cult peripherality comes into its own for refocusing a greater emphasis on the inclusive domain.

So far all of my comments about this set of cases have been made with a general condition, common to them all, left implicit. It must be made explicit now in order to bring the next, contrasting set of cases into relief. Briefly, none of the first are clearly hierarchical whereas four of the second are; as will be seen, this poses further problems for analysis. Along with the lack of hierarchy goes a variation from cult to cult in the degree to which there is a unified ranking of cult shrines and holyplaces, although none rival the more formalized ordering in some of the second set’s cases. The variation is as follows. At one extreme is the Tonga cult which is without overall ranking but whose prophets and shrines are distinguished according to their participation in different orders of relations. Next comes the South Sinai marabout cult with its one acknowledged senior but somewhat ambiguous or unranked standing for other shrines. After these egalitarian regional cults come the two elitist ones, the Mbona cult and the Sherqawi lodge in that order. The regional medium of Mbona alone speaks for the whole of his region, and the lesser territorial mediums are restricted to narrower local constituencies. Finally, apart from the local maraboutic tombs which are points of domestic pilgrimage for women and children, the Sherqawi lodge has two grades of shrine within the pilgrimage center of the town of Boujad. Each quarter has its own shrine for rural and urban pilgrims, and above this grade is the one main shrine for them all, that of the lodge’s founder Sidi Mhammed Sherqi.

Hierarchy and the Waxing and Waning of Cults

To understand the importance of hierarchy and how it develops along with other cult transformations we must now consider our second set of cases. Our analysis of the waxing and waning of cults must be extended to cover a further range of problems, and shifts from non-regional to regional cults have to be explained. Already our discussion has moved towards the third of Turner’s conclusions, which is about cult transformations, and it helps to clarify our views for us to examine this conclusion explicitly, at this point, with further light from the rest of our cases.

Turner argues that ‘Pilgrimage centers, in fact, generate a “field” ’ (1974 : 226). Around pilgrimage centers a process of oscillation takes place or, more forcefully, gets its impetus: shrines, once peripheral and apart from ‘politico-economic centers’, attract people who make them structurally central; and then, over time, they revert back to peripherality. The model of the perpetual pendulum has no hold here, however, because it is a cumulative process, in Turner’s view. He considers that ‘pilgrimages sometimes generate cities and consolidate regions’ and yet ‘they are sometimes also the ritualized vestiges of former sociopolitical systems’ (1974 : 227). In other words, “the field” generated by pilgrimage centers is, at once, more and other than any contemporary ‘sociopolitical system’ with which it is, in fact, in a dialectical relation: increasing boundedness bears within itself the seeds of its own destruction, and conversely. Yet the question which has to be answered here is this. Across which boundaries, and in which respects, can a regional cult extend its hierarchy and yet still assert the kind of boundedness which, by its very nature, the hierarchy requires?

Turner’s own classic studies of the Ndembu cults of affliction, as reconsidered by van Binsbergen in a wider context and within an historical perspective, open the way to an answer. The first step that van Binsbergen takes is to free the concept of cults of affliction from any necessary bondage to the unitary social system. It is a radical departure from Turner’s view which left in the background crucial continuities in the ritual field:

‘But, by and large, the function of maintaining the unity of the widest social unit, the Ndembu people, devolves mainly upon the ritual system, although the ritual system spreads even wider than this’ (Turner 1957 : 292 my italics).

Turner’s view is tied not merely to Ndembu but, more fundamentally, to a single tribe or society taken as a ritually integrated whole. Yet successive cults of affliction sweep like waves of fashion across Western Zambia’s cultural and tribal boundaries. Hence van Binsbergen suggests that

‘The non-ancestral cults of affliction represent an attempt to come to terms (both conceptually and interactionally) with the reality of extensive, inter-ethnic, interlocal contacts’.

It is no paradox, however, that van Binsbergen’s observations, even more than Turner’s, illuminate the significance of local factors in the cults. Elsewhere I have argued (1971 : 318f) that Turner’s evidence does not sustain his conceptual model of rituals of affliction. According to Turner’s model, individuals associate in these rituals on an ad hoc and ephemeral basis; their community is the community of shared suffering, no other; they form a congregation, as van Binsbergen puts it, ‘in a pattern that cuts across, rather than reinforces the structure of distinct local communities’. Turner’s evidence, on the other hand, shows that a major performance of a ritual of affliction has primarily an invited congregation whose members nearly all have prior ties of one kind or another with the sponsors. The occasion provides an index, in effect a Who’s Who, of people and relationships important for the sponsors, but there.is no ritualization of a structure of localized groups, their divisions or boundaries. This kind of indexical occasion mobilizes as a congregation the interpersonal sector which is dispersed around a village, rather than the village itself in ritual isolation. Thus it is not the kind of event which emphasizes virtually random inclusiveness or participation ‘from all over the Ndembu region’ (Turner 1957 : 295). Here van Binsbergen examines the local organization of the cults more closely than Turner does, and thus establishes that, at least among the Nkoya of Western Zambia, cults like those of the Ndembu are fragmented into numerous, small and unstable, localized factions, each with its own leader, distinct songs, and peculiar paraphernalia. Not that van Binsbergen regards merely the politics and economic transactions in the cults; he is careful to bring into perspective also the pull of belief and liturgy in the healing ritual.

The next step van Binsbergen takes follows from his historical approach. He distinguishes types of cult of affliction and examines their development. Already familiar is the non-regional type, which covers the Ndembu examples also: the cult has its pecular emblems, its own highly specific idioms and subjects, and it is widespread, perhaps with many local variations, among a multitude of culturally different people. What it lacks is any central place, hierarchy, or overall organizational focus. Only the regional cult of affliction has these features along with the others and, significantly, even more of a conceptual emphasis on inclusiveness, with a shift towards greater attention to a High God. To explain the transformation, van Binsbergen traces a creative process of religious innovation, with its taproot continuously drawing on (or sloughing away) the transcultural, general idiom of the earlier, but still current, non-regional cults. He rejects an ‘imitation theory’, which reduces the essential dimension of the change, and makes it depend simply on the existence of a Christian church as an organizational model to be copied. Moreover, he shows that innovation itself becomes a proof of spiritual grace: whereas followers of any leader in a non-regional cult regard him as a mere transmitter of a received idiom, the regional cult’s members attribute its basic design to their founder’s prophetic visions and insights. In fact, much of the cult innovation is on a pragmatic, ad hoc basis. It has to be understood as a gradual series of choices in response to growing acceptance or rejection of the cult by a wider public. Furthermore, it is not an evolution in a single direction, and a cult may lapse from the newer type into the older one and become non-regional.

How does one cult — and not another — firmly establish a hierarchy, with secure centralized control of cult assets and services, while expanding the regional catchment area over town and country and the span of cult membership across socio-economic strata and ethnic groups? Detailed historical comparison is necessary for the explanation, and van Binsbergen compares the career of one cult, formerly regional, with another that is now vast, multi-ethnic and still expanding not only in Zambia but in Botswana also: respectively, these are the Bituma and Nzila cults. The relative placement and mobility of the founder’s own group counts along, of course, with the founder’s capacity to mobilize his fellows in it. Nzila’s founder initially appealed to Luvale immigrants like himself and even beyond them to others who also came in a great influx from Angola into an area of Zambia with towns of major economic and political importance. With the increasing penetration of the immigrants went the greater spread of the cult. At first, the diagnosis and final treatment of member’s afflictions remained (with the possible exception of the limited autonomy allowed a single deputy and branch in another district) the monopoly of the founder and his designated heir. They moved the cult’s main shrine to a central place right by one of the two major towns and thus secured a suitable headquarters for urban as well as rural supplicants. Eventually, other branches were also allowed some autonomy in healing, but central control was stabilized by the careful registration of all members, ‘by the Annual Convention at the cult’s headquarters, by the [centralised] distribution of essential paraphernalia, and by formal examinations concerning the cult’s beliefs and regulations’. Not only did this result in an idiom quite distinct from that of any non-regional cult of affliction — and thus made it harder for the new cult to lapse into the old cultic background — but it also established, van Binsbergen argues, ‘clearly recognised principles of legitimation, controlled from headquarters, outside which no cult leader pursuing the Nzila idiom could claim ritual efficacy’.

A more illuminating contrast could hardly be found than the career of the Bituma cult. Having failed from the start to convert many of his fellow Angolans, the founder, himself an Mbunda, concentrated his appeal on rural villagers, mainly Nkoya. His cult became trapped in their relative poverty, their hinterland’s local power-blocks, and the limitations of their labour migration to distant towns where they never gained much of a foothold. Anchorage in their district stultified the further expansion of the cult’s region. The district had only a small township as its administrative centre and was ethnically more homogeneous as well as economically less differentiated than the breeding ground of the Nzila cult. When Bituma’s founder died, the cult fragmented; it became little more than other non-regional cults. Never did it develop, in van Binsbergen’s terms, ‘an adequate structure through which the founder’s charisma could be routinised and channelled without becoming dissipated or usurped’.

In both cases van Binsbergen sees a constant factor as foremost: the structural characteristics of the immediate arena in which the cult arises. His argument is that the viability of the cult is contingent upon these characteristics above all and that the structure of the cult has to be commensurate with them.

It is an argument which Thoden van Velzen takes a stage further in seeking a materialist explanation of the waxing and waning of regional cults in Surinam. Just as waves of cults of affliction sweep across Central Africa, so do High God cults come and go in Surinam among Bush Negroes and, at times, among Creoles also. Like van Binsbergen, Thoden van Velzen considers it essential, therefore, that the cultic background out of which a regional cult emerges and into which it lapses be given its full weight in analysis. But for Thoden van Velzen even more problematic, and thus more central in his argument, is the accumulation by the cult of valuables or assets: How does such accumulation affect the survival and raison d’etre of a cult hierarchy and, even more fundamentally, how does the overall religious transformation relate to change in modes of production and distribution? His argument also contributes to other, perhaps even more general debates, because his main case is about a cult, Gaan Tata, which invites comparison with cargo cults, that favourite concern in dispute about the rationality or irrationality of religious belief and practice. Gaan Tata is, in a sense, a cargo cult in reverse. In it, too, the moral redemption of men, purity, cleansing from sin, all depended on waste, and the waste was on a grand scale: great heaps of European trade goods — household utensils, Victorian bric-a-brac, mantlepiece decorations, gold chains, shoes, plastic ware — were left to rot in dedication to the High God (Gaan Tata). The sudden, unexpected enrichment of the Surinam interior’s boat entrepreneurs, due to the discovery of gold which brought with it a deluge of imported goods quite without precedent, gave the cult its initial opportunity and basic concern. It was not that the cult demanded, as some millenarian movements do (cf. Douglas 1970 : 137), that people reject their own material values or those of Europeans. The wasted goods were quite highly valued imports, as in a cargo cult. For that very reason, they became God’s business: like other new interlocal relations, such new riches, to be made in river transport across communal boundaries, were not to be left to look after themselves, but were made sense of, and controlled, religiously.

If, as Douglas suggests, cargo cults are ‘cults of revolt against the way the social structure seems to be working’ (1970 : 139), then we may say that in reverse what we find is a ‘cult of regularization’ (cf. also Moore 1975 : 219f). A cult of regularization enhances economic inequalities, rather than levels them; and it legitimizes a new status quo, rather than a restoration of the old. Gaan Tata is such a cult. It hid its waste, from outsiders especially, in a secret bush shrine. It reduced the circulation of European material goods and limited them even more to an elite, rather than providing, after the fashion of a cargo cult, the ritual means to bring the goods flowing in from the outside. Moreover, it redirected the circulation away from the main European centre, Surinam’s capital, on the coast and towards the cult’s own headquarters in the interior. Fittingly for a cult of regularization, its headquarters was sited at a centre of current and continuing political importance, the capital of the Paramount Chief (Gaanman) of the largest Bush Negro tribe.

Thoden van Velzen points out:

‘As part of Gaan Tata’s commandments, the [boat] entrepreneurs were given a set of standards that proved their usefulness in contacts between Bush Negroes from various areas and with different cultural backgrounds; the cult, for example, gave religious sanction to peacefulness: even the shedding of a single drop of blood could arouse Gaan Tata’s wrath ... Significantly, any boatsman who established a good relation with the Gaan Tata oracles through timely presents and acts of deference — soliciting help in case of illness — could rest assured that his reputation would be protected [from accusations of witchcraft]’.

He argues further:

‘The cult was directed at eradicating an internal evil, witchcraft, that was felt to dog the steps of the nouveaux riches’.

Hence the cult introduced an alternative to the customary conciliatory ritual for the dead: one treatment for the witches and sinners, another for the “respectable”. Sinners or witches were believed to die from Gaan Tata’s wrath, so no elaborate death rites for them. Their legacies, polluting to their relatives, were brought as “God’s cargoes” to the cult’s priests, who gave back a small, cleansed share, kept part for themselves, and left great heaps dedicated to Gaan Tata. The relatives and local community made a death the occasion for costly celebration among themselves — the customary death rites — only when someone was declared, after a post mortem inquest, to be “respectable”.

Thoden van Velzen argues that the career of the cult was closely geared to radical economic change introduced from outside the cult region. The boom in river transport entrepreneurship produced along with wealth a qualitative change in interlocal and cross-cultural relations which was essential for the cult’s overall centralization and effective hierarchy. Then came a drastic economic slump when the interior’s extractive industries, gold and Balata, collapsed. Not long afterwards the cult itself fragmented into a number of parochial cults, without any central control or uniform ritual practice and dogma. The cult’s career, until the next radical economic change, took the course of what might be described as a downwards spiral. The more the priests retrenched through short term solutions to their own pressing religious and organizational problems, the less was the cult able to respond to the current concerns of its public. Once a new labour elite emerged with the next radical economic change, a different regional cult began in opposition to the remnants of the old one. The opposition was based in a rival political centre within the old cult’s heartland, and the new cult, like the old in its heyday, regularized and legitimized the wax of inequality. Protest and revolt there was, perhaps, but in defense of new privilege: once again a fresh cult gave the newly prosperous members of a rising labour elite protection against witchcraft and also a fresh imprimatur in the form of special certificates to their own innocence of witchcraft.

A key point must be kept in sight, given the stress in Thoden van Velzen’s argument on radical economic change. Yesterday’s cult of regularization becomes today’s anachronism, not merely through external change or radical discontinuities in the milieu of the cult but through cumulative change in the cult itself. Indeed, every regional cult builds up a momentum of its own over time which has to be taken into account in explaining its response to radical economic change or any challenge to the viability of its regional organization. That is why in my own study of a High God cult, southern Africa’s cult of Mwali, I pay special attention, as does Thoden van Velzen, to the underlying orientation of the cult’s leadership and to endemic instability within its organization. The question is, How far does this orientation alter over time, and what directions does such instability take regularly as well as in sudden, perhaps unprecedented shifts?

Historical evidence and my own observation shows that from precolonial times to the present there has been a continuity in orientation within the cult of Mwali. The central concern of the messages from the cult’s oracles was and still is to conserve order in the world and maintain the welfare of the land, its people and their economy. It is not surprising that such conservatism is fostered by a cult with a certain, lasting fixity in its hierarchical centralization. What is surprising is the combination of such fixity with instability in the coverage and definition of regions, impermanence in the sites of oracles, variability in priestly succession and great fluctuations in affiliated congregations and their staff. On the one hand, this High God cult has a small, limited number of regional oracles that are ranked in a stabilized, unified order according to their restricted distribution within a central heartland. Similarly, the ranks of its staff form a hierarchy with relatively exclusive criteria for recruitment: the higher the rank, the narrower and more exclusive are the criteria, and recruitment to the inner circle of priests is on the most exclusive basis. On the other hand, its land shrines lack an overall pecking order; they are widely dispersed and highly variable in standing, a multitude without a known size. Like their congregations, they rise and fall in importance as foci of ritual collaboration (some disappear completely), and pass through what I regard as three grades — local, interlocal, and regional.

Significantly, this cult is one of those to which Turner seems to refer when he suggests that earth cults stress inclusiveness, rather than ‘selfish and sectional interests and conflict over them’ (Turner 1974: 185). Yet the cult fits no polar extreme in Turner’s terms. Admittedly, never does its ritual offer a straightforward representation of the various political divisions within its vast domain (see Map 1). But it does thrive on indexical occasions — on the very stuff of micropolitics — when the relations between competitors are put to the test of ritual collaboration or avoidance, when the scale of local commitment and investment in the cult is proved, when still unresolved territorial and political ambiguities are given a somewhat anticipatory definition, heralding new power divisions. At the same time, its ritual, characteristically rich in symbolism with transcultural meaning, memorializes events that are broadly significant for every congregation, be it mainly Kalanga, Karanga, Venda, Ndau, Khurutse, Ndebele or any other. Moreover, its theology directs attention to the macrocosm, the order beyond that of the congregation or any single community. I show in my essay that the cult’s macrocosmic conceptions ‘are a significant factor in behaviour because they are of such a kind that communities can continue to define their broadest consensus through them irrespective of their differences, hostilities and competition’. With its capacity to override or encompass quite marked cultural differences, the cult continually admits newcomers irrespective of their ethnic origins. Thus highly sensitive and responsive as the cult is to micro-political change, it is no less concerned with ‘wider bonding ... disinterestedness and shared values’ (Turner loc.cit.) Indeed, it is dynamic tension between inclusiveness and exclusiveness which gives the cult much of its momentum and renewed viability from one historical context to another.

It may be, as Turner suggests, that studies of African cults of the earth, most strikingly the Tallensi case, tend to have to say little about ‘factional conflict’ (idem). But the cult of Mwali can hardly be studied adequately without a discussion of politics and policy making in various arenas, from the most petty to one grand enough to embrace the cult’s vast domain across southern Africa. Indeed, it has been much debated in the literature on the cult, though mainly from the viewpoints of white settlers, how the cult responded in crises, especially during an early war against white rule in 1896. I suggest that, even during crises or rather then most of all, the cult has been constrained to favour political compromise in its widest policy — its cardinal oracles make pronouncements which do not commit the whole cult to one side only during a struggle — and that this balancing has been in response to the diversity of peoples and interests which the cult must encompass. In no way does this suggestion deny the cult its part in rallying moral sentiment against exploitation and the abuse of power: some of its songs ring with moral outrage against alien rule; its oracles have urged resistance against the inroads of a cash economy; they have also warned about the dangers of party politics. Rather this suggestion takes account of the fact that the cult’s ultimate dependence is on voluntary consent; its appeal is to the most general consensus, which it may focus but cannot invent.

A two-way process of politics has taken place. The cult has had its impact on changes in the distribution of power within various arenas and so, too, have nationwide conflicts and factional struggles, both within and between local communities and Christian churches, contributed to political competition within the cult itself. To demonstrate this point in detail part of my discussion concentrates on one region, its development, its leaders’ careers, their disputes, managerial problems and tactics. Necessarily, territorial encroachment and the expansion of political communities and churches figure prominently in my analysis, because individual and communal migration, in part due to the appropriation of the central highlands of Botswana and Zimbabwe for European ranches, has been a considerable factor in the region’s politics, contributing to a growing shortage of land and increasingly diversifying the region ethnically. Similarly weighted analytically are shifts in the flow of transactions, because these have affected the control of assets and the accumulation of great wealth by the region’s leaders, especially its priest and his heirs. Furthermore, I show why it has been critically important to have a secure base in a political community on the wax in order to command a centre for the region as a whole. It would be a gross error to turn the region’s sequences of political competition into a series of successful manipulations by Machiavellian masterminds in collusion with each other, or on the other hand, to exaggerate the ad hoc element in decision making in the cult. My argument avoids both these pitfalls by establishing how regularities in the sequences relate to competitors’ long and short term objectives as well as to the highly pragmatic, situational adjustments which they make without any grand design in view.

The Interpenetration of Cults

Through an examination of how different kinds of cult interpenetrate in Zimbabwe, Garbett reaches a closely related argument. Garbett rejects the view that the ritual field formed by the interpenetration of the cults can be understood as extensively controllable or manipulable by a single focus of ritual authority, such as one spirit medium. He argues,

‘Only within limited local arenas is it meaningful to speak of strategy ... The appearance of strategy, when the total ritual field is viewed over time, arises not from the manipulations of particular mediums, coordinated in terms of some overall plan — the key actors rarely if ever meet and do not collude — but from the emergent properties of the ritual field itself’ (Garbett’s italics).

He is careful to point out, also, that methodological problems arise — problems which come up repeatedly in our studies of regional cults — because the vastness of this ritual field puts it well beyond the practical reach of a single fieldworker. He regards the ritual field, therefore, from the perspective of the Korekore cult which he has studied in depth in Zimbabwe, and relies on historical evidence and anthropological studies of linked cults in order to contextualize his own observations.

The processes he analyses link a regional cult with quite different, noncentralized cults which overlap from one part of the ritual field to another. These cults, so different in organization and ideology, might loosely be categorized as earth cults: they are broadly concerned with general morality and man’s dependence on the earth, and they receive supplicants from wider vicinities extending beyond individual political communities. However, only the clearly centralized regional cult has ranked mediums in a recognized hierarchy with jurisdiction over well-defined ritual territories. Its sacred topography, while changeable and flexible like certain other aspects of the cult — and Garbett stresses that some aspects are even more flexible and unstable than these — is regarded by the people themselves as enduring and unchanging from mythic time, not an expression of current political boundaries. The whole landscape, ritually divided and sub-divided into ‘spirit provinces’, is full of fine boundaries ‘established’ in mythic time, though they may be actually ‘discovered’ by spirit mediums quite recently. By contrast, the most widespead of the linked cults, the Zezuru local cult, is more like the Tonga prophet cult, lacking demarcated constituencies and with unranked but graded mediums whose clientele varies accordingly: the minor medium serves the local order, clients from his immediate community or kin, whereas the major medium has a wider clientele, dispersed and more diverse (cf. Fry 1976: 34f, in his terms the Zezuru mediums are ‘lower-level’ and ‘higher-level’, respectively). Moreover, like Tonga and other Zambian prophets, the Zezuru ones have recently multiplied in droves, with a striking exaggeration of the familiar pattern of meteoric careers for many. Here the resemblances, like certain crucial differences, are not fortuitous or merely matters of typological comparison: they are evidence of the properties of a wider ritual field, and continuities and discontinuities within it. Taking the Korekore cult as the field’s centre, it extends from a centralized regional cult to its periphery in non-centralized or weakly centralized cults, respectively the Zezuru to the south and the Tonga to the north (see Map 1). Similarly, across the field, there is a correlated variation in the conceptual or symbolic elaboration of hierarchy relative to the Korekore cult. At the periphery the elaboration is somewhat less among Tonga than Zezuru, who make a conceptual analogy between their political hierarchy — village headman, subchief, chief — and their pattern of spiritual authority — near ancestors, remote ancestors, heroes — although the mediums possessed by these spirits may be competitors with mutually disputed and ambiguous standing. Among Zezuru yesterday’s disciple is often in the process of becoming today’s rival and perhaps even tomorrow’s master:

‘Although the system is thought of in stable hierarchical terms, and although there is a certain stability of authority relations between spirit-mediums, the pattern of relationships between the spirits and their mediums at any one point of time depends more on the abilities of the mediums than on anything else’ (Fry 1976 : 44–45).

The relations between the disparate cults are unstable as are the cults themselves, in certain respects. Yet the processes in the wider ritual field cannot be understood as merely changeless oscillations between hierarchy and non-hierarchy or centralization and non-centralization. Garbett rejects the all-too familiar pendulum model in order to comprehend the changing interdependence between the disparate cults. To appreciate the advance he makes it is useful to contrast his approach with that of Fry (1976) who also discusses the Zezuru and Korekore cults but from an alternative vantage-ground, Fry’s own, sensitive observations of the Zuzuru. As Fry sees it, each cult corresponds to its society; one society being the opposite of the other, the cults are opposites, too. The Zezuru are ‘dynamic and radical’ ‘in the front line of “progress” ’ and ‘cultural nationalism’ (Fry 1976 : 66); correspondingly their cult is highly flexible:

‘While the Zezuru experience rapid change, and unpredictable future, rapid economic or political success on the basis of personal charisma ... so their religious life features dynamic charismatic spirit-mediums and a highly fluid and little-controlled pantheon of spirits’ (Fry 1976 : 66–67).

Fry is aware that the contrast has an element of caricature in it, but he pursues it nevertheless:

‘It is perhaps an overstatement to regard the Korekore as a stagnant society, living in the past, but it would nevertheless go some way to explaining the ‘bureaucratisation’ of [their] spirit-medium cult as part of the running down process. The mediums perpetuate the glorious past, which is changeless and as rigid as the cult organisation itself’ (1976 : 66).

The trouble with such a picture of ‘bureaucratisation’ and ‘rigidity’ is that it selects features, somewhat arbitrarily, out of context and so gets them out of proportion. Hence the Korekore cult appears to be hierarchical and stable beyond belief. Partly, Fry gets it wrong, because from his grounding in the Zezuru cult (whose mediums have a clientele without an apparent or well-defined constituency), Fry sees as the sole alternative a cult having mediums with constituencies only. In fact, however, the Korekore mediums have both constituency and clientele, and it is essential to recognize the area of indeterminacy in the relations between constituency and clientele. Thus Garbett observes:

‘A medium who establishes a reputation for effective cures and for bringing rain may draw supplicants [i.e. a clientele] to him from a wider vicinity that extends far beyond the boundaries of his spirit province [his constituency] or even his region ... Some mediums, by expanding their vicinity of supplicants, come to have reputations which far outweigh their, possibly, humble positions in the formal hierarchy of mediums. Ultimately, this may lead to friction between a junior medium and one more senior about their respective jurisdiction and spheres of ritual authority, and the intervention of the senior medium may be required’.

Quite unlike the caricature, the Korekore cult in reality, Garbett shows, emerges through the mediums’ lobbying for consensus, through their competition for backing among themselves, for greater reputation with followers, and through their continuous reconstruction of the mythic past in the light of their understandings of the present and their aspirations for the future. The ‘bureaucratic’ Korekore mediums turn out to be ‘charismatic’, ‘dynamic’, after all; and to demonstrate this point Garbett enables us to follow closely a regional medium’s behavour in the critical context of the seance, because this is the context par excellence for development in the cult. It is where ritual authority is proved or disproved, where displays of ritual knowledge and competence put power and authority at risk. In his analysis Garbet also gives full weight to transactions over material resources and tangible cult assets. Indeed, he evaluates the critical factors in a pattern of expansion through his detailed case material on the regional medium of Mutota’s spirit, George Kupara. Kupara abided by the cult’s doctrines discouraging personal enrichment at the expense of the wealth in trust to the spirit. His aim was not accumulation in the spirit’s name, but rather largesse which enhanced the spirit’s reputation along with his own, of course, and thus contributed to the continuing expansion of his region and its vicinity. Such expansion thrived on the redistribution and recirculation of cult income and the conversion of cult fees into debts or loans subject to interest charges that were ‘quickly forgotten’: i.e., became the halo of good-will needed for wider ritual authority.

So far I have barely touched upon the subject of the interdependence between the disparate cults. Yet it is this part of Garbett’s study which does most to advance our understanding of the wider ritual field. The problems for the people themselves, like the anthropologist, are in good measure conceptual and demand that they, too, make a translation of culture. Seen from a Korekore perspective, non-hierarchical relations among the Zezuru have to be comprehended by and through the hierarchical cult among the Korekore. One of the ways that it is done is through a shared genealogical idiom which allows the establishment of kinship and filiation between the Zezuru local spirits and the Korekore territorial ones and thus relationships between their mediums. Garbett draws our attention to the distinctive configuration that the genealogy of the Korekore ancestor Mutota has: in graphic terms, it can be roughly described as a ‘T’ shape rather than the inverted ‘V’ the shape so familiar in the genealogies of segmentary lineages. The apical ancestor is Mutota, and his ‘children’ (the top of the ‘T’ ) are not only virtually without issue but also somewhat indeterminate genealogically — or rather, what their number and birth order may be is ignored as a subject of cultural knowledge. Hence there is leeway for the inclusion of more or less spirits and thus mediums, without definition of their rank or seniority vis-a-vis each other: they are all ‘children’, peers alike subordinate to the one ‘father’, Mutota. As for the rest of the genealogy, after a fixed line of names, which a candidate for mediumship has to cite as a proof of competence and legitimacy, the rest does define hierarchy and rank but for the mediums of the Korekore themselves. Thus with one simple conceptual framework Korekore are able to comprehend cultural and transcultural relations, because that \framework includes both areas of ambiguity and areas of apparent definition of the rank and limits to the inclusion of spirits and their mediums.

His discussion of the various ways that the people themselves bridge the disparate cults by making translations from one culture to another leads Garbett to a further question. Under what conditions do mediums or would-be mediums of the non-hierarchical local cult turn for legitimation to the regional medium of the hierarchical cult? And what consequences for stability or instability follow due to the reaction of each cult on the other? It is a measure of the force of his answers that he takes into account both the long-term properties of the wider .ritual field and the most recent transformations in the last phase of colonialism.

Future Research

These essays begin a new phase in the anthropological study of regional cults. It is the first time that the cults are treated as a major focus of comparative analysis. Hence much of our work is devoted to the tightly argued presentation of a substantial body of evidence on a limited number of main cases, rather than on any broad survey of the subject. We are well aware that some of our generalizations are tentative, as befits the initial stage of enquiry into a vast subject. Our aim is to provide a major body of hypotheses, to raise fresh questions even beyond our immediate reach in answers, and thus to prepare the ground for future research into change in religious organization and ideology.

University of Manchester

January 1977

References

© Richard P Werbner 1977

Acknowledgements

Max Gluckman was a creative force in the A.S.A., as a Founder Member, Secretary (1951–57), and Chairman (1962–66). He took a special interest in the A.S.A. Monographs, and began the first series with the introduction which he and Fred Eggan wrote. With characteristic energy, he helped us in our early efforts to plan this conference, shortly before he died in 1975. We dedicate this book to his memory as a tribute to his inspiration.

This book is based on the 1976 A.S.A. Conference which Kingsley Garbett and I convened at Langdale Hall in the University of Manchester. I am grateful to him for sharing the burdens of the Conference. We had planned to edit the book together. However, he took up a Readership in Adelaide after the Conference, and we decided that having two editors so far apart would mean delaying the book beyond the year we had fixed for publication.

The A.S.A. wishes to thank warmly the Warden of Langdale Hall and her staff for providing so perfect a home for the Conference. We are grateful also to the University of Manchester and the Wenner Gren Foundation for financial help.

The chairpersons of the Conference’s sessions were: Jean La Fontaine, Mary Douglas, Ernest Gellner, Paul Baxter and David Parkin; the discussants were Malcolm Ruel and Hector Blackhurst along with the conveners. We would like to thank them and others, especially Terence Ranger, Raymond Firth, Peter Fry, Benson Saler and Jaap Van Velsen, for their helpful comments. For much practical help at the Conference we owe a very great debt to Mary Lea, Trudy Armstrong, Nick Mahoney, Beth Goldblatt, Bob Fredericks, and Jay Close.

Our present understanding of regional cults is the product of much joint effort and intellectual cooperation by all the contributors to the book. As editor, I am grateful for the opportunity it has given me to work from the vast darkness of the subject towards the light.

Finally, I wish to acknowledge the creative work of members of the Audio-Visual Service of Manchester University: Paul Bailey designed the book, and Sue Frost carried out the studio production. Summer Harrison, Paul McCartney, and Derek Schofield also gave last-minute assistance.

The royalties on the sales of this monograph go entirely into the Radcliffe-Brown Memorial Fund.

Part One: Islamic Cult Regions: Historical Interaction or Symbolic Congruence

1. Ideological Change and Regional Cults: Maraboutism and Ties of “Closeness” in Western Morocco

by Dale F. Eickelman

This study presents an account of the main structure of the ties between the Sherqawi religious lodge (zawya) of western Morocco and its clientele, the idealogical base for the maintenance of these ties, and how both the actual ties and the values on which they are based have altered and, in turn, have been altered by changing historical and political contexts.[1] Through these topics I pursue a more general concern with the classic sociological theme of the interrelation between symbolic representations of the world and patterns of social conduct. In particular, I am concerned with maraboutism, a predominant (although not unique) version of Islam as locally received in Morocco. I argue that like most implicit ideologies (or in a sense that is explained shortly, a part-ideology), maraboutism is closely tied to basic Moroccan conceptions of the social order, notably the notion here translated as “closeness.” I further argue that such ideologies must be understood as socially maintained activities and related to specific historical contexts. I seek to place maraboutism in such a perspective and to indicate how this approach is more satisfactory than alternative ones which treat maraboutism as analytically “timeless” or which posit a stable interrelation between the realms of ideology and social action.

Ideologies and the Social Order

Because the relation between symbolic representations of the social order and patterns of social conduct is such an enduring sociological problem, I think it useful to indicate what I regard as the shortcomings of certain earlier approaches. An older academic tradition asserted a correspondence between the two realms without regarding the exact nature of their interrelation as problematic. Thus Robertson Smith (1919) related successive modifications in the messages of Old Testament prophets to the specific historical circumstances in which they originated. Smith clearly perceived both social changes and inconsistencies in the prophetic ideologies, yet regarded Victorian notions of progress and Christian assumptions about the gradual enlightenment of mankind as sufficient to account for them (Beidelman 1974: 38). Similarly, the “simple” societies that Durkheim and his collaborators chose to investigate allowed them to presume an elegant one-to-one correlation between ideology and social action (Durkheim 1915; Mauss 1966; see also Evans-Pritchard 1940). Edward Shils is a contemporary representative of the “correspondence” tradition to the extent that he regards the core symbols, values, and beliefs of a society as concretely represented by such physical manifestations as capitols, cult centres, and social and political boundaries (Shils 1975: 3).

In some contexts the straightforward postulate of an at least partial reflection between the two realms still possesses analytical utility. In Shils’ case it is applied in an original way to the symbolism which underlies modern nation states. Yet, in general, it is now competent but sterile to reiterate that ideologies sometimes support actual social arrangements and vice versa. Or, in the less hyperbolic language of Sally Falk Moore, one consequence of accepting such a straightforward correspondence between the two realms is to imply that “social arrangements” are “simply imperfect approximations of ideal models” (1969: 400).

The notion of correspondence masks the two-way interaction which occurs between symbolic conceptions of the social order and patterns of social action. The idea of interaction, as opposed to correspondence, necessarily implies a lack of fit between the two analytical levels at any given moment. Relatively few anthropologists have sought to trace out in concrete detail the consequences of such interaction. Those who have tend to be concerned with only one direction of it, the manipulation of ideological principles at the level of social action. Below I briefly analyze one such approach because it places my ensuing discussion in a more meaningful perspective. In an excellent article entitled “Descent and Legal Fiction” (1969), Moore argues that systems of descent, legal systems, and religious ideologies describe and explain the nature of the social universe. At the same time they provide a means for manipulating it. To illustrate her general argument, she argues that any descent system should be seen “as an ideology of identities that can be adapted for use as an organizing principle, rather than as an organizing principle in the first place” (Moore 1969:381). When adapted as an organizing principle, descent ideology or any other system of symbolic representations can vary considerably as it is manipulated for practical ends. Even when jural rights are ideologically claimed to emanate from a single principle such as descent, they usually derive from several and cannot effectively be traced to a single source (Moore 1969: 396).

In anthropological analyses which are concerned primarily with the ramifications of symbol systems as they serve as practical guides to and rationalizations for social action, there is a tendency to simplify the argument by treating symbol systems analytically as “timeless.” Occasionally this is made explicit (e.g., Schneider and Smith 1973: 6). A consequence of this approach is that what might be regarded as the communicative aspects of the relation of symbol systems to social action remain unstressed. After all, ideological systems are social activities that are maintained through various forms of expression, including ritual action. In the course of being expressed, such ideologies become modified and in turn modify the direction of social action. Thus, to be fully comprehended in relation to social action, such ideological systems must be seen in at least some of their successive historical modifications.

There is a closely related issue which it is appropriate to introduce here. Moore (1969: 400) implies that most ideological systems are analogous to the descent systems which directly concern her. As difficult as the analysis of descent systems might appear in some respects, I would argue that from an analytic point of view some ideologies, especially those considered religious, are often less distinctly bounded. Thus in most world religious traditions there is an inherent tension between formal ideological tenets propagated and accepted by an educated, religious elite and, coexisting with them, an implicit ideology of religion as locally practised and understood. The latter substantially overlaps with equally implicit, locally held assumptions about the nature of the social world. From an analytic point of view it is difficult to categorize such beliefs as “religious” and as Runciman has indicated, labelling such beliefs as ideology or simply as beliefs about the conduct of life is just as adequate (1970: 75–76). Indeed, from the viewpoint of the social elite in most religious traditions, the religious practices and attitudes of subordinate groups or persons frequently are considered uncivilized, pathological, or, in the case which concerns this article, un-Islamic. Yet for those who maintain such beliefs, they usually are taken for granted as authentic expressions of the religious tradition in which they participate. North African maraboutism is one such ideology. Precisely because it appears so closely tied to aspects of local society, maraboutism affords a particularly useful means of tracing the interaction between ideology and the social order.

Maraboutism as a Part-Ideology

Since the terms marabout[2] and maraboutism have been used in a variety of ways, I specify my own usage below and then indicate how marabouts form a part of Islam as locally understood. In the Moroccan context, marabouts are persons living or dead (dead, that is, only from an outside observer’s point of view) thought to have a special relation toward God which makes them particularly well placed to serve as intermediaries with the supernatural and to communicate God’s grace (Baraka) to their clients. Maraboutism is the usually implicit ideology and the ritual activities associated with such beliefs. Like any successful ideology, maraboutism and the ritual comportment associated with it have been flexible enough to accommodate changing historical situations. Marabouts and similar holy men elsewhere in the Muslim world are seen by their supporters as elements in a particularistic hierarchy through whom the supernatural pervades, sustains, and affects the universe. In a Weberian sense such a conception might be seen as a “magical” accretion to Islam; but from the viewpoint of tribesmen, peasants, and townsmen who hold such beliefs and act upon them, intermediaries with God are an integral part of Islam as they understand it. They regard themselves as Muslims, pure and simple, although they are aware of the disfavour in which such beliefs are held by Muslim scripturalists and reformists. Consequently, clients of marabouts often dissimulate their beliefs in front of such persons, especially in contemporary contexts.

Now and in the past, those who express and accept maraboutism are consciously aware of an alternative, formally articulated vision of Islam antithetical to their own. For this reason, maraboutism can be considered a part-ideology. It is never found in isolation, but co-exists with formal Islamic ideology. Further, it largely overlaps with locally maintained assumptions concerning the social order. Because of its implicit patterning upon (and patterning of) existing social realities, maraboutism can be considered a conservative ideology in Mannheim’s sense of the term (1971: 152–153). It does not have to be consciously articulated and defended in order to be maintained since it takes as its legitimation “the way things are,” to use a popular Moroccan expression to refer to the social present. This implicitness undoubtedly is one of the reasons why there have been relatively few discussions of maraboutic ideology in North African ethnographic literature, as opposed to police-dossier style accounts of the leadership, genealogical claims and mystic ties, and regions of influence of maraboutic orders (e.g., Depont and Coppolani 1897; Drague 1951).

The conjuncture of conflicting patterns of belief can be found in most world religions. Along these lines, Ernst Troeltsch (1960) has argued that Christianity can be seen as constantly in tension with social reality. Consequently its social history is one of shifts at various levels of compromise and noncompromise with the world. I think that a similar argument is critical to comprehending maraboutic beliefs. The Islamic tradition as it has emerged in various societies has generally contained conceptions of man’s relations with God both as comprising and lacking intermediaries. The Qur’an, as scripturalists and Muslim reformists vigorously stress, portrays all men as equal before God with no privileged intermediaries toward him.

In contrast, the strength of maraboutism rests on the implicit assumption that relations between men and God work just like all other social ties. Moroccans who act upon maraboutic beliefs explicitly see the relation of God with his marabouts as nearly analogous to patron-client ties in ordinary society. As one informant related, God, like a minister or the king, is too powerful to be approached directly, so the client works through a marabout who is “close” both to him and to God (Eickelman 1976: 161–162).[3]

Taken together in various syntheses according to historical context, the antithetical notions of formal Islam and maraboutism (or similar part-ideologies elsewhere in the Muslim world) have enabled Islam, analytically considered as a religious tradition, to encompass many varieties of social experience and to be regarded by its adherents at any given time as a meaningful religious representation of reality.

The Sherqawi Zawya

Since maraboutism is largely tied to specific social contexts, here I briefly sketch the Sherqawi religious lodge’s social history and the changing scope of its activities before considering the maraboutic ideology and cult associated with it. Asa major regional pilgrimage centre, the Sherqawi zawya of Boujad annually draws tens of thousands of pilgrims, singly and in groups, and currently can claim as clients most of the tribal collectivities of the western plains (see Map 1), many individuals from nearby towns, and scattered ones from elsewhere in Morocco. Approximately one-third of the town’s inhabitants, roughly twenty thousand as of 1970, claim Sherquawi descent; continuing Sherqawi dominance of the town’s economy and administrative offices gives Boujad the flavour of a “company” town. This dominance has levelled off, but the town is still primarily identified with the zawya.

The religious lodge was founded at the end of the sixteenth century by the marabout Sidi Mhammed Sherqi (d. 1601). Within a century it became a leading intellectual centre in Morocco. Through a series of shrewd political moves, the successive lords (Sid-s) of the zawya managed to enhance its influence in a tumultuous political climate. Hagiographies and court histories of this early period give little indication of the relation of the Sherqawa with their clientele or their internal rivalries, but by the end of the 18^th century its prestige was such that Sultan Muhammed ben ‘Abdallah (reigned 1757–1790) feared its prominence and destroyed most of the town, including a collective granary used by the region’s client tribes. The Sultan also led the principal Sherqawi marabout of the time, Sid i-Hajj l-‘Arbi (d. 1819), to forced residence with the royal entourage in Marrakesh. The marabout used the occasion to accept as donations extensive agricultural estates (‘azib-s) from tribal collectivities in the south (see Map I) and to extend the Sherqawi sphere of influence (Eickelman 1972–1973: 44). By the early 19^th century Sid 1-Hajj l-‘Arbi was considered one of the two most powerful marabouts in Morocco (Ali Bey 1816:1, 176–177).

A fuller picture of the activities of the zawya emerges by the latter half of the 19^th century. Leading Sherqawa were able to influence the sultan and his entourage and in turn were considered significant enough to be watched, accommodated, and carefully manipulated by them. The Sherqawa exercised intermittently a de facto control over Makhzen (central government) appointments in the area, acted as intermediaries to the Makhzen on behalf of clients, secured the safe passage of commerce through the area, mediated tribal disputes, and performed a myriad of other roles. Clients from Rabat, Sale, Fes, Meknes, and the Tafilalt area to the south regularly sent offerings to leading Sherqawi marabouts, as did rural local communities (dawwar-s) and tribal sections (fakhda-s). Most collective client relations with the Sherqawa were in the area indicated on the map, but there were also a few groups as far south as the Sus and in the High Atlas mountains. These links, as with those in the main sphere of Sherqawi influence, depend largely upon the social landscape as perceived by the Sherqawa and their clientele at any given moment, as in the case of the lands acquired by Sid l-‘Arbi (d. 1819). His descendents still possess these lands and maintain ties with clients in the area.[4]

Client allegiances frequently shifted among various internal Sherqawi factions and occasionally away from them entirely. The vicissitudes of Sherqawi agricultural estates and subsidiary religious lodges, especially those located on the Shawya plain (for which documentation is more abundant), indicate how fragile and shifting was the balance of power among competing persons and factions. The sultan, his entourage, and local Makhzen officials often were at odds with each other. Each calculated the possibilities and advantages to themselves of enhancing or detracting from Sherqawi prestige, depending upon their estimation of interests at any given moment. Rival Sherqawa, merchants, rural strongmen, men of learning (‘ulama), and religious lodges elsewhere in Morocco made similar calculations of their own interests, the strength of the Sherqawa (or more accurately, of individual dominant Sherqawa), and the immediate opportunities available to them (Eickelman 1976: 31–64).

Growing French influence in the early 20^th century at first added only another component to this multicentered shifting of alliances, but the social and political roles of the Sherqawa abruptly altered with the imposition of the protectorate in 1912. Prominent Sherqawa lost many of their preprotectorate privileges but were given local administrative sinecures, preferential access to French educational facilities, and favourable treatment as notables in administrative matters. The bases for social honour (in Weber’s sense of the term) rapidly shifted. For instance, to the end of the 19^th century, leading Sherqawi marabouts maintained reputations as men of religious learning and were acknowledged as such by ‘ulama in Morocco’s leading cities. Yet by the 1920’s only a few isolated Sherqawa held such reputations. Still, given their competitive edge as notables, leading Sherqawa were generally able to shift their social, political and economic roles so as to maintain their social honour and elite status during the protectorate and after independence in 1956. Even as some leading educated Sherqawa deliberately dissociated themselves from maraboutic practices, clients of the Sherqawa continued to regard the success of some Sherqawa in any significant activity as a consequence of their maraboutic status and sought out those Sherqawa willing to maintain maraboutic ties with them.

Although the scope of activities performed by marabouts or kinsmen acting as proxies on their behalf has contracted and the size of offerings and sacrifices to them also has diminished, most rural Moroccans in the sphere of Sherqawi influence and a large number of urban Moroccans continue to maintain ties with the Sherqawa. In a few cases, as I have indicated, new ties have been established. The pattern appears to be the same for at least some religious lodges elsewhere in Morocco (e.g., Hatt 1974). Competition for acquiring a maraboutic reputation is no longer a principal means of maintaining social honour or power in Morocco, so few men actively seek reputations as living marabouts. But enshrined marabouts are another matter, and some persons who claim descent from them continue to “work” (ta-ykhdem) clients and to receive substantial donations of cash and produce for their efforts. Educated Sherqawa prefer to emphasize the piety, scholarship, and illustrious descent of their maraboutic predecessors rather than the popular conception of marabouts as “close” to God an J thus capable of efficaciously conveying his grace. In some contexts the association of the Sherqawa with maraboutism can be a distinct embarrassment. Yet maraboutism is far from residual in Moroccan society, for reasons that are described below.

Maraboutism and “Closeness”

Any ideological system is influenced by the social order in which it is expressed, and part-ideologies such as maraboutism are so closely tied to existing social arrangements that they cannot be comprehended without reference to them. Maraboutism is especially linked to two key concepts through which Moroccans make sense of existing social arrangements: obligation (haqq) and closeness (qaraba) (Eickelman 1976: 96–99, 141–149). Obligations are contracted and exchanged through services or offers of support. They are the culturally accepted means by which persons bond themselves to each other in Moroccan society and which determine the social honour of persons in relation to each other. Moroccans speak of having obligations “in” or “over” other persons, indicating the asymmetrical nature of the ties so created. All relationships impose obligations of calculable intensity. There is a considerable latitude within which such obligations can be contracted; thus ties of kinship and descent in themselves do not determine the nature and intensity of such ties. In general, persons strive for flexibility in relations where they are under obligation to others, while at the same time fixing as firmly as possible relations in which they hold obligations “over” other persons. Their goal is to preserve autonomy whenever possible and to be free to shift obligations as opportunities arise to enhance their social position.

“Closeness” is a form of relationship which is said to exist between persons bound together by multiple personal obligations and common interests and who regularly can be expected to act on each other’s behalf. Closeness is acting as if ties of obligation exist with another person which are so compelling that they are generally expressed in the idiom of kinship. This is because kinship ties, at least those involving common descent (‘blood’),were considered at the symbolic level to be permanent and unbreakable. Some relations of closeness are based upon kinship, although as stated, closeness based upon kinship is generally not sharply differentiated from closeness based upon other grounds. Closeness also develops out of patronage and clientship, residential propinquity or neighbourliness, membership in rural local communities or other “tribal” entities, and common occupation. Quite frequently these bases for closeness overlap and persons deliberately seek to make them do so. A brother can also be a client and a neighbour may seek to be considered a kinsmen, if it will enhance his status or secure practical advantages.

Ties involving marabouts are a particular instance of the concept of closeness. Below I first discuss the closeness said to exist between marabouts and their clients, then the complementary issue of the maintenance and legitimation of elite status within a maraboutic descent group.

The primary consideration in the client’s eye view of marabouts is that they are close (qrib) to God and therefore capable of influencing him or at least of relaying his grace. Having or acquiring a reputation for grace is sufficient for being a marabout, but the most-influential of them are distinguished by a number of complementary attributes. One of these is the claim of patrilineal descent from long-established marabouts such as Sidi Mhammed Sherqi, who have descendants actively maintaining their shrines and other cultic apparatus and thus indicating closeness to their ancestor. Marabouts claiming descent from the Prophet Muhammed, or, in the case of the Sherqawa, from Sidi ‘Umar (d. 644), the second caliph in Islam, are especially prominent. Such persons intermarry as equals but do not give their daughters in marriage to “commoners” (fawam), those who cannot claim either type of privileged descent. Second, marabouts are generally attributed with some degree of religious knowledge (‘Urn) and always with mystic insight. Both of these qualities are reputedly transmitted to them through dyadic ties with earlier generations of maraboutic scholars and mystics.[5] Finally, marabouts must be willing to transmit their baraka to their clients or to have someone claiming closeness to them who does so on their behalf.

Despite the formal ideological claim that only God knows who is “close” to him, clients necessarily seek this-wordly signs to determine who marabouts are. Within the limits of their prosperity and other considerations, clients seek ties with those marabouts or their descendants thought most capable of sustaining-or enhancing their interests. A primary contemporary sign of baraka is the ability of marabouts or some of their descendants to maintain a prosperous life-style based at least in part upon their clients’ offerings. In the past another sign of the mystic powers of marabouts was their ability to act politically on behalf of their clients. Sherqawa can still occasionally do so, as I have indicated above. Such political successes are popularly attributed to Sherqawi baraka whether or not the Sherqawi involved actively seeks to be regarded as a marabout or takes material advantage of Boujad’s pilgrim traffic.

Like other social ties, those with marabouts are thought to depend upon continuing exchanges of obligations. The principal problem for believers in marabouts is how to create and maintain such ties with marabouts or those “close” to them (descendants or other persons) in order to ensure a flow of grace to sustain their particular concerns. They seek to contract personal, dyadic bonds of obligation with marabouts in order to compel marabouts to act on their behalf.

Maraboutic Descent

Marabouts with descendants and with established cults are considered the most prestigious, since the well-being of their descendants and their cult is a concrete indication of the benefits of closeness to them. A maraboutic descent group is one which shares certain rights and prerogatives on the basis of a recognized claim to common maraboutic descent. The concept of closeness is crucial to understanding such claims. Ideologically, Sherqawi status is expressed in the idiom of patrilineal descent from the marabout Sidi Mhammed Sherqi. Yet such formal determination of status is rarely invoked and is frequently impossible on a practical basis. Hence it is possible to acquire Sherqawi status, even in Boujad itself, and not have such claims directly challenged, by “known” Sherqawa. In practice, one is “known” (me‘ruf) asaSherqawi on the basis of a combination of attributes. These include residence in a Sherqawi quarter, marriage patterns, political and economic alliances, and other aspects of comportment. Most Sherqawa are known through face-to-face relations among themselves and to their clients.

The acquiesence of known Sherqawa to such assertions of Sherqawi status can be explained on two levels, the analytical and that of social practice. Analytically, it is useful to employ Scheffler’s distinction between descent categories and descent groups (1965: 62). The idiom of descent in itself gives form only to categories of persons which can remain only vaguely bounded. Membership in such a category does not imply an obligation to act in common with other members of that category. However, perceived common interests can encourage the formation of descent groups out of such categories. On the related practical level of actual social arrangements, I already have indicated that Moroccans constantly re-evaluate their ties of closeness in terms of maintaining and whenever possible enhancing their social status. Thus in some instances the attribute of being a Sherqawi is a relatively minor component of social identity; in others it can play a major role.

Sherqawi descent groups show the same flexible patterns of boundary as do other collectivities in Moroccan society. Currently, the Sherqawa are divided into eight descent groups.[6] Six are considered to descend from sons of Sidi Mhammed Sherqi, although no one claims to know the intervening genealogical links. Some persons refer to written records which legitimize such links, but these never are produced and cannot be located. The remaining two groups are the most prominent. They claim descent from another of the marabout’s sons, but in addition claim as eponymous ancestors 19^th century Sherqawi “lords,” so designated because they effectively controlled the zawya or at least a substantial part of its resources. Each of these descent groups has a quarter (derb) in the town where the core of its membership resides, although non-Sherqawi also reside in these quarters. Each descent group possesses a shrine of its common ancestor, although this is not necessarily located in its quarter and many Sherqawi shrines are not directly linked to any specific descent group.

The same concept of closeness underlies claimed relations of kinship among the Sherqawa as among commoners. Substantial material and social benefits are associated with Sherqawi descent for at least some Sherqawa. Hence the genealogies tracing their descent are generally more complex than is the case for commoners. Or so the Sherqawa assert. Yet the only component of their genealogy which the Sherqawa commonly know is the name-chain (silsila) of the “lords” of the zawya. This name-chain stretches in an unbroken line back to Sidi Mhammed Sherqi, with variants occuring only over the “lords” of the last century. Earlier factional disputes are not socially significant and are not reflected in the name-chain.

All Sherqawa point to this name-chain as legitimating their status, but only members of the two most prominent “mainline” descent groups can actually trace agnatic ties with the Sherqawi lords. With few exceptions, the most prominent Sherqawa of the past century come from these two descent groups. For reasons of shared material interests and prestige, members of the two “mainline” groups show a much greater concern over the boundaries of their groups than do other Sherqawa. They generally can trace in detail their genealogical ties over at least three generations and possess a more precise knowledge of the scope of their kinship ties with contemporaries. They sharply distinguish between persons who claim closeness to them solely on such bases as residential propinquity and those related through agnatic and affinal ties. These two mainline groups have substantial common interests and in general have managed to defend them. They have kept largely undivided estates from their prominent ancestors of the last century. Through Islamic inheritance many of them are thus entitled to substantial revenues in addition to their shares in the offerings at the main shrine (shares available in principle to all Sherqawa).

Sherqawa active in the maraboutic “trade” seek to manifest closeness to leading Sherqawa before their clients. This is despite the fact that most prominent Sherqawa of today, even when popularly regarded as marabouts, no longer take an active part in maraboutic enterprises. Identification with the two mainline descent groups gives persons a competitive edge in maintaining status, but participation in the aristocracy or inner elite of the Sherqawa (or esteem in clients’ eyes) is not determined exclusively by descent group membership. For example, a member of one of the non-mainline descent groups, a textile merchant, successfully entered politics shortly after independence. At first he actively dissociated himself from Sherqawi maraboutic activities (because the mainline Sherqawa failed to support him in election campaigns) and in fact became prayer leader (imam) of one of Boujad’s two principal mosques. In recent years he has consolidated his influence by claiming, at least away from Boujad, to be lord of the zawya and of the mainline Sherqawa. He employs his visible hold over a rural clientele as a means of legitimating his political influence.

Until the beginning of this century, Sherqawi descent and maraboutic status were largely coterminous. The principal means for a Sherqawi to acquire social dominance was to seek to be regarded as a marabout and to attract a clientele. The means of acquiring social prominence are now more varied. The immediate descendants of late 19^th and early 20^th centuries Sherqawi marabouts have often maintained elite status, at least when they have acquired an education or adapted successfully to modern commercial and political activities. Through these activities it is possible to maintain a social position that identity primarily as a marabout no longer allows. Many of the Sherqawi elite dissociate themselves from the maraboutic interests of some of their kinsmen because such activities would open them to attack and ridicule from educated persons. Such Sherqawa seek to redefine the activities of some of their kinsmen and ancestors in light of the formal doctrines of reformist, modernist Islam. By any criteria, some Sherqawa have been prominent in Morocco for centuries. — as marabouts, counselors of sultans, religious scholars, and recently as educators, merchants, ministers of state, doctors, judges, administrators, and in one case, as a sociologist. The very success of these Sherqawa is taken by maraboutic clients as a sign of continuing Sherqawi religious prominence. The elegant Sherqawi shrines are maintained through their donations.

Ideology and Practice

The analytic distinction has already been made between maraboutism as an ideology and maraboutism as a concrete set of social practices. The contracting economic bases of marabouts, especially living marabouts, has had an impact on the ideology of maraboutism. It is this impact that I now wish to discuss in reference to the Sherqawi maraboutic cult. Ideologically, marabouts and maraboutic cults form a symbolic base for order in society. Maraboutic cults such as that of the Sherqawa are identified with specific towns and regions, although the correlations between maraboutic cults and specific rural or urban local communities are not immutable and some groups may even divide their allegiances among several maraboutic cults. The implicit ideology of such cults also provides a bridge between the divine template for human conduct as revealed in the Qur’an and periodically renewed by reformist movements in Islam, and the realities of “the way things are” in Moroccan society. Maraboutic cults are linked to specific social forms but are more than a reflection of such forms. This is indicated by the analysis of the means through which specific ties with marabouts are maintained.

There are both collective and individual ties with Sherqawi marabouts. Presently, only tribal groups maintain collective ties with the Sherqawa, although in preprotectorate times urban quarters and occupational groups in western Morocco also did so. In certain parts of Morocco, including Marrakesh (Jemma 1971), collective urban ties still exist. Individual ties are maintained by both townsmen and tribesmen.

Personal and collective ties of obligation with marabouts rest on a similar base, so that the form which collective and personal ties take and their distribution are not sharply distinguished. In both cases the fundamental unit of social structure is the person rather than their attributes or statuses as members of groups. Thus, persons participating in collective sacrifices and offerings say that they do so because in that manner they are able to present a more substantial sacrifice or offering to a marabout and thus more forcefully indicate their faith in him. In other words, persons deal with marabouts collectively rather than individually in order to have more “word” with the marabout. Of course, such ties also denote the boundaries of particular social groups at any moment.

The accompanying map indicates a large, contiguous sphere of Sherqawi influence covering the Shawya and Tadla plains. In rural contexts, the Sherqawa are the dominant religious lodge in the area, although to a lesser extent other maraboutic orders also are represented. The urban situation is slightly more complex, but only due to post-1912 developments (with the exception of a few coastal towns including Casablanca). It also should be kept in mind that Boujad was the only town of significance in the interior of the area prior to 1912. An important earlier form of ties with the Sherqawa took the form of urban religious brotherhoods[7] (tariqa-s), whose members met weekly, usually on Friday afternoons, in a local lodge (zawya) maintained for that purpose. There are indications that these lodges may have been related to trading networks based upon Boujad in the late 19^th century (Eickelman 1976: 199). They exist alongside non-Sherqawi religious lodges, most of which gained adherents in the early years of the protectorate, when merchants and craftsmen joined them as a means of asserting independence from the Sherqawa. From the early 1930s onward these brotherhoods declined in face of reformist Islam and the emerging nationalist movement.

Personally contracted ties with the Sherqawa are much more widespread than those of brotherhoods in the urban setting. Significant numbers of townsmen occasionally travel to Boujad to visit Sherqawi shrines there, as well as prominent non-Sherqawi shrines elsewhere. Generally, however, only townsmen of rural origin in western Morocco make regular offerings to the Sherqawa over extended periods of time. More substantial in terms of aggregate value are the donations of animals and cash irregularly made by persons who seek the intervention of Sidi Mhammed Sherqi or other principal Sherqawi marabouts in cures of illness and infertility. These offerings occur only at times of personal or household crisis and to a lesser degree are also made to non-Sherqawi marabouts.

The ties between the Sherqawa and specific tribal groups in western Morocco at first appear more stable than the amorphous, discontinuous ones sustained in urban settings. The major sphere of Sherqawi influence has remained relatively constant since at least the 1880s. This is in large part because a succession of sultans in the late 19^th century deliberately sought to extend Sherqawi influence in the area at the expense of other maraboutic lodges. In 1901 this policy was formally reversed, but by then the sultanate was unable effectively to impose its will in the area. From 1907 onward the French used Sherqawi influence for their own ends. Once the pax gallica was formally established in 1912, spheres of maraboutic influence were largely frozen in place. This is in contrast with the 19^th century and what is known of earlier periods, when the geographical sphere of Sherqawi influence expanded and contracted with political and economic currents. In the 1880s, for instance, the Sherqawa had clients in the Middle Atlas mountains and actively sought to extend their influence in that direction (Eickelman 1976 : 55). The existence of Sherqawi agents, holdings, and clientele in key Moroccan cities outside their main sphere of influence, such as Fes, Meknes, Marrakesh, and Sale can similarly be explained in terms of their seeking to widen the scope of their activities. Of course, the geographical pattern of Sherqawi influence is partially related to ecological constraints. Some geographical features tend effectively to impede communication in certain directions. Consistently more important, however, were the patterns of social and political alignments at any given moment. This explains in part the patterns of Sherqawi influence in an area as far away as the Sus valley. Certain Sherqawa were effectively able to create and maintain ties with groups and individuals there and the development of ties with the Sherqawa was considered more advantageous than the development of ties with local maraboutic descent groups. Such decisions depend upon judgements of the efficacy of particular marabouts and the availability of particular marabouts or their descendants.

Myths relating the exploits of particular marabouts are a crucial element in maraboutic ideology. Those for the Sherqawa are known throughout western Morocco and resemble in form those associated with marabouts and maraboutic descent groups elsewhere. To the extent that these myths frequently invoke place names and other features peculiar to specific geographic divisions, they are localized. Yet these characteristics are almost incidental to the stress in the myths upon how marabouts become tied to specific groups or persons. For example, one myth relates how the entire area that surrounds Boujad was wilderness when Sidi Mhammed Sherqi first arrived there (Eickelman and Draioui 1973: 201–205). The marabout located water, cleared the forest, and forced wild animals out of the area. Then he called upon various groups to settle there and establish a covenant (‘ahd; bay’a) with him. So far I have collected roughly sixty myths, including variants, dealing with Sherqawa. Virtually all of them concern the means by which covenants are established and honored between named groups and Sidi Mhammed Sherqi and his sons. Such covenants are represented as being maintained through such means as sacrifice, the giving of women to the Sherqawa, the claim of a common, distant ancestor between the marabout and his clients, and the claim of mere physical propinquity in the distant past.

In exchange for sacrifices and the renewal of the covenant, Sherqawi marabouts appear as guarantors of the moral order insofar as it affects their clients; in some cases the clients are made Muslim at the same time a covenant is established. One common theme in the myths is that of Sherqawi marabouts defending covenanted groups from unjust tyrants. The ties of the Sherqawa with other maraboutic centres and with Mecca are also frequently stressed. Although in real life Sherqawi marabouts and other Sherqawa have made the pilgrimage to Mecca, there are few sustained ties with other maraboutic descent groups. The social order depicted in the myths is constructed out of particularistic bonds of obligations, but the myths also depict the submission of both marabouts and clients to an Islamic code of conduct.

Covenants with the Sherqawa are particularistic but far from being primordial. Most client tribal groups in western Morocco claim that their covenant with the Sherqawa was sealed in “early” times (Bekri). “Early” is that range of time, internally atemporal, which forms a backdrop to the ordinary, fully known social horizon (see Eickelman 1977). Events that occurred in early times are thought to have a significant impact upon current social alignments but the full social context of such early events is considered unknown and unknowable. Significantly, the tribal groups mentioned in the covenants of the myths are not “real life” collectivities in the sense of being effective units of social action either now or in the historically known past. Tribal names mentioned in the myths are generally those from which many existing groups claim to have derived, but even this is not always the case. It is useful to think of the names mentioned in these myths as together forming a conceptual grid through which covenants can be legitimated as well as shifted as occasion warrants through the “discovery” of alternate formulations of the grid.

The notions of obligation and closeness underlie the dyadic bonds through which marabouts and their clients remain “connected” (mtesslin). A “connection” can be accomplished in three complementary ways. The first is through offerings and communal meals with those living descendants of Sidi Mhammed Sherqi, not themselves considered marabouts, who claim specific ties with certain tribal sections and rural local communities. These descendants are called “visitors” (zewwar-s)[8] because they literally visit clients’ households and tents twice each year — in spring when herds are traditionally divided and in early summer just after harvesting — to collect the offerings (zyara) considered the due of the marabout. When they make donations or offer communal meals, clients specifically remind the visitor that he is to accept them on behalf of his ancestor. In return, visitors offer an invocation (da‘wa) to their maraboutic ancestor in front of their clients. Visitors know each individual involved and carefully specify the particular concerns of each household: the desire for male offspring, good harvests, the wellbeing of persons and animals. In the past, visitors also mediated tribal disputes and brought major matters to the attention of more influential Sherqawa. When clients are in town, their Sherqawi visitors occasionally assist them in various matters, but in general the visitors no longer can serve as efficacious intermediaries in practical matters. An interesting consequence is that clients now explain their offerings to visitors as alms to the poor (sadaqa) and treat the visitors in an offhand way. Still, contact with visitors is considered necessary in order to maintain ties with their maraboutic ancestors.

The second way of remaining connected is through annual sacrifices at the shrine of Sidi Mhammed Sherqi and those of certain of his descendants. These are made in the presence of living Sherqawa, as well as the visitors associated with particular pilgrims and pilgrim groups who again act on behalf of the marabout. Such sacrifices can be made by individuals who seek specific favours from marabouts and are made regularly throughout the year. Here I discuss primarily the collective offerings that are made. Most collective sacrifices are made each fall during the festival of Sidi Mhammed Sherqi. This is not a single event, but a period of roughly two weeks during which client groups make a pilgrimage to Boujad. For transhumant groups, the festival generally coincides with the move from summer to winter pastures. Especially over the last two decades, the Sherqawa have discouraged actual sacrifices, although many still occur. One reason for their discouragement is pressure from other townsmen, including Sherqawa, who are embarrassed by what they consider the more “primitive” forms of maraboutism (although a similar argument is never made concerning the sacrifice of an animal for feast days recognized by “formal” Islam). Another reason is the desire on the part of the Sherqawa involved in the pilgrim “trade,” as they refer to it, to live from the proceeds of the pilgrim traffic. To this end, they persuade pilgrims whenever possible that the offer of a sacrificial animal or the equivalent in cash has the same benefit as a sacrifice.

In some cases, sacrificial animals are slaughtered but immediately sold at regular prices at a butcher’s shop; in most cases, however, Sherqawa dispose of the animal (or offerings in kind) through regular market channels.

An unanticipated consequence of the gradual abandonment of actual sacrifices and the greater demand for marketable offerings is the rapid attrition of client donations. Traditional offerings of cereals and of animals to a lesser extent were considered largely outside of the domain of the cash economy prior to the severe economic dislocations of the Second World War. It is true that to a certain extent obligations to marabouts were always thought of in terms analogous to the market situation. In all forms of “remaining connected,” acrimonious bargaining often occurs over the size of an offering or sacrifice. Nineteenth century correspondence attests to the fact that such bargaining is not recent. I stress analogous because informants emphasize that a cash value cannot be placed on any bond of closeness or obligation, including those with marabouts.

Yet under current conditions the scope of maraboutic powers is much more circumscribed than it was in the past and virtually inexistent in the realms of politics and economics. Clients get less out of maintaining such obligations than they did in the past. In addition, when Sherqawi visitors and the caretakers of major Sherqawi shrines convert offerings into cash at the market, often before the eyes of their clients, the latter are encouraged to consider their ties with marabouts as identical with those of the marketplace. Doubt in the prevalence of marabouts has always been present (cf. Firth 1974), but under present conditions this doubt becomes systematic and encourages consideration of the tenets of formal Islam as the only valid religious expression.

The final means of remaining connected with the Sherqawa is through offerings and communal meals with living marabouts. I have already mentioned why this practice is rapidly disappearing. Living marabouts have difficulty in maintaining their reputation due to decreasing influence and revenues and are stigmatized by educated Muslims. Unlike enshrined marabouts, living ones are not insulated from direct calculations of their efficacy and comments upon the lavishness, or baraka, shown by their hospitality. Once again, as the benefits of maintaining obligations with marabouts become more vague and indistinct, not only do pilgrims place less value in maintaining such obligations, but maraboutism as an implicit ideology ceases to make sense in terms of the concepts of obligation and closeness.

Sacrifices and Social Groups

Another dimension of maraboutic ties is the means by which groups are formed that collectively sacrifice and make offerings. As an example I use the Sma’la of the upper Tadla plain, located about twenty kilometers to the north of Boujad. They are one of the most “traditional” tribes of the area in that they remain largely transhumant (as were virtually all the tribes of the area prior to the 1920s) and vigorously maintain their covenants with the Sherqawa. Therefore their current practices (and those documented for the protectorate and the pre-1912 period) provide more insight into precolonial practices than is the case for groups that have been more radically affected by subsequent political and economic transformations. The nature of tribal society in this early period is significant because maraboutism has been primarily linked in Morocco to “segmentary” tribal societies. Here I present an alternate account of traditional rural society as it existed in some parts of Morocco and which does not necessitate postulating a radical dichotomy between past and present social forms.

The pasture lands of the Sma’la are still collectively owned; only recently has legislation been passed which, when eventually implemented, will convert all “tribal” lands into private ownership. Agricultural lands are owned by individuals, as was also the case in the region during the preprotectorate period. Herds are also individually owned, although in practice the head of each tent (household) or herding unit exercises effective control over the disposition of animals owned by individuals residing with him. Agreements over grazing rights were made with neighbouring groups both on a person-to-person basis and on the basis of larger collectivities. Even among the “traditional” Sma’la, therefore, individuals had a considerable area of latitude in which they could manipulate their social identities. Such latitude is administratively inconvenient, so the colonial administration and its post-1956 Moroccan successor discouraged small-scale, informal pastoral agreements, sometimes using the pretext that only agreements binding on the larger, administratively recognized tribal subdivisions were really “traditional.” Neat administrative organigrams of tribal organization and administrative jurisdictions would be damaged by acknowledging a continuous flux in social alignments.

Tribesmen employ the idiom of agnation to describe relations among themselves, but political relations are conceptualized in other idioms as well. Nor was the idiom of agnation more pervasive in the historically known past, which in this case means since the late 19^th century. Tribesmen often use the analogies of the tree and the human body to describe their interrelationships. To follow through the first analogy, as one tribesman said: “We are like the branches of a tree, but each branch is on its own.” Similarly, in some contexts tribesmen assert that they are all linked through ties of blood, but explicitly deny that the claim of such ties obliges them now or obliged them in the past to act in specified ways. In other words, social and political action flows along numerous lines: kinship, patron-client relations, and other bonds of necessity or mutual interest. The area of latitude in which rural Moroccans can shift their social ties is more circumscribed than in an urban milieu, but it is still fairly considerable.

When tribesmen are asked to describe their relations with the Sherqawa, they frequently begin by discussing the nature of the rural local community (dawwar). The term literally means “circle.” In the era of preprotectorate insecurity, household heads of local communities pitched their tents in a circle during transhumance. They collectively agreed upon pastoral movements and upon the tenor of their relations with other groups. Members of local communities claim to be linked through agnatic ties but are unable to demonstrate specifically that such ties encompass all members of their group. When asked to trace such relations in detail, tribal informants invariably responded that only someone older could do so. In turn, elderly informants vaguely referred to “earlier times” as the period when the local community was entirely linked through traceable agnatic relations. In actuality, local communities commonly consist of several agnatic cores linked by complex affinal and contractual ties.

There are two significant criteria for the existence of a rural local community. First is the general willingness of its members to support each other in mutual interests. Second is the capacity to act collectively on certain ritual and political occasions. The comparison with the earlier discussion of maraboutic descent groups (or for that matter any other such groups in Morocco) should be clear. Each local community has a council (jma’a), although this is not a formal body as has frequently been assumed by protectorate administrators or their Moroccan successors. Any adult male household head who has an interest in the issues at hand participates, although the opinions of those who are economically or politically weak tend to be discounted or ignored in favour of those with more substantial material interests and persuasive skills. When necessary or convenient, such councils informally agree upon a spokesman (mqaddem) to represent them to outsiders. Each local community also has its own mosque, burial ground, and maraboutic tomb. The latter is usually a derelict structure to which no offerings are made except for candles and small coins left by the local community’s women and children. At any time, each household or tent belongs to only one local community. Membership is demonstrated by sustained ties of “closeness,” a concept which I earlier illustrated in relation to the urban Sherqawa.

The refusal or inability of a tent to contribute to certain collective obligations does not in itself indicate dissociation from a local community, but it affects the capacity of a household head to have a say in the community’s affairs. When a definitive break occurs a household shifts its allegiance to another local community, although its rights in agricultural land remain unchanged. When a group of tents or households breaks away and forms a separate local community, it is distinguished by its own mosque, burial ground, and maraboutic tomb. One of the numerous unused structures usually is appropriated for the purpose, although new tombs occasionally are built.[9]

When Sherqawi visitors arrive in local communities, they usually deal with a spokesman agreed upon for that purpose. He sees that the offerings of grain and animals are forthcoming from the households of his group and cajoles recalcitrant members. A similar informal arrangement is made at the level of tribal sections (fakhda-s) in jointly contributing sheep or bulls for sacrifice (for all Sma’la tribal sections refuse substitutes for sacrifices) at the main Sherqawi shrine in Boujad. No such sacrifices are made at the level of the Sma’la acting as a “tribal” collectivity, nor were they in the known past.

Existing social alignments can be discerned through the clusters of persons that contribute to various maraboutic obligations. More specifically, the conduct of individuals and groups at the annual festival of Sidi Mhammed Sherqi constitutes a publication of new social alignments. These are revealed in the arrangement of tents, patterns of offering and sacrifice, and the part taken by local communities and tribal sections in “powder plays” (tehrak-s)’ competitive displays of horsemanship. Such “powder plays” among the Sma’la, as with groups participating in similar activities elsewhere in Morocco, are always accompanied by heavy, serious betting between the groups involved. In part these competitions reflect the lack of formal rules inherent in the Moroccan concept of closeness. I he judgment of their outcome is left to the competitors themselves and to any support they can muster among onlookers. Differences that develop during the annual festival often themselves engender new lines of social demarcation.

Once such shills in the composition of groups are effected, appropriate realignments are made in how these groups conceive of their relations with one another. Similar adjustments occur when groups or individuals decide that their link to a specific Sherqawi “visitor” can or should be replaced by one more efficacious.

Available evidence for the preprotectorate past suggests that the rural social order and its relation to maraboutism was basically similar in pattern to what I have described above. Of course, it can be argued that the political and economic transformations of the last sixty years have been so substantial that tribal organization described primarily from contemporary ethnographic evidence or the contemporary remembrance of things past differs fundamentally from that of the preprotectorate era. Some basic shifts have occurred. Since the protectorate, for instance, administrative intervention has greatly restricted the possibilities of intergroup hostilities. Matters previously settled among the tribes themselves or through maraboutic mediation now almost inevitably involve government intervention.

Two sets of events involving the Sma’la and other groups in western Morocco during the preprotectorate era are known in sufficient detail to suggest an essential continuity in local conceptions of the social order. One was the manner in which resistance was organized against the French for three years beginning with 1910. Rapid shifts of alignment occurred that indicate a flexible, opportunistic perception of the mechanics of alliance formation. A similar flexibility is evident in examining the repercussions of the struggle for the sultanate between the brothers Mulay Abd l-‘Aziz and Mulay Hafed which began in 1907. Protracted local disputes over water rights and other matters, maraboutic allegiances both to Sherqawi factions and other maraboutic centers, the attitudes of local communities and tribal sections about the legitimacy of the two rival sultans, and assessments of the wisdom of resistance to the French all entered into consideration. The alignments of the various collectivities involved bore minimal relation to asserted agnatic relationships.

My concern with extending my argument to at least the known precolonial past is to indicate that there is no need for a “special” explanation of the relation of maraboutism to the tribal social order. The most widely known discussion of this issue in English is contained in the writings of Ernest Gellner. He is primarily concerned with the role of marabouts in social structures which qualify as “segmentary” and which no longer prevail in Morocco. Thus Gellner’s argument is necessarily relegated to the precolonial past when he considers it reasonable to assume that the society in the High Atlas region which he studied was in a stable “kind of Social Contract situation” over an indefinite period (1963: 146; 1969: 158–159). Insofar as present-day institutions can be assumed to have been in “a kind of sociological ice-box” (Gellner 1973: 59), he argues that they can be taken as reliable indicators of past situations.

A principal drawback to this set of assumptions is that it offers no explanation for the role of marabouts and the ideology of maraboutism in non-segmentary situations in the past or for the continuing, albeit modified, significance of both the ideology of maraboutism and the activities of maraboutic cults in the present. Prior to the appearance of the Islamic reform movement in Algeria in the early decades of this century, maraboutism was so persuasive that most urban and rural Algerians took it for Islam pure and simple (Merad 1967: 58). Merad’s statement can be applied even more firmly to Morocco. As for the present, as indicated earlier, rituals associated with marabouts have held their own in many urban and rural milieux (Hatt 1974: 25; Pacques 1971; Jemma 1971; Eickelman 1974a: 229). Contrary to the fervent desires of some Muslim reformists and other educated Moroccans, the ideology and practice of maraboutism has not just melted away in the face of recent social and economic transformations. At least with respect to religious belief and practice, these facts argue against assuming a radical discontinuity between past and present in Morocco’s social history, and especially against any one-to-one relation of ideology with the social order at any historical moment.[10]

Conclusion

The Moroccan concept of closeness relates to maraboutism on two levels. As an organizing principle for social arrangements, it underlies such diverse social forms as descent groups (including maraboutic ones), urban quarters, tribal sections and rural local communities, and ties between marabouts and their clients. Closeness as an organizing principle is essential for comprehending the particular implicit ideology and social relations described in this paper. I have indicated the diversity of historical and social-contexts in which closeness operates to suggest its variable relation to maraboutism taken as social ideology and practice. Any one set of “social arrangements” considered in isolation with maraboutism as an ideology would lead to an analytically unjustified assumption of a one-to-one link between them.

Maraboutism like any other ideology is socially maintained. It is based partially upon an implicit analogy between ordinary social relations and those which prevail between men and the supernatural. In fact, for those who accept maraboutism, there has traditionally been no sharp division between the two realms. Persons considered “close” to marabouts, especially members of their descent groups associated with cult activities, are thought to have a special influence over them. The notions of obligation and closeness make these ties comprehensible.

This paper has been concerned with two sorts of transformations which occur in relations between marabouts and their clients. First is the continuous revaluing of the utility of specific social ties, including those which exist between marabouts and their clients. These can lead to the rise or fall in the fortunes of specific marabouts, maraboutic descent groups, or regional pilgrimage centres such as Boujad. In themselves, this form of realignment does not substantially modify the content of maraboutism as an implicit ideology.

The second form of transformation directly modifes the form of maraboutic ideology and analytically is the more interesting (and neglected) form. It concerns shifts in social action which engender ideological change. Put succinctly, my position is that to be understood sociologically, maraboutism has to be considered in a range of historical contexts. In the preprotectorate past, now only a lived experience for the most elderly of Moroccans, marabouts had a much greater range of social action than they have had in later periods. Living marabouts complemented enshrined ones and directly or through proxies could choose to affect the circumstances of their clients. More recently, the political and economic activities of marabouts became markedly circumscribed, until what remains to them is primarily a “religious” sphere of influence distinct from practical concerns. In earlier periods such an explicit distinction did not exist. With marabouts in the past, the intercession of the supernatural in such concerns was a daily occurrence. Now the benefits which derive from sustaining ties with marabouts and their descendants are vague and ill-defined.

The form in which ties between marabouts and clients are established and reaffirmed has also shifted and this has had more radical consequences for maraboutism as an ideology. Few persons can sustain reputations as living marabouts, and clients in any case have less to do with them. Both living marabouts and maraboutic “visitors” have encouraged the substitution of sacrifices for other forms of offering, and when offered goods in kind have directly converted them for cash through regular market channels. A consequence of this shift has been to encourage clients to consider their transactions with marabouts just as they would all other transactions. When viewed in this fashion, there is no perceptible benefit in maintaining ties with marabouts. Maraboutism ceases to make sense as part of “the way things are.”

Finally, the strength of maraboutism is precisely in its being taken as an implicit ideology. Those who support marabouts are aware in outline of the tenets of “formal” Islam and indicate their support for it through building mosques — most rural local communities have one — and hiring Qur’anic teachers. Although formal Islam also claims to be related to all aspects of life, there is no direct, personified means by which the supernatural can pervade the social order as is the case with maraboutism. Formal Islam thus requires no direct analogy with the nature of the local social order. In contemporary Moroccan society this disjuncture allows it increasingly to be taken as a religious representation of the world more meaningful than maraboutism.

References

© Dale F Eickelman 1977

2. Communal and Individual Pilgrimage: The Region of Saints’ Tombs in South Sinai[11]

by Emanuel Marx

Introduction: The Region and the Wider Universe in the Cult

Regionalism among the Bedouin tribes of South Sinai is more or less confined to their area’s geographical boundaries. It has one main concern: the sharing of certain of the area’s resources for the benefit of all the tribes who inhabit it. This is clearly represented by the Bedouin themselves in their cult of saints and holy tombs. But their cult refers to an even wider entity also: the Bedouin’s insistence that the saint is only a mediator between man and God indicates that they refer to a wider universe. In their experience this is the world in which they work and from where they obtain most of their food and other supplies. This world is located chiefly outside the borders of South Sinai and, even more importantly, the Bedouins do not exert influence on its activities. A distant government and other external organizations determine much of their lives, but contact with, not to mention influence over, these forces is minimal. For the bureaucratic chain does not fully extend to the Bedouin; and their own appointed representatives, the sheikhs, do not greatly influence officials at the centre. So the Bedouin place their reliance on a ‘spiritual’ representative, the saint, who is at least close to God and able to influence him.

The cult of saints has been reported from many parts of Islam (Levy 1962: 258). The majority of Islamic leaders do not object to this practice. Linant de Bellefonds (1974: 355) claims that “The permissibility of visiting tombs was admitted very early on by idjma’ [consensual canonical law]; all the [law] schools....even went so far as to recommend the practice”. Some authors claim that it “represents a striking deviation from the genuine prophetic tradition” (Grunebaum 1951: 67), perhaps because, by western standards, it is logically inconsistent with a strict monotheism. It is practised not only in regions where “official, genuine literate Islam....is present only in a minimal form” (Gellner 1972: 59); it is found also in cities, villages, and even in association with mosques (Canaan 1927: 2; Eickelman 1976: 112). Nor is it an “expression of that necessity....for a world religion to come to terms with....a multiplicity of local forms of faith and yet maintain the essence of its own identity” (Geertz 1968: 48). For centre and periphery are united in the same faith, and its various orthodox and heterodox forms are spread almost uniformly throughout its realm.

The cult of saints is, then, neither a parochial phenomenon, a series of quaint local traditions, nor a deviation from official Islam. Its very ubiquity should preclude such conclusions. It should be viewed as a link between local societies and the “wider society”, as conceptualized by the locals. Much of the essay is devoted to a demonstration of this thesis.

However widespread the cult of saint is, it appears to be less common among nomadic pastoralists. Musil (1928: 417) states categorically:

“The Bedouins know of no communion with the saints. In the whole inner desert there is not a single holy grave or shrine erected in honour of a saint....When they make short sojourns in the settled territory, where by every village the dome of a shrine rises above the real or imaginary grave of some man or woman whom public opinion considers to be a saint, they never pay any attention to these domes.”

But among the Bedouin of South Sinai and, incidentally, among those of Cyrenaica (Evans-Pritchard 1949: 66–7: on recent changes see Peters 1976),the visiting of shrines and pilgrimages is an important feature of social life. If we gain some insights into the sociology of the cult of saints in Sinai, we might improve our understanding of such cults elsewhere, and interpret their absence among some nomadic pastoralists. I suggest that tribal pilgrimages elsewhere, too, develop, in response to similar situations: in tribal or other economically undifferentiated hinterlands that depend on external forces (natural forces, or an external state and economy) over which they exert little influence. A society that controls its affairs to a considerable extent, will not resort to communal pilgrimages.

A further basic point must be made. To understand the importance of the cult in South Sinai it is essential to appreciate variations in communal pilgrimage. In this essay such questions as why tribes should celebrate an annual reunion, why the reunions of each tribe follow a different pattern, and why the attendance at them fluctuates from year to year will be asked. Lastly, the significance of the holy tombs will be examined. Why do they mark the sites of the gatherings of one or several tribes? This touches on some of the problems discussed by Robertson Smith nearly a century ago in his Lectures on the Religion of the Semites (1889). It will become evident that some of his views have stood the test of time.

Three Tribes and their Setting in the Peninsula

Every year towards the end of summer members of the Bedouin tribes in South Sinai congregate at certain Saints’ tombs in their area. There are about twenty well-known tombs in the southern half of the peninsula, and they are all located at major cross-roads and not far from water sources. Some are found close to centres of population, while others are far from them. Throughout the year individual Bedouin visit tombs whose resident saint, they think, can help them with their problems. Only at some of these tombs do annual tribal gatherings take place. I intend to look closely at the annual pilgrimages of three tribes, the Muzeneh, the Awlad Sa’id, and the Jebaliyeh. The Muzeneh meet for three consecutive days and make the round of three holy tombs located within a range of 20 kilometres; from the tomb of their founding ancestor Faraj, they move to that of the prophet (nebi) Saleh, and thence to that of Aron, the brother of Moses. The Awlad Sa’id gather once a year at the tomb of ‘Ali Abu Taleb, the third Khalif. As for the Jebaliyeh, their sections gather on separate occasions at the tomb of Sheikh ‘Awad, a holy man whose antecedents are unclear.

Other saints’ tombs at which tribal gatherings take place are: Sheikh Qra’i for the ‘Aleqat, Sheikh Suliman Nfe’i for the Qaragha, al-Hashash for the small Hamada, and the female saint Sheikha Swerha for the Sawalha. All these major shrines are classified by Bedouin as maqam (literally “place”) a term widely used in the Islamic world to denote an important holy place, mostly, but not always, associated with a saint’s tomb. Here, in Sinai, maqam means a shrine whose patron saint is buried elsewhere.

Tribesmen attend these gatherings in order to meet their friends, relatives and visitors from other tribes, and to reaffirm their membership of the tribe and the right to use its resources. For the greater part of the year tribesmen are dispersed in small groups and many men work most of the time outside their tribal land and even outside South Sinai. The tribal reunions are the culmination of their efforts to maintain the ties with other Bedouin. The variations in the tribal reunions are connected to the different habitat and economies of the three tribes, as shown, among other things, in their annual migrations.

South Sinai is the mountainous southern part of the Sinai peninsula up to the Tih Plateau. It covers an area of about 17,000 square kilometres. Most of its interior is bare and rugged; high plateaus interspersed with mountains, build up to the Mount Sinai massif at the centre. The highest peak in the peninsula is Jebel Katarina, 2642 meters and other mountains are nearly as high. Asphalted roads run close to the sea shore round the tip of the peninsula. In the interior of the country there are only mud tracks, mostly running on an east to. west axis. Some of these are suitable for trucks and other rugged vehicles. There is no regular public transportation of any kind, but the Bedouin car fleet, of over a hundred vehicles, mostly American and Russian jeeps, pickups and a few trucks, mostly in a bad state of repair, is up for hire. Parts of the peninsula are inaccessible to motor cars, and can be reached by camel, often only by circuitous routes.

The peninsula is arid, with small amounts of rain-fall in the cool winter months, between November and March. In the coastal sector, the annual average precipitation is about 10 milimetres and in the high mountains— about 60 milimetres (Gdnor 1973: 35). Rain-fall is also very irregular, and in all parts but the high mountains there are often two to three consecutive years without rain-fall. The spatial distribution of rains is also very changeable, and in consequence there are few reliable pastures. Any sizable rain causes floods in the lower reaches of wadis and often results in serious damage to Bedouin property. Sometimes most of a year’s rain-fall comes in a single downpour. Some of these torrential rains may occur in early summer. A flood that occurred in May 1968 destroyed hundreds of palm trees, houses and gardens, and ruined several major motor tracks. Another occurred in February 1975, with results almost as disastrous.

The winter months are usually cool. In the mountains there are each year cold spells, with temperatures occasionally falling as low as —10° (14°F). Summer temperatures are uniformly high, and in the low-lying areas often climb up to 40° (105°F) during the day, and in the mountains to around 30°.

The vegetation is adapted to the harsh climate, and relies largely on ground water. During the long dry summer, most plants wither, and only the ubiquitous ba’tharan, (arthemisia), ajram (anabasis) and some other shrubs seem to withstand the heat successfully. But after rains colourful annuals appear. Some plants come to life even in the absence of rain; milder weather after a cold spell entices them out of the ground. There are few sources of surface water in the region, except immediately after rains. Water is mostly obtained from wells dug by enterprising Bedouin. In recent years government sunk several deep wells to which all Bedouin have equal access.

There is reason to believe that a small and highly mobile population of nomadic pastoralists could maintain themselves in South Sinai, provided they had access to all parts of the peninsula. It appears that a few Muzeneh and Awlad Sa’id households subsist exclusively on their flocks.

No grain is grown in South Sinai, and the fruit and vegetables grown labouriously in small plots suffice for the needs of a household during a few weeks in summer. Practically all the food consumed in Sinai is imported from outside the peninsuja. In the past, when employment was scarce and communications slow and laborious, Bedouin devoted more attention to their gardens. Some households, especially among the Jebaliyeh, relied on them for their livelihood. Since the early 70’s, wage labour became easily available and the incomes of Bedouin rose rapidly. As a result Bedouin spend less time in their gardens, but still maintain them in reasonable condition.

South Sinai is inhabited by over 6,000 Bedouin, giving a population density of less than one person per 2 square kilometres. A considerable part of this population lives close to the main east-west passage through the mountainous interior, Wadi Firan—Saint Catherine—Wadi Nash. Although every Bedouin has a place or at least an area, that he considers his home ground, there are few permanent settlements of Bedouin and these have stabilised in recent years as local employment opportunities improved, and new wells were put in operation. The trend began in the mid-fifties when smuggling of narcotics accelerated and Bedouin invested some of the profits in wells and orchards. Some of these settlements, such as Bir Zrer and Bir ‘Oqda, were established in relatively inaccessible places, for understandable reasons. The copious ground water supply in Wadi Firan allowed Bedouin to remain for many months in one place and to cultivate fruit and vegetables. The permanent employment found in Saint Catherine’s monastery in the mountains gave rise to a Bedouin village in its vicinity. These trends were intensified after the Israelis took over in 1967. Though smuggling became almost extinct, more jobs became available locally, and new wells were dug. As a result, Bedouin households now tend to spend much longer periods in their home grounds.

Permanent settlements were few until the 50’s. There was the little port of al-Tur on the Red Sea with a population of about 500 souls, mostly sailors and fishermen. Some of these men were Bedouin. Near al-Tur there was a quarantine station for Egyptians returning from the annual pilgrimage to the holy places of Islam in Mecca, which provided seasonal work for many Bedouin. The other permanent settlement, Saint Catherine’s monastery on Mount Sinai, is inhabited by 10—12 Greek monks. Founded in the sixth century and continuously occupied since, it is rightly viewed by Bedouin as the only stable institution in the peninsula, more permanent and reliable than ephemeral governments. In the 50’s, the Egyptians began to develop the peninsula. They set up civilian headquarters in Abu Znema, north of al-Tur, built army camps and an air field, constructed roads along the Red Sea coast, and began to exploit on a commercial scale the oil fields, and the gypsum and manganese deposits found near the Red Sea coast.

‘Tribal Organization’ as Ideology

The Bedouin of South Sinai consider themselves to be tribesmen, in two senses: first, they belong to an administrative unit at the head of which stands a chief (sheikh) chosen by the members, whose appointment is confirmed by government. Second, membership of the tribe confers various territorial rights: to “build a house” anywhere in their roughly defined territory, which in practice means settling in an oasis; to have preferential access to pastures and to employment in it; and to participate in smuggling activities passing through it. This last is a recent transformation of the former monopoly on conducting travellers through the territory (Neibuhr 1799, 1: 141,153). The tribe claims not so much to control a clearly bounded territory, but defined points in it and paths leading through it. When asked to describe their tribal territory, Bedouin present one with a list of salient points, such as oases, wells and pastures, but there is much disagreement between their statements about boundaries and considerable overlap between tribes. This is to be expected in a country whose practical value to the Bedouin is located in particular tracts. At any moment a considerable proportion of the men may be working outside the territory, and their flocks herding in the grounds of other tribes. This did not deter administrators and travellers from sketching tribal maps which neatly apportion all of South Sinai between the tribes (Murray 1935: 247; Israel Army 1962: map 3; Glassner 1974: 35).

The Bedouin conceptualize.their social organization in two ideologies: the tribe and the annual pilgrimages. Each ideology refers to different aspects of social life, and is invoked in different contexts. And yet they seem to complement each other in important ways, which will be discussed later.

The so-called “tribal organization” is, on one level of discourse, a device through which people conceptualize their belonging to the tribe and their rights of access to the resources of the tribe, and to the pastures of South Sinai. The individual becomes a tribesman through belonging to two groups. First, he belongs to a small group of agnates, some of whose members reside in the tribal area, while others may be outside it. Second, through this group he connects up with a more inclusive descent group which is associated with a vaguely defined part of the tribal territory; and third, he views this group as a sub-division of the tribe, the group that owns the territory. The transition from the agnatic group to a territorial one is accomplished in several steps, because each group serves a different function. Yet the conception of the tribe as a homologous series of ever-widening groups is valuable to them, for it allows them to visualise the tribe as a combination of small forces capable of defending territory, in a country where the central government is far away and rights are maintained by might. The tribal chief does not necessarily figure in this conception of the tribe. Bedouin usually describe the chiefs as “government chiefs, who really represent only thier personal interest”. They do, however, admit that the chiefs act as liaison officers between the government authorities—and in the case of the Jebaliyeh also between the monks of Saint Catherine’s—and the tribesmen.

In more detail, the tribe is made up in this manner: each Bedouin is born into an agnatic descent group. The group called a family (’eleh), is named after an ancestor who lived two to four generations ago and bears either his real or his nickname (naqbah or nabadh). Some people cannot trace the details of their connection to this patronym. Bearers of the patronym do not necessarily camp together or own joint property, and there is no formal leader. Some of the bearers of the patronym, often close agnates, may, in company with cognates or even non-relatives, engage in joint economic ventures, for instance in smuggling. Others, and even the members of the economic partnership themselves, may then consider this as an activity of the agnatic descent group, oblivious to the fact that not all members participate in the joint activity and that some non-members are involved.

Several patronymic groups are affiliated to a large unit, which is, at one and the same time, considered a large descent group and a sub-unit of the tribe and which, supposedly, is associated with a not too clearly defined part of the tribal territory. This unit is called a quarter (ruba’, pl. rubu’), or branch (fara , pl. furu’), indicating that it is a sub-division of a larger unit, the tribe. Tribesmen freely admit that their patronyms are not necessarily descendants of the ancestors of the ruba ’ and that some tribesmen originated outside the tribe and joined the ruba’ within living memory. The members of the ruba’do not all reside in the territory associated with their ancestor. A respected member of one of the constituent groups—not necessarily of a large or wealthy one—is elected elder, and his home becomes a meetingplace for members of the group. This man distributes resources allocated to the group by external agencies, such as work supplied by the government, and in this context acts as representative of the tribal chief. He also makes the arrangements for periodical reunions of the group. The tribe is conceptualized as a group founded by a single eponym, whose sons—and as the term ruba’ implies, there are often four—are the ancestors of the ruba’. This tribe, as distinct from the administrative tribe headed by the chief, exists exclusively as an organisation for the control of territory and initiates few joint political or economic activities. Bedouin often argue that the tribe as a whole pays blood-compensation (diyah), but no instance has come to my knowledge. The tribesmen are united by joint ownership of territorial resources, and some of them combine forces to prevent encroachments of outsiders. Thus I was told that some years ago an Awlad Sa’id man had constructed a stone house near a route which the Jebaliyeh claimed as their own. One of the Jebaliyeh elders warned him that he was welcome to set up a tent in that place, but not a house. The Awlad Sa’id man moved away, and the house was then torn down. The tribesmen do get together only at the annual pilgrimage.

The genealogies of the tribes and the rubu ’ may remain unchanged for many generations. They may, however, alter because of territorial realignments. Patronymics change in accordance with demographic processes. Individuals and patronymic groups move from ruba’ to ruba’, but do not usually shift from their tribe to another.

Bedouin often migrate relatively short distances in their tribal territory. But they claim that “membership of the tribe also gives them access to all the pastures of South Sinai”. They consider their tribes to belong to an alliance, which “in the past elected a common leader and moderator”, and together they are called the people of Mount Sinai (Tawarah, s. Turi). The confederation is not represented in the genealogies by an apical ancestor. Government occasionally appointed one of the tribal chiefs as paramount chief, but this appointment was mainly ceremonial, and the chief never exercised control over the population. Pastures are frequently shared by members of several tribes; whenever this happens, the herdsmen may camp together, irrespective of their tribal affiliation.

Six tribes are usually considered members of the alliance of Tawara: the Muzeneh and Alequat who together constitute one moiety, the Sawalha, Qarasha and Awlad Sa’id—the other; and the Jebaliyeh, who are slightly inferior members of the alliance. Two other small tribes, the Hamada and the Beni Wasel, are viewed as remnants of the ancient inhabitants of the country who lost their land, but are tacitly accepted as belonging to the Tawara. Only the Huwetat tribesmen are thought to be intruders without right to land; and indeed most of them appear to have entered the area as late as after the First World War. They only do not perform the annual tribal pilgrimage.

Here is a list of tribes arranged by size. The figures are based on Ben-David (1972: 113), taken from a census carried out by the Israeli authorities in 1968. Later information largely confirms these figures, and I made only-slight corrections.

Bedouin Tribes of South Sinai

The Muzeneh are the largest tribe and also occupy the largest territory. Murray (1935: 265) describes them as camel and sheep breeders many of whom “have lately taken to fishing”. Their control of the east coast put them in an ideal position for receiving narcotics from Saudi Arabia, and passing them on to other middlemen in Sinai. The final destination was usually Egypt. Smuggling thrived until about 1970, when it was stopped almost completely by the Israeli authorities. Some of the best known organizers of smuggling bands stem from this tribe, and Bedouin estimate that about 30 per cent of its income was derived from smuggling. There are some fishermen and sailors among the Muzeneh, but only few tribesmen own flocks of goats and sheep large enough for subsistence. Bedouin claim that 50 to 60 animals are the minimum needed. One of the few men I met who owned a flock of that size claimed that he obtained an average monthly revenue of IL 400, in addition to which he obtained IL 200 from National Insurance. Altogether he had a monthly income of IL 600 (approximately £40). Most households live on the wage labour of one or more males, who earn IL 40 a day or more (nearly £3). While almost every household raises five to six goats or sheep, most people claim that they do not expect to make a profit on animals. For part of the dry season the animals are fed millet or corn, imported from outside Sinai. The small flocks are herded by women and girls. This seems to have been the situation for a long time, since H.S. Palmer (1906; 69–72) found in 1869 that flocks were tended exclusively by girls, that men earned their living away from home (though not as wage labourers but as caravaneers), and that they bought corn with the money earned (similarly E.H. Palmer 1871: 81–82; Keller 1900: 26).

While their movements are influenced by the needs of their animals, Bedouin must also take into account other requirements. Women and girls are expected to return home in the evening and perform household chores. Therefore, theyxio not venture too far afield with their flock. Camps must remain close, but not too close, to water, usually at a distance of 2—3 kilometres from it, and near sources of firewood, such as broom (ratam), acacias (seyal), or tamarisks (tarfa), and amenities, such as shops and roads. Therefore, they cannot provide the most favourable herding conditions for the animals, and cannot make their flocks pay.

The Awlad Sa’id live in the mountains west of Saint Catherine’s. They straddle all the roads along the east-west axis, and control some of the best pasture areas. Only a small area of high mountains has been left to Jebaliyeh tribesmen. The Awlad Sa’id, too, used to engage in smuggling. They rely to a somewhat greater extent than the Muzeneh on flocks, and not a few of them possess small and relatively neglected orchards in the mountains. In the past their camels carried pilgrims from Egypt and supplies from al-Tur to the monastery. Now they, too, depend largely on wage labour.

The habitat of the Jebaliyeh is the high mountain. Their flocks are even smaller than those of the other two tribes, and there are households without animals. As their small territory includes no major mountain pass, they are more or less excluded from smuggling activities. They specialise in horticulture. Every tribesman owns an orchard in the high mountains, which is regularly watered and cultivated, even when its owners live elsewhere. Tribesmen claim that in the past they used to market their fruit in al-Tur, and with the money they just managed to buy a year’s supply of grain, but could not afford other items such as tea, sugar and clothing. So the tribesmen sought employment in the area and outside it. Saint Catherine’s monastery traditionally employs between 20 to 30 Jebaliyeh tribesmen and also a few Awlad Sa’id. While its workers are paid much lower wages than those in the labour market, it can offer them two advantages:— work near home and work that is relatively secure. The monastery’s employees often remain in their jobs throughout their working life, and even after retirement the monastery gives these men small but regular food rations. Tribesmen complain about the meagre pay, but remain in their jobs, for no-one knows what the future has in store. The Jebaliyeh also used to supply guides for pilgrims who wished to visit the many sites in the vicinity of the monastery associated with the works of Moses.

Jebaliyeh “explain” that their close ties with the monastery go back into history. Their ancestor was a Greek slave, named Constantine. When the monastery was established, he was one of a number of slaves sent out to serve the monks. This story both establishes the link to the monastery and the tribe’s claim to a territory “for it is owned by the monastery”. While individual tribesmen admit to various origins, their attachment to one of the tribe’s four branches gives them a right to enjoy the advantages of their links.

As their territory is too small for herding and as the monastery could employ only a few men, the Jebaliyeh were among the first wage labourers. As a result some of them acquired skills, and as builders and well-diggers worked all over the peninsula. They still rely on wage labour for most of their income, and the men regularly work in the towns or the Abu Rudes oil installations. Between 30 to 50 men are usually employed by the Israeli administration.

Jebaliyeh households move as a rule between two locations: for most of the year, and especially during the cooler months, they stay in the “low-lying” areas at a height of about 1,600 metres; during the hot summer months they move into the mountains to enjoy the fresh air and the fruits of their orchards. A similar transhumance is observed by the Awlad Sa’id but it takes place at a lower altitude. Their cool summer sites are at about the same height as the “sheltered” winter sites of the Jebaliyeh.

All three tribes, then, depend on wage labour for their living, and men spend long periods working away from home. At the same time, they behave as if they were pastoralists or gardeners, although many people admit that they either make little money out of their flocks and gardens, or even lose it on them. Why do Bedouin then resist moving their families nearer to the places of work? The answer is twofold. Some Bedouin who established themselves in secure jobs settle in the towns and cities of Egypt, and are often followed by kinsmen. But these people are generally out of the view of the other tribesmen, who maintain the fiction that they are always drawn back to their homeland. A Jebaliyeh man put it thus:—

“Our country is arid, nothing grows in it. Last night the dew froze and burnt (sic) my tomato plants. The same happens to the almond and apple blossoms. Nor is it a good country for goats, you always have to supplement their food and lose on them. Only its landscapes are beautiful and in summer the climate is good, but our livelihood (rizq) is obtained outside. I always return to it though, because my family lives here. And I do not leave because my country is dear to me. I am tied to it by my navel-string. For when a child is born, the father buries the navel-string and the placenta deep in the ground”.

The sentimental attachment of this man and others to the land conveys its importance as a secure base. When compared to the income obtained by wage labour its resources are of little value. The gardens and flocks are maintained although they do not pay. For work is insecure, and the political future uncertain. If anything should go wrong Bedouin could always fall back on the resources of their country and by some serious effort they could increase their garden produce, and they could migrate with their flocks so as to make the best out of the available pasture. Men leave their homes in the tribal area, and allow their gardens and flocks to “tick over”, because these are their economic base. Should the need arise, these could be built up, and provide a small income. This became evident late in 1975, during the negotiations between Egypt and Israel over an “interim agreement” in Sinai. Suddenly there was a flourish of activity in the basic economy. Bedouin stopped the sale of animals and there was great interest in gardens. Some people acquired gardens, others dug wells and planted new gardens. Thus they prepared for the possibility of losing their work.

Smuggling is to be viewed as a secondary accommodation to prevailing conditions. As people must stay in this wild and inhospitable country, they are available for any work that presents itself there. In spite of high risks, and the not extravagant wages, numerous people were engaged in the business and viewed it as an accepted and respectable way of making a living. Only a handful of men made fortunes out of smuggling. Smuggling reinforces tribal allegiance as the control over routes is a precondition for entering the game.

Annual Migration and Pilgrimage

Uncertainty about the political and economic future, then, ties these people to their country and to the tribe, and makes them behave as if they were pastoralists and gardeners in the accepted sense. As a result, not only do many men spend a great deal of time away from their families and kin, but families are also widely distributed in small camps. While individuals move about a great deal, for much of the year they meet only some of the people on whom they depend for mutual assistance and reassurance. Relationships require maintenance. When persons do not interact regularly, they lose confidence in their relationship. Would it stand the test of a critical situation? Therefore, each Bedouin strives to visit his relatives and friends, and the opportunity to see many of them at one time presents itself in late summer. At that time the dates ripen, and the Bedouin congregate in the oases until the harvest is over. These large concentrations, in turn, form the basis for the organization of tribal gatherings.

Both the Muzeneh and the Awlad Sa’id go through annual migrations. Their pastoral year begins in spring (rabi’), the time when the weather becomes a little warmer and verdure springs up. This usually happens around February. In the mountains spring comes round every year, but in the lower areas only if there has been some rain in winter. Then the tribesmen leave their winter quarters, relatively large stationary tent camps, and some of them divide up into small camps of two—three tents in order to exploit the pasture wherever it is available. Even hquseholds with few animals may join in these movements, which are largely determined by the requirements of the larger flocks, because they wish to be close to relatives. Other catnps may be larger and remain almost immobile. They are made up of households whose income depends on wage labour, and who put little emphasis on herding.

Spring pasture does not last long and the owners of flocks move into the few reliable early summer pastures, such as Wadi Rahaba and Wadi Slaf. The stationary camps move a short distance from the sheltered winter camping site to one that allows in the refreshing breezes.

The annual migratory cycle reaches the point of maximum concentration in summer. The pasture is exhausted and water sources dry up. People converge on the remaining wells in the oases or on the sea-shore and harvest their dates or, in the case of the Jebaliyeh, repair to their gardens in the mountains. Here tribesmen gather in large groups, there is much visiting and among the Muzeneh young men organize dances nearly every night. After the fruit have been picked and the time comes to move, the elders of the tribe call for a tribal gathering. The gathering not only fits into the seasonal pattern of maximum concentration of tribesmen, it also enhances it. This is the only occasion when the ‘tribe’ visibly becomes a group. At other times it does not operate as one political unit, though members are frequently aware that they act as tribesmen, for instance when they plant gardens, build houses or engage in smuggling.

The tribal gatherings, like the individual pilgrimages to saints’ tombs, are called “visits” (ziyara, locally pronounced zuwara). This is the term used elsewhere, too, for pilgrimages to holy places, with the exception of that to Mecca, for which the distinctive term hajj is employed. While the pilgrimage to the summit of Jebel Musa is often compared with that to Mecca, it is nevertheless a zuwara. Each year a number of men and women make the still arduous and costly journey to Mecca.

What is it that makes a considerable number of tribesmen participate in the communal or individual pilgrimages? There is very little persuasion involved; the organizers of a communal pilgrimage are respected members of their communities, but few of them are tribal chiefs or leaders with a strong hold on people. Their main preparations are to collect from the people in the neighbourhood contributions for the tribal sacrifice, which are eventually returned to the donors in the form of boiled meat, and jointly to set the date of the pilgrimage. An attraction of the pilgrimage is that the individual considers it as beneficial. The reputations of the major shrines have spread widely. Not only the tribesmen on whose territory they are situated believe in their powers, but also many others. Some shrines are known to Bedouin from central Sinai. All these places attract individual pilgrims from all over the peninsula; The eves’of Monday and Friday are considered as propitious for pilgrimage. On these days, throughout most of the year, individual households make their way to a shrine of their choice. In many instances this will be one of the holy tombs, but there are also a few other sites, among which the summit of Jebel Musa (where Moses is said to have received the Tablets of the Law) is the most venerated. In these pilgrimages people either redeem vows made when in danger or make requests. As the saint only mediates requests to almighty God, a visit to any holy tomb, perhaps the closest one, should do. Yet, people are slightly sceptical, and trust their own experience and that of their relatives, which shows that some saints intercede more efficiently in some matters than in others. As individual experience varies widely, no real specialization has developed among the saints. But one result has been that people from all parts of South Sinai visit all the holy tombs, and that they often range far afield. Thus members of one patronymic group of the Jebaliyeh visited at various times during the last four—five years the following sites: A women who had not conceived for two years ascended Jebel Musa with her family; a man sought to protect the wellbeing of his family by taking them on an outing to Sheikh ‘Awad; two men took their flocks by car to Sheikh Habus in order to safeguard their health; one man carried his family by camel to Nebi Saleh, on recovering from an illness; and an old lady regularly expressed her wish to make the expensive pilgrimage to Mecca “next year”. Similarly, people all over South Sinai visit nearby and distant shrines, and all the saints are respected throughout this part of the peninsula.

This situation affects tribal gatherings in two ways. First, the commonly shared belief in saints’ tombs, and in saints as intercessors with God, is one of the foundations of the tribal pilgrimage. Each of the individuals who participates in the gathering believes that he obtains a spiritual benefit. It must be added, however, that he may also expect material advantages, perhaps meet debtors or offer a car or a camel for sale. Bedouin see nothing incongruous in the seeking of spiritual and material benefits at the same time. They do not conceive of purely spiritual or purely material occasions, as found in specialised industrial societies.

Second, all the participants in the gathering, whether belonging to the celebrating tribe or not, respect the saint’s tomb, and consider it the equal of the saints’ tombs of other tribes. Even Nebi Saleh, whose tomb is the recognised centre of the tribal gatherings, is not considered superior to the others. Individual pilgrims do not prefer him in any way to the others, and often many weeks pass between one pilgrim and the next. This equality of saints expresses an important facet of the gatherings: that they are not held exclusively for the members of one tribe, but that people from other tribes can also attend. Indeed, people say explicitly that the gatherings are expected to attract visitors from other tribes. Thus, the meeting-place of the Bedouin of South Sinai shifts periodically from one site to another. At each gathering one tribe slightly dominates because of its numerical preponderance, but never becomes superior to the others. Not so long ago, people say up to 1965, at least five tribes, the Sawalha, Awlad Sa’id, Qararsha, Muzeneh and Jebaliyeh, held separate annual tribal gatherings at Nebi Saleh. I can only surmise that the location turned this tomb into the central meeting place of the Bedouin of the mountains; it is situated on the cross-roads of the major east—west and north—south mountain passes, and close to water. In view of the great importance of smuggling in those days, each tribe attached major significance to its rights of passage through the mountains, as well as to co-ordination of activities and maintaining peaceful relations with other tribes.

Today each tribe has its own meeting-place, in the areas it customarily frequents. There are two clusters of such sites. In the west, not far from the sea are the pilgrimage centres of the ’Aleqat, Qararsha and Hamada tribes. They are located close to the main concentrations of the population, whose employment is found in the towns in and around Sinai. Before the Israelis took over, some people worked in the mines and industries of the area, but the majority worked all year round as migrant labourers, and approximately half of them eventually settled in Egypt proper. For these mobile people, the saint’s tomb is close to home, as it points out their claims to the territory. The other cluster of pilgrimage ventres, which will be examined in greater detail, are those of the Muzeneh, Awlad Sa’id and Jebaliyeh. They are located in the mountains, at the edges of each tribe’s territory, and here pilgrimage requires most people to leave their homes and to move elsewhere. They go through a rite of passage, “beginning in a Familiar Place, going to a Far Place, and returning, ideally, ‘changed’, to a Familiar Place” (Turner 1973: 213). These tribesmen are already tied to the land, for here their basic economy lies. Here, where the attachment to the land is given, Bedouin seek in their pilgrimages to forge or maintain relationships in the tribe and outside it. The individual is drawn away from his round of daily life and its circumscribed field of relationships, and brought to another place where he meets people from a wide area.

The date of the tribal gathering is set by a few elders, who take into account all the relevant factors. A Bedouin explained these;

“The date of the gathering is fixed perhaps two months in advance, so that the news will get to all tribesfhen. Conditions of work are taken into account as well as the end of the date harvest (masif) in Wadi Firan (the dates there ripen later), and the position of the stars. Only after canopus (thraiah) and other stars rise, the mutton becomes suitable for eating, before that the goats are too lean and their meat is hard to digest”.

To complicate matters further, the feasts are held on Thursday afternoons, the eve of the Muslim day of rest, and during the full moon.

The dales of the tribal gatherings vary but little over the years. I he earliest reported date gives the end of July (Burckhardt 1822: 489). ‘I he gatherings I attended, or on which I possess details, all took place between the end of June and the beginning of September. In this respect there was no difference between the Muzeneh, the Awlad Sa’id and the Jebaliyeh.

My data on attendance are incomplete, but they do show that in recent years it varied considerably. Ihus for the Muzeneh gathering at Nebi Saleh in 1970, 100–150 men are reported (Mcshel 1971: 95). In following years the numbers declined sharply, until in 1973 only 50 men were reported to be present. In 1975 the figure climbed up to nearly 200 men.

The Awlad Sa’id attended their tribal meet at Abu Taleb more regularly. At the gathering in July 1973 that I attended about 70 men gathered in the rectangular open-sided hut opposite the saint’s tomb, l or the following year Kapara (1975: 193) reports 40–50 men in the big hut, while others were still arriving. In contrast to the Muzeneh, whose attendance declined up to 1973, and went up again in the following year, the Awlad Sa’id pattern was stable. Most of the men attended each year.

I he Jebaliyeh were different again. I ribesmen claimed that in the past they had had annual tribal gatherings which started at Nebi Saleh, who in the eyes of all tribes is the senior saint, and at whose tomb several tribes gathered, on separate occasions, f rom there they used to move on to Nebi Harun (the reputed tomb of Aron, the high priest), near Saint Catherine’s monastery. This gathering was held for the last time in 1965. After that there were no more gatherings of the whole tribe, but in August 1973 I attended a gathering of about 65 to 70 men of the Awlad Jindi sub-tribe at Shekh ‘Awad, and in August 1975 another one at the same place at which nearly all the Awlad Jindi men were present. A month later another sub-tribe, the Wuhebat, held a gathering of their own at Sheikh ‘Awad. I was told that previously they had not had such gatherings.

A clear pattern emerges here: though the gatherings never ceased, attendance declined when more work was available and it was expected that the favourable conditions would continue. This was particularly true for the Muzeneh and Jebaliyeh, who largely depended on wage labour. People who regularly turned up at the gatherings were those who had a predominant interest in herding, and also the former smugglers, who presented themselves as keepers of tradition. When I asked why others did not turn up I was often told that they were away working and would not jeopardize their jobs. So the men who had an interest in maintaining the tribe showed up at all meetings and those whose employment was secure did not. After 1973, the employed people also took part in the gatherings. This was the result of a new kind of uncertainty. At the time work was plentiful, and well paid, but the political future of Sinai was uncertain. The Egyptians and Israelis were negotiating under the aegis of Dr. Kissinger, the Israelis were giving up territory in Sinai, and no one could know where this was going to end. The extended negotiations for peacemaking only brought doubts and insecurity for the Bedouin. An Awlad Jindi man employed by the government as driver had often told me that he never attended tribal gatherings. These were for the old people he thought. In 1975 he attended for the first time and explained that his wife had begged him to go and that he could not refuse her wish. He did not admit, perhaps not even to himself, that the future was uncertain and that he might have to rely to a larger extent than hitherto on the help of agnates and tribesmen. At the time he was toying with the idea of setting up a repair shop for motor-cars. Tribesmen could thus I become more important even as customers.

The two levels of the individual’s involvement in the pilgrimage are shown in the eating arrangements. At each gathering there is a communal meal. Several sheep or a young camel are slaughtered and all the men who contributed to the price attend the meal. The animals are bought and cooking arrangements made by an elder of the group. In the late afternoon or evening all the men gather for a festive meal. They pray together and then sit down in a large circle for a meal. Each man’s name is announced for all to hear, then his portion of meat, wrapped in a flap of flat bread, is passed along a chain of servers up to the announcer who hands it to the man. There is no visible leader at these functions, no speeches and announcements are made. Visitors from other tribes are welcome and also given portions of meat. Chiefs and notables mingle with other men and no special area is reserved for them. At one of the Muzeneh gatherings the paramount chief of the tribe put in a short appearance, to pay respect to a gathering at which he did not preside. At the Awlad Sa’id gatherings the two tribal chiefs were present all the time and moved among the people. The Jebaliyeh chief did not attend the Awlad Jindi meetings at all.

This was the formal part of the gathering of the tribe as a territorial unit. But men did not stay all the time in the circle of tribesmen. All around the saint’s tomb related families set up small circles of their own. Here men, women and children shared in a joint meal. Each group slaughtered a goat or a sheep and in addition to meat other dishes were prepared. The women in particular wore their finest dresses, embroidered with silver and gold thread. The atmosphere was relaxed and women spoke up freely, even making jokes with sexual innuendos. There was much mutual visiting. The men, most of the time, made the rounds of friends, now and again returning for a few minutes with their families. At these gatherings people reaffirmed their friendly relations and, incidentally, settled debts and disputes and initiated commercial dealings such as the sale of animals and cars, and other joint activities. Shopkeepers, too, had arrived and made brisk sales of cola bottles and sweets.

Leadership there was, but it was muted. The tribe as an administrative organisation was not brought into play, so that the appointed chiefs either did not attend or came as visitors. The Awlad Sa’id chiefs, the Awlad Jindi headman and the other elders acted as organisers, they provided a service for members of their group and, just as in daily life, they facilitated communication among them. But at the gathering itself they did not stand out. Thus Kapara (1975: 192–194) spent a day at an Awlad Sa’id gathering and did not notice the chiefs.

The Symbolism of Saints’ Tombs

The significance of the tribal gatherings has gradully emerged. There remains the question why the meeting-places should be saints’ tombs. I shall examine the symbolism of these holy places, in the hope of obtaining an answer, even be it an incomplete one. I take a ‘symbol’ to be a sign replete with many interrelated meanings. In this working definition, which bears resemblance to Sapir’s (1934: 493) characterisation of the ‘condensation symbol’, I wish to stress the richness of meaning as an identifying mark of the symbol, as opposed to the mechanical multiplication of mental associations that may attach to a particular sign. The meanings of the condensation symbol express, as Sapir (1934: 493) pointed out, “a condensation of energy, its actual significance being out of all proportion to the apparent triviality of meaning suggested by its mere form”. To this I would add that much of the emotional impact of ‘condensation symbols’ derives from the fact that they evoke numerous meanings simultaneously. The secret of this immediate impact probably lies in the symbol’s full integration in the daily routine of its users. Because of this, many students of society today agree that an external observer can tap the multiple meanings of symbols only through intimate knowledge of both native conceptions of their society and of the social contexts in which the symbols are invoked. The whole gamut of a symbol’s contextual uses, as well as the associated symbols and contexts, must be explored, before it is properly understood. But once this is done one may find that each segment of the meaning of a symbol has a degree of precision that is often unsuspected.

I shall attempt to unravel three of the major strands in the mesh of meanings attached to the saint’s tomb, which have until now remained in the background: the significance of the building erected over the tomb, the tomb itself, and details of the saint’s person.

I would argue that the building is a very precise image of the Bedouin’s conception that territorial rights are vested in the tribe, and that their essence is the tribesman’s exclusive right to build a house on tribal land. Geertz’s (1968: 49) concise description of the Moroccan shrine fits that of South Sinai too: “[It] is a squat, white, usually domed, block-like stone building set under a tree, on a hilltop, or isolated, like an abandoned pillbox, in the middle of an open plain.” (See also Canaan 1927: 11). One might add that each shrine has a door, and usually also a small window. Only the dome, often crowned with a crescent, and in a few shrines a prayer niche facing in the direction of Mecca, distinguish the shrine from an ordinary house, and indicate that it serves a religious purpose. We already know that Bedouin consider the building of a house as the essential territorial right, and do not permit outsiders to build houses in areas controlled by the tribe. The shrine signifies, among other things, the territorial claims of the tribe whose pilgrimage centre it is. As other tribes may also visit the shrine, the owning tribe’s rights obviously must be limited, while others obtain rights to pass through the tribe’s land and to pasture their animals on it. On another level, the use of the saint’s tomb as a symbol of tribal ownership indicates that territorial rights are protected by supernatural sanctions, but not by the might of the whole tribe. Bedouin store property, such as tents and farming implements, inside or near the holy tomb, secure in the belief that it is protected by the saint. They recount stories about misfortunes that befell would-be thieves, until they returned the stolen property to the tomb. Construction of a saint’s tomb or repairs to one, are acts intended to claim rights in land. When Ghanem Jum’a, a wealthy Muzeneh, built the shrine of Faraj, the tribe’s ancestor, in the 1960s, he reasserted ‘the tribe’s sovereignty over the area which was at the time becoming the main supply centre in South Sinai, and the major sales depot of the mobile merchants from al-’Arish who control nearly all the commerce in the peninsula. Similarly, Jebaliyeh elders every few years invite their tribesmen to subscribe money for repairs to the shrine of Sheikh ’Awad. This building activity reasserts a tribe’s rights to the holy tomb, and through it—to territory; and incidentally, in this way all the shrines are kept in a good state of repair.

The saint’s tomb occupies the centre of the building, to allow visitors to circumambulate. The tomb rises to a height of about three feet. It is covered by a shroud made of white or green cloth, on which sometimes the Islamic creed and the saint’s name are embroidered. There are usually several layers of shrouds, for each time a shroud is worn and torn it remains in place and a new one is placed on top of it. It is always doubtful whether the holy tomb actually contains a human body. While the tomb of Nebi Saleh is solidly built and one cannot therefore know whether anyone is buried in it, the tomb of Ali Abu Taleb consists of a flimsy wooden frame, just sufficient to carry the layers of embroidered shrouds. Canaan (1927: 22) remarks, with regard to Palestinian sanctuaries, that “the tomb is often not in the shrine, but outside of it ... it is not at all necessary that there should be a tomb ... connected with the place to make it a shrine.” In South Sinai the Bedouin have gone a step further: they insist that no-one is buried in a maqam. As proof they cite the

“well-known fact that Nebi Saleh is buried in Ramla, in Israel, and that even the Sheikhs Habus and Abu Shabib have each two holy tombs. Only one of these holds a grave (turba), while the other is a place for gatherings of men (maq’ad).”

All the shrines with tribal gatherings are, in Bedouin eyes, maqams, irrespective of whether anyone is buried there or not. This attitude releases the tomb of an association with a tribal ancestor. The saint himself is disembodied and, shed of his earthly vestiges, he moves into closer communion with God.

Most people cannot say anything definite about the saint’s person, except that he was a holy man during his lifetime, as proven by instances in which he rescued individuals from various tribes in miraculous ways. Occasionally people hazard guesses about the antecedents of a saint. While some say that Nebi Saleh was a pre-Islamic Arabian prophet, others think that perhaps he was the ancestor of the Sawalha. The same uncertainty seems to have prevailed over a century ago, when Robinson (1841: I 215) reported: “The history of the Saint is uncertain; but our Arabs held him to be a progenitor of their tribe, the Sawaliha.” Sheikh ’Awad is also said by some Jebaliyeh to belong to their tribe, while the Muzeneh consider Sheikh Abu Zaid to have been a Sawark Bedouin (from North Sinai). The only saint who appears to be an acknowledged tribal ancestor is Faraj of the Muzeneh. But his tomb is not reckoned among the maqams it is named after a group, the Faranja, and not after the ancestor Faraj; and only in recent years was a shrine built over the tomb. Yet the fact remains that Faranja is one of the Muzeneh’s three pilgrimage sites, and one has to conclude that, in a sense, the saint is an emblem of the tribe.

The three holy tombs visited by the Muzeneh allow us to define more clearly the symbolic significance of the saint, for the person of each of the three saints gives special emphasis to one aspect: Faranja symbolizes the unity of the tribe, as well as its descent from the North Arabian Muzeneh (Doughty 1937: Il 381). Nebi Saleh represents both the unity of the Tawara, of South Sinai as one region to which all its inhabitants have access, and the inclusion of South Sinai in the Islamic world and, by extension, in the state that control Sinai. Nebi Harun again stands for the connection with the monastery of Saint Catherine’s and the privileges associated with it on one hand, and on the other—for the Islamic community. For Moses and Aron are included in Islamic hagiography. I cannot explain this specialization of the three Muzeneh saints, for in other tribes all the symbolic aspects merge in a single saint.

On a higher level of abstraction, the three Muzeneh saints stand for ideologies that are deeply engrained in Bedouin thought and concern areas of their social life: agnation and tribal organization, the saints’ cult and the regionalism of South Sinai, and the universalism of Islam and, by extension, all the external world impinging on the Bedouin and not controlled by them. These are not, of course, the only ideologies employed by Bedouin. Even in the saints’ cult another ideology, namely kinship, plays an important part: it underlies the belief in the efficacy of the saints’ intercession with God, and it is expressed in the family sacrifices that are made parallel to the tribal ones.

The ideology of agnation informs “bounded” activities, such as the control over rights in land, protecting property and the setting up of groups. To this kind of activity applies Robertson Smith’s (1894: 150) dictum that holiness is “a restriction on the licence of man in the free use of natural things”. In the Arabian context, he believes that “at most sanctuaries embracing a stretch of pasture ground, the right of grazing was free to the community of the god, but not to outsiders” (Smith 1894: 144). In South Sinai the rights thus reserved are those over oases and their fruit-trees, irrigation from wells and the tribal smuggling routes, and they are compressed in the right to build a house on tribal land.

For each restriction on the use of resources there must be a corresponding opening up of resources, as exemplified in this statement, also taken from Robertson Smith (1894: 111):

“The Arabs regard rain as depending on the ... seasons, which affect all tribes alike within a wide range; and so when the showers of heaven are ascribed to a god, that god is Allah, the supreme and non-tribal deity.”

Evidently these tribesmen reserve certain rights in territory for themselves, but also grant rights to pasture on it to others, on a basis of mutuality. Robertson Smith cannot admit that the “tribal deity” also stands for Allah; to imply that the same people believe in two gods, one tribal and one supreme, at the same time, is as far as he dare go. In South Sinai the saint is clearly associated with God. While he intercedes with God on behalf of his tribe, he also works on behalf of individuals from any tribe, and is thus never an exclusive patron of his own people. He represents both restrictions on the use of the country’s resources in favour of some people, and access to the remaining resources to others. There is no paradox in this, for rights are always specific, so that every prohibition has its complementary permission. Thus in South Sinai certain rights in land, wells and roads are reserved for tribesmen, and other rights are expressly opened up to all Tawara: land everywhere is available to anyone for pasture, drinking water for men and animals may be drawn from any well, even when surrounded by a fence, and every Bedouin can travel freely anywhere in the peninsula, as long as he does not infringe local privileges of employment or uses another tribe’s smuggling route.

The ideologies of agnation and of the saints’ cult not only complement each other; they are also linked up, through the person of the saint. The ambiguous position of the saint, who is sometimes thought of as a tribal patron, perhaps even an ancestor and at others just a holy man about whom little is known, or maybe even a Bedouin from a distant tribe, is evidence of his two faces. One pulls him toward the tribe, the other towards the world.

References

© Emanuel Marx 1977

Part Two: Regional and Non-Regional Alternatives: The Waxing and Waning of Cults

3. Disparate Regional Cults and a Unitary Ritual Field in Zimbabwe[12]

*by Kingsley Garbett

“I am a well-built house in which the little spirits shelter ... Nyabapa, Gweera, Chingoo, Goronga and Chaminhuka are the poles and fibres which built me, the Outspread One, who stretches out over all the earth. The lesser spirits are the fibres, but my children are the poles.”

George Kupara, speaking as Mutota.

Introduction: The Interpenetration of Cults

Perhaps the greatest temptation in a discussion of regional cults is to treat each one as a separate entity. This arises because, what strikes an observer immediately is the presence of sharp cultural and organisational differences. Often, regional cults that are closest in location are quite distinct in kind. But to yield to this temptation easily leads to a wrong reading of the dynamics of the cults. Here the fundamental problem is how one cult or region waxes and wanes in relation to others. Not that this necessarily requires observation of a series of cults. Admittedly, to cover in depth as much as one cult or even a single region is, often enough, beyond the reach of a lone fieldworker. But the need is to bring one cult or region into focus within the context of its relations with others. It is essential to take account of the interpenetration of cults within a wider field of ritual relations. Only thus can analysis reach a right explanation of transformations in a single cult along with changes in the connections between and across the cult’s regions.

In this paper I centre my analysis on the expansion of one region in Zimbabwe, that of the Korekore ancestor Mutota. I take into account ritual relations among people who are ethnically distinct neighbours, the Korekore, Tavara, and Zezuru, but who speak dialects of the same language and take part in the same wider field of political and economic relations. In its spatial extent this field extends over an area of about 30,000 square miles and includes several million people. My analysis shows how two quite unlike cults interpenetrate and connect with each other within this wider field. In some respects the two kinds of cults do resemble each other. Both are broadly concerned with regulating man’s relation to the earth and with matters of general morality. Both involve beliefs in spirit possession and are organised through spirit mediums. Given these similarities, however, the cognitive and organisational differences appear all the more striking.

To take the cult of territorial spirits first, this includes Mutota’s region along with those of at least two others, Chingoo and Nyabapa. The cult’s domain among the Korekore extends throughout the Zambezi valley between the Angwa and Mukumvura rivers, northwards into Mozambique,[13] and into part of the adjacent southern escarpment (see Map 2). Here all land, whether now occupied or not, is divided into ritual areas which are regions and parts of regions, ‘spirit provinces’ in my terms. These spirit provinces vary considerably in size from a few hundred acres to several hundred square miles in extent. Each is associated with a territorial spirit who is believed to be the remote ancestor in control of the territory from ancient times to the present. The boundaries which demarcate each spirit province are well known. Moreover, they have an autonomy of their own. They are not adjusted to fit other divisions, such as current political ones. They do not necessarily coincide with the boundaries of chiefdoms or with modern administrative and international boundaries.

As for the territorial spirits themselves, some are said to be mhondoro (‘lions’). Among Korekore speakers this designation is-reserved specifically for those remote ancestors who, according to belief, ate special medicines (mbanda) when alive, and thus ensured that their spirits would return to inhabit animal hosts, usually lions, but occasionally pythons and birds, and speak, from time to time, through human mediums (svikiro). Each of these remote ancestoral spirits has a place in a genealogy and thus in a conceptual hierarchy. Similarly, the spirits’ mediums are hierarchically ranked and allocated to the defined territories over which their possessing spirits are thought to have particular control. Mediums are usually installed as young men and remain in office until they die. Very occasionally, one hears accounts of mediums being ousted from office because of some gross misdemeanour but, in general, a medium’s position is secure. Each medium is assisted by an acolyte (mutape or nechombo). This is an hereditary position and, when a medium dies, it is the duty of the acolyte first to announce the death to the senior medium of the region, then to repair and maintain the spirit house (zumba) in which possession occurs, until a new medium appears.

In ritual contexts one never hears the term for the High God, Mwari, and a region’s apical ancestor, such as Mutota, is regarded as a creator being. The objects of the cult are, thus, the remote ancestors who, according to myth, established each region and its spirit provinces. Significantly, they did so through conquest, according to the origin myth of Mutota’s region. In myth, too, expansion must be explained, but it is made out to be a matter of the remote ancestor’s ancient history, rather than a current process, as I shall show it to be.

The second and contrasting cult is that of local spirits, such as Nehanda and Chaminhuka to name only the most prominent. Its domain among Zezuru extends broadly through Central and Western Mashonaland to include the administrative districts of Hartley, Marandellas and Mrewa, and the southern parts of the districts of Lomagundi and Mazoe (see Map 2). Some of the most important of the local spirits, such as Chaminhuka and perhaps Nehanda, are held never to have been mortal and are believed to mediate between man, lesser spirits and the High God, Mwari.[14] Other local spirits are chthonic beings of the underground and pools or ancestors without known genealogical links to the living; and still others are important ancestors of chiefly lineages whose influence extends beyond their original chiefdoms and descendants. Moreover, some of the spirits are considered to be more powerful or more pervasive in their influence, and to this extent there appears to be a conceptual hierarchy, albeit an ambiguous one, in the cult of local spirits. However, there are no clearly defined and regulated hierarchies of mediums associated with particular centres. Nor is there always a single medium for one spirit. A number of mediums, often in widely separated areas, may be accepted as subject to possession by the same spirit and yet not clearly linked with each other. Finally, and perhaps most critically, for purposes of an initial comparison with the cult of territorial spirits, there is the difference in ritual jurisdiction over blocks of land and particular parts of the earth. In the cult of local spirits, specific mediums may be communal ones closely associated with certain sets of chiefdoms, but the local spirits themselves are not, in general, figures of territorial authority. Conceptually, these spirits, quite unlike the territorial ones, lack distinct territories of their own. In terms of cult organisation, their mediums lack jurisdiction over clearly defined ritual areas, other than those which correspond to current political divisions. In brief, the two cults are like opposite sides of the same coin. Viewed in the abstract, as kinds of cult, one appears hierarchical and territorially defined, and the other virtually the contrary, non-hierarchical and territorially undefined, though sometimes communally or locally defined. However, to go beyond this appearance in order to understand how the cults interpenetrate and how one transforms into the other is the main aim of this paper.

Lest my argument be taken to be merely another analysis of an oscillation between ideal types, I must stress from the outset that I am concerned both with the regional expansion and the interpenetration of cults. Briefly in advance of my argument, and perhaps too briefly, three of the ways in which the cults interpenetrate each other’s organisation need to be highlighted. First, there is a tendency towards connection through at least one shared inner circle of leaders: at least one set of close patrikin gains key or senior mediumships in regions of both cults. Second, there is cross-cult affiliation: mediums in one cult may enter into a formal relationship and attach themselves to those in another. Thirdly, cross-cult confirmation occurs: mediums of one may validate the authority or legitimacy of some of the other’s mediums. It would take me well beyond the limits of this paper to consider fully the cognitive management of such organizational interpenetration. Clearly, there must be to some degree a translation of one set of beliefs into another, given the cultural differences in belief in local and territorial spirits. To show how this and other cross-cult problems of cognition are managed along with organisational problems, I shall examine the situation of a medium’s seance. My account will document in particular one such translation of beliefs and the way that Zezuru spirits are presented as, and transposed into, the remote ancestor spirits of Korekore. At various points, I deliberately extend my analysis to the limits of the data. Occasionally, I go beyond the data to indulge in informed speculation. This is particularly so when I use data from secondary sources. I am dealing with a large and complex field seen mainly from the perspective of one cult. Many of the leaders and important mediums of other cults I did not meet: they lived and operated many miles away from me. Yet to ignore them, and the connections between them, leads to a failure to comprehend significant issues concerning the interpenetration of Korekore and Zezuru cults.

The Region of Mutota

The Korekore region of Mutota is but one of the cult regions which exist in and near the Zambezi Valley. The various regions are formally connected to each other by perpetual kinship (see Genealogy 1); each region’s senior medium, and occasionally lesser mediums, take on relationships held to have existed in the past between prominent ancestors in the regions. In Mutota’s case, perpetual kinship links his region to those of Chingoo and Nyabapa: he is held to have had these two ‘brothers’. The relative seniority of these ‘brothers’ is not known or recognised and they are presented as equals. Chingoo was recently represented by a senior medium operating with a small hierarchy of mediums under him in the administrative area known as Sipolilo Tribal Trust Land. Similarly, the medium of Nybapa heads a hierarchy of mediums in the Kandeya Tribal Trust Land near to Mt. Darwin.

Once or twice a year messengers travel between the senior mediums. This maintains formal relations between the three ‘brothers’ through a symbolic procedure: they never meet face-to-face. The messengers carry small containers of snuff (nhekwe) into which a senior medium has ‘breathed’ a message. In a seance, the container is unstoppered, the message ‘listened to’, and a similar ‘message’ is breathed into the container. The messenger is then sent on his way. The same procedure is used within a regional hierarchy: similar messages are regularly exchanged between mediums in Mutota’s hierarchy, for example. However, within a region such ‘messages’ when initiated by a junior medium are transmitted to the regional medium more often through a chain of mediums more senior in the hierarchy, and often the messenger carries a brief verbal message about ritual matters in addition. Between regions the symbolic exchange of ‘messages’ expresses formal equality among regional mediums: within a region it expresses and emphasizes hierarchical relations.

Messengers, who receive a small fee, are chosen from a medium’s staff or from among trusted associates. Significantly, a messenger bearing a snuff container should be allowed to pass through any territory without hindrance: it should be sufficient for him to show the snuff container and to state that he is a messenger of a mhondoro for him to be given safe passage. Nor should he be questioned as to his mission as an ordinary traveller might: it is considered secret.

The need for the formal exchange of messages among mediums arises, as Korekore view it, from the belief that it is mystically dangerous for two or more spirits active as mhondoro to occupy simultaneously the same ritual space. By extension, it is believed to be mortally dangerous for an established medium to come into Ihe presence of, or enter the territory of, another medium without both having taken extensive ritual precautions. Thus, a medium may not cross the boundary of his spirit province without taking such precautions; in some cases he may even have to be blindfolded and carried over the boundary. When established mediums visit the senior medium of the region, they stay outside his hamlet at a special place until, protected by medicines, they are called into his presence at a seance.

In addition to expressing the formal relations among mediums, the regular transmission of symbolic messages also serves to convey to a senior medium which mediums within his region still form part of it and which are still functioning. The repeated failure of a medium to send a messenger to the senior medium over a period of two or three years may be a cause for concern and, unless the senior medium meanwhile receives formal notification of the medium’s death, he may well send his own messenger to establish exactly what has happened..

In general, Korekore conceptualize regions as autonomous ritual territories internally subdivided into areas of various sizes: both the region’s boundaries and those of the sub-divisions are held to be ancient and immutable. If pressed, Korekore acknowledge that when an ancestor becomes a mhondoro and speaks through a human medium for the first time, then an adjustment of ritual areas may be necessary, for each mhondoro and its medium must have a territory; there are no itinerant mediums in Korekore regional cults. However, they argue that it is not that a new ritual territory has to be demarcated but rather that an ancient one, long forgotten, has now been rediscovered. When the medium of the recent ancestor Mzarabani (Genealogy 2: Nl) appeared for final testing in 1963, he was told by Mutota’s medium to seek out the land centred on Mzarabani’s grave. This was held to be located on the west bank of the Msengezi river, adjacent to Chuzu’s spirit province (Map 1: VI,k), in what until then had been accepted as the spirit province of Chivere and Chikuyo.

Cognitively and organizationally, there is sufficient flexibility to accommodate a variety of contingencies without the basic conceptions of the antiquity and immutability of ritual boundaries appearing to be challenged. For example, some territorial spirits, such as Negomo, are credited with more than one territory. A medium of such a spirit, therefore, may have the option of living in either of the territories. In fact, there is a general belief that all territorial spirits have additional, small, ritual territories scattered through the region but few of these are known. Korekore argue that the mhondoro will make them known should the occasion warrant it. This belief, and the fact that it is accepted that a medium may operate in a spirit province other than that of his possessing spirit, enables mediums who have been displaced from their own spirit provinces to operate from a legitimate territorial base. For example, the medium of the Korekore Nehanda, whose spirit-province (Map 1: II, 1, m) extends into Mozambique, fled in the twenties from her village there to escape the attention of the Portuguese authorities and took up residence in the land of her ‘brother’ Kagoro. A more complex case concerns the relations between the senior medium of Mutota and Chihwahwa’s medium. In 1958, the senior medium was displaced by the colonial administration from Gota, on the plateau, in terms of the Land Apportionment Act. He chose to settle as near to his spirit province as possible on the southern boundary of the spirit province of Kasekete and Chihwahwa (Map 1: VI,g). Chihwahwa, however, was represented by a medium with a considerable reputation and a large vicinity. He argued that Mutota’s medium had now placed him in mortal danger. The senior medium countered this by arguing that since the whole of the region was Mutota’s, he could settle wherever he chose. Eventually, after a protracted dispute, Chihwahwa’s medium moved in 1962 to take up residence at a site said to be sacred to Chihwahwa within the small spirit province of Chuzu (Map 1: VI,k).

Table 1: Index of Spirit Provinces Shown in Map 1[15][16][17]

Spirit	A	B	Boundaries of Spirit Provinces
Mutota	A1	VII,d-f	N: Escarpment; W: Nyahoko; S: Chimuwere; E: Kadzi
Chiobvu	B1	VI-VII,h-k	N & W: Utete; S: Beyond Escarpment (?) E: Slope break
Sundo	B2	VII,o-p	Unmapped. N: Escarpment
Kagoro	B3	III-VII,k-n	N: Edge of Forest; W: Kadziruve & Msengezi; S: Escarpment; E: Nzou-Mvunda
Nyajore	B5	IV-V,a	Unmapped, Approximate location shown
Nehanda	B7	I-II,k-m	N: Into Mozambique (?); W: Nyatwao & Karo; S: Hwatonda; W: Msengezi
Nebedza	B8	I-III,j-m	N: Kadzi & into Mozambique (?); W: Katsawa; S: Kafungura & Msengezi; E: Nyakutitatu, Karo & Nyatwao
Madzomba	B15	I,g-h	a knoll ringed by baobabs; N: Mututa & Chamukura; E & S: headwaters of Kusuka & Karemwe
Kadembo	C2	II-IV,h-j	N: Kada & Kadzi; E: Katsawa, Kafungura & Msengezi; S: Sapa; W: Kada & Pote
Nyahuma	C4	II-V,d-g	N: Bwase; W: Hunyani; S: Ambi; E: Mahuwi
Kachenje	D1	II-IIl;g-h	N: Headwaters of tributaries of Kadzi & a Musiga tree, Mujinga. W: Watershed between Kadzi & Bwase; S & E: Kada and headwaters of Bwase
Karambera	D2	IV-VI,d-f	N: Gwira & Kautyora; W: Karoi; S: Escarpment; E: Ambi
Chikuyo	D3	III-V,a-d	N & E: Dande; S: Escarpment; W: Hunyani
Svembere	E1	III,d	N & W: Ambi & Dande; S: Kautyora; E: Gwira
Marumba	E2	IV-VI,c-d	N & E: Karoi; W: Dande; S: Escarpment
Chivere	E4	IV-VI,h-j	N: Sapa & Nyanzou; W: Kadzi, Mahuwi, Karo & Katonha; S: Escarpment; E: Msengezi & Utete
Seremwe	F1	III-IV,b	Unmapped. Approximate location shown
Negomo	F2	II,m	N: Msengezi; W & S: Msingwa-Dande; E: Hoya & Stream (?)
Negomo	F2	II,c	Unmapped. Approximate location shown
Katuu	G1	III,b-c	Unmapped. Approximate location shown
Siti	H1	III-V,k	N & W: Msengezi; S: Moyo; E: Kadziruve
Mukombwe	H2	V,p	Unmapped. Approximate location shown. Extends into Mozambique
Chiurinyanga		II-III,l-m	N: Karo; W: Kariramombe & Mutaure; S: Forest edge; E: Msengezi & Msingwa Dande
Zezi	J1	III,k-l	N: Hwatonda; E: Nyakutitatu; S & W: Msengezi
Kasekete	J2	V-VII,g-h	N: Headwaters of Mahuwi & Kadzi tributaries; W: Mahuwi; S: Escarpment; EfKatonho
Chipute	K1	V-VI,f-g	N & E: Sakonde; S: Escarpment; W: Ambi
Chihwahwa	K2	V-VII,g-h	See Kasekete
Chitsungo	M1	V-VI,a	N & W: Hunyani; S: Escarpment; E: Nyakakutu
Chuzu	M2	VI,k	Ox-bow Meander, Msengezi
Kasindi	M4	III-VII,n-p	N & E: Hoya; S: Escarpment; W: Nzou-Mvunda
Chiruka	M5	II-III,n	N & E: Hoya; W: Msingwa Dande; S: Forest edge
Bamwomwe	T	I-II.i-j	N & W: Karemwe; E: Headwaters of Kusuka; S: Kadzi & Kada
Musuma	Ch.	VI,c	The Nyagwikwi Depression
Pfunze	Ch.	III,l	N: Msengezi; W: Kadziruve; S: Forest edge; W: Mutaure
Madzivamhako	S	III-V,k	See Siti
Mukarazi	?	I-IV,d-g	N: Mozambique!?); W: Hunyani; S: Bwase, E: Watershead of Bwase and Kada

Whilst anomalies arise, I stress that I recorded very few and that many appeared to have resulted from the intrusion of the colonial powers into the domains of the cults. Those which did occur were accommodated in such a way as to maintain the basic elements of cult organization and belief, the hierarchical relations among mediums and the close relation between the medium, his possessing spirit and a particular, bounded, ritual Territory. Thus, the Korekore regions and their constituent parts appear as ancient in origin and part of an enduring and unchanging reality.

Lengthy origin myths supported by complex genealogies validate the present-day distribution of regions and the relation between ritual territories within regions. The myths describe how Mutota’s sons and their descendants swept down upon the Tavara people over-running them and dividing the land among themselves.[18] In myth the Zambezi valley at this time is depicted as being under the secular control of Tavara chiefs and under the ritual authority of powerful chthonic beings, among them Dzivaguru and his sons Karuva and Musuma Nyamukokoko and the latter’s guardian, Chimombe.[19] The Tavara chiefs are said to have been defeated by superior physical force and the chthonic beings, who are usually held to have disappeared into the earth, by trickery and the use of superior mystical powers. In some versions of the myth, the chthonic beings, as ultimate ‘owners of the earth’, are associated with a territorial cult and are held to have provided rain for the Korekore invaders who supplicated them and offered tribute at a number of shrines. In other versions of the myth, particularly those related by the regional medium of Mutota and other mediums senior in his hierarchy, the over-riding ritual supremacy of the chthonic beings and their role in providing rain is minimized, and they are either reduced, as in the case of Musuma, to spirits of purely local influence, or designated as the objects of veneration of separate, circumscribed, regional cults such as those of Dzivaguru and his son Karuva, and of Chimombe.

Evidence from the ethnography of the Zambezi Valley east of Kariba, strongly suggests that the way in which the mediums of the Mutota cult relate the origin myth so as to circumscribe and minimize the influence of the chthonic cults is of relatively recent origin and can be related to the decline of the Dzivaguru cult and the concomitant rise of the Mututa cult which began in the twenties. I shall argue that this resulted from the need of the Mutota cult to appear manifestly autonomous particularly in its relations with Zezuru local cults.

The extent of the region of the Dzivaguru cult in the late nineteenth century is reasonably well documented and it covered a wide territory including at least part of what is now held to be the autonomous region of the Mutota cult. Tribute offered to the mediums of Dzivaguru and his son Karuva are reported to have come from many different Korekore groups both east and west of the cult’s centre at Choma (Map 2). Old men living in the east of Mutota’s region recalled how, in their youth, tribute was sent to Dzivaguru at Choma. Bourdillon (1972a, 1972b) has reported that among the easternmost Korekore in the Valley, a number of local cults of chiefly ancestors occur in conjunction with a territorial cult of Dzivaguru. In the east, beyond the Angwa river, there is a territorial cult centred on Chimombe, a chthonic being (Campbell 1957; Nicolle 1937), which also exists alongside local cults of Korekore ancestors. I am tempted to speculate that the decline of the Dzivaguru cult and the rise of the Mutota cult accelerated processes which had begun earlier with the collapse of the Mwene Mutapa dynasty in the eighteenth century (cf. Abraham 1959; 1964; 1966; Beach 1972) whereby what had been a Korekore cult of royal ancestors, possible centred on royal graves, operating alongside a separately organized chthonic earth cult was transformed into a regional territorial cult. Chiefly ancestors came to be accorded the powers of chthonic beings, and the mediums who represented them became associated with fixed territories. Such an interpretation would go some way to explaining the general form of the Mutota cult and also account for some of its unique features.

However, whatever be one’s reading of the meagre documentary and oral evidence concerning the remote past, I would strongly maintain that the variations presently observed in the regional cults of the Zambezi Valley are not to be interpreted simply or mainly as the ossified residues of a remote history as, for example, the historian Abraham (1966) has argued.

They result from on-going secular and ritual processes which continue to affect both inter- and intra-regional relations among the cults. Contrary to Korekore belief which, as I have stated, stresses the antiquity of the cults and the immutability of their ritual territories, the evidence indicates that ritual territories are adjusted both inter- and intra-regionally in response to secular and ritual processes and that concomitant changes occur in the validating myths. Intra-regional adjustment occurs in response to largely secular processes which affect the frequency with which spirit provinces are created and new mediums appear, whilst inter-regional adjustments occur in relation to secular and ritual processes, set in a much wider field, which affect the waxing and waning of cults in relation to one another. The two processes of adjustment, which I consider in more detail later, though ultimately connected are of different orders and have different durations.

Nevertheless, whilst I have emphasised that there is adjustment in ritual territories both inter- and intra-regionally, let me not convey the impression that this occurs rapidly. The process is relatively slow and occurs over several generations so that for any one generation of Korekore, the appearance of continuity in the organization of the cult and the lack of change in the number and size of ritual territories accords reasonably well with their belief. Korekore, however, are a highly mobile people. A period of uxorilocal residence and bride-service, coupled with the operation of exogamic rules, complicated nowadays by a high rate of male circulatory labour migration, scatters men and women over wide areas. These processes, together with competition among senior men within hamlets and neighbourhoods for followings, and the need to seek out, periodically, new areas of forest suitable for swidden agriculture, make hamlets unstable units which are periodically shifting and proliferating. It is not surprising, therefore, that the smallest unit of cult organization at the local level, the land-shrine community, is itself ephemeral, often persisting for only a generation or two.

The Land-Shrine Community

Within a spirit province, hamlets tend to cluster in relation to a source of water and a stretch of alluvial soil with a high water table suitable for almost continuous cultivation. Often, each cluster contains a small core of patrilineal relatives who may claim to have all once lived together in one hamlet. Each cluster usually constitutes a separate community with its own land shrine, and as the cluster waxes and wanes with the movement of people so does the land shrine community alter. In small spirit provinces there may be only one such community but frequently there are several. Each year when the rains are due (December) and, less commonly nowadays at harvest time (April/May), members of a land shrine community meet at a tree shrine, often a prominent Baobab tree, to make offerings to the territorial spirits. (While first fruits are still offered to the territorial spirit, this is now more commonly done on a separate and individual basis by heads of households in their own fields.) All households within the community are expected to contribute grain for the beer and flour to be offered at the shrine. This is prepared ritually by pre-pubertal girls and post-menopahsal women chosen from all the hamlets which constitute the community.

Those who offer together at a land shrine constitute a moral community (see Garbett 1969). They act together to deal with such polluting offences as murder, suicide, incest, witchcraft and sorcery, or to resolve disputes which may lead to such offences. It is believed that these offences pollute the earth, incur the wrath of the territorial spirit and render the land shrine ritual inefficacious. The territorial spirit shows his wrath principally by withholding rain, by sending birds, locusts and insects to attack crops, by inflicting cattle or humans with sterility and by killing with thunderbolts. Thus when people living within a spirit province speak with the spirit through his medium, they may seek advice on such matters as rainfall, offences against the earth and the fertility of humans, cattle and crops. But they may do this as individuals or as representatives of a particular land shrine community. Elsewhere (Garbett 1969) I have described how spirit mediums stress the dangers of mystical retribution flowing from an unresolved secular issue and thus come to play a prominent role in the resolution of local or communal disputes. However, whilst offences which are the particular concern of territorial spirits may involve a secular chief, he usually participates at a shrine as an ordinary elder. A chief does not organize or coordinate the ritual at the various land shrines within his chiefdom. In this respect, ritual and secular spheres are organizationally distinct.

Lineal and Territorial Aspects of a Regional Cult

Lest this point be misunderstood, I must summarize an argument which I have put more fully elsewhere (Garbett 1966). Korekore who claim patrilineal descent from Mutota form chiefly descent groups in a number of chiefdoms in Mutota’s region. Invariably, one finds that the founder of a chiefdom, an ancestor two or three generations removed from the oldest living, is recognized as a mhondoro (‘lion’) spirit and is represented by a medium. An especially important duty of a medium of this type is, after a chief’s death, to select from a number of candidates the next chief and to install him in office. Generally, the chief’s patrikin first raise such a matter with the medium who represents the most junior lineal ancestor speaking as a mhondoro. If he is unable to resolve it satisfactorily, it will be taken to the next medium senior in the hierarchy and so forth, until the senior and regional medium is reached. I have described in some detail (Garbett 1966) the kind of politicking which occurs nowadays over chiefly succession and which comes to involve not only various factions and several mediums of the hierarchy but also officers of the district and provincial administrations. I stressed in that account the delaying tactics which mediums employ as they attempt to assess the balance of support in favour of one candidate or another and wait for some consensus to emerge.

From this summary, and my discussion above, it can be seen that there are two aspects of the Korekore cult which relate to the crucial position of the regional medium. The first or lineal aspect derives from the fact that patrilineal descendants may come from wherever they live to consult a medium representing the ancestor.. Consultations usually relate to illness and misfortune due to the affliction of ancestral spirits and to matters of inheritance and succession. The second or territorial aspect derives from the fact that people living within a spirit province may consult its medium and seek to placate its territorial spirit’s wrath. The regional medium’s position is crucial, because it is held at a level that is maximal for the intersection of the two aspects. He alone can be consulted by any person living anywhere in the region’s provinces and also by a patrilineal descendant of the region’s apical ancestor coming from wherever he may live.

There is a further aspect which relates to the regional medium’s position but which entails relations that are personal primarily, rather than hierarchical and lineal or territorial. A medium who establishes a reputation for wisdom in settling disputes, for exorcising malevolent spirits, for effective cures and for bringing rain may draw supplicants to him from a wider vicinity that extends far beyond the boundaries of his spirit province or even his region. For example, the vicinity of Mutota’s regional medium extended south as far as Hartley District (130 miles), south east as far as Bikita (300 miles) and as far west as Urungwe (95 miles) (See Map 2). The supplicants from the wider vicinity, unless they are lineal descendants of the territorial spirit, use the medium very much as they would use an ordinary diviner (ng’anga). Some mediums, by expanding their vicinity of supplicants, come to have reputations which far outweigh their, possibly, humble positions in the formal hierarchy of mediums. Ultimately, this may lead to friction between a junior medium and one more senior about their respective jurisdiction and spheres of ritual authority, and the intervention of the senior medium may be required.

Change in Ritual Territories

I stressed earlier that Korekore maintain that the present distribution of ritual territories between and within regions is ancient and immutable. I now consider the processes by which intra-regional adjustments in ritual territories occur before turning to consider the wider field of inter-regional relations.

Evidence indicates and logic suggests that spirit provinces must be adjusted in response to secular processes which affect the frequency with which new mediums appear and new spirit provinces are created. These processes have their origin in the competition among royals for chiefly titles which, in the past, often led to fission and the establishment of new chiefdoms and chiefships. I have indicated that invariably within two or three generations of the foundation of a new chiefdom a medium appeared to represent its founding ancestor and to regulate the succession to the new chiefship. Each new medium was installed in a ‘new’ spirit province which was often within or near the territory of the chiefdom, centred on what was said to be the first chief’s grave, but was always defined within some previously existing spirit province. Mediums who had responsibility for regulating succession to the parent chiefdoms relinquished their direct responsibility for this to new mediums who became their subordinates in the hierarchy. However, they were available for appeals against the decisions of their subordinates and appear to have continued to have had an important role in mediating relations between the new chiefdoms. Hence, the evidence indicates that the secular processes which gave rise to the establishment of new chiefdoms also led to the proliferation of spirit provinces and the appearance of new mediums. However, there were contrary processes operating. Chiefdoms were established through conquest and thus formed at the expense of existing ones. As the mediums linked with former chiefdoms became less directly involved in secular issues and matters of succession, there was a tendency for their positions not to be succeeded to after their deaths, though their spirit provinces remained and the names of their possessing spirits were retained as ancestors in the genealogy. It is noticeable in Mutota’s genealogy (Genealogy 2) how none of his descendants from Generations D to J is represented by a medium. Thus, the factors which led to the demise of some chiefdoms also led to the eventual loss of significance of the mediumships associated with them.

Whilst it is clear that certain mediumships junior in the hierarchy have been terminated, certain senior mediumships, by contrast, appear to have been filled by a succession of mediums over long periods of time. At this level, where mediums are remote from direct involvement in day-to-day secular issues and in chiefly succession, different processes operate. Spirit provinces associated with ancestors at higher genealogical levels tend to be large and‘new’ spirit provinces are established as sub-divisions of them. In the west of the Dande, for example, Nyahuma’s spirit province is held to have once extended from the Bwase stream (Map 1: II,e) in the north to the escarpment in the south but became sub-divided among his sons, grandsons and other descendants. Similarly, Kagoro’s spirit province to the east (Map 1: IV-VI;K—N) is hqld to have once extended from the Msengezi in the west to the Hoya river and then later became sub-divided. Where such sub-divisions are said to have occurred, the medium representing the ancestor of the original, now divided territory, is held to have some degree of control over the mediums and spirit provinces within it, irrespective of those mediums’ formal ranking in the genealogically derived hierarchy. Nyahuma’s medium in the past, for example, is said to have received tribute from Chitsungo’s medium despite the fact that in the formal hierarchy the medium of Chihwahwa was his immediate superior. There is therefore, some tendency for the hierarchical affiliations of junior mediums to be cross-cut by their territorial affiliation.

One can therefore reconstrue Map I as a set of large ritual territories associated with remote ancestors such as Kagoro, Nyahuma, Nebedza and Nehanda who, significantly, have all been represented by mediums over long spans of time. These large ritual territories, therefore, appear to provide an enduring territorial and organizational basis for the cult and persist despite sub-division within them. The mediums allocated to them are above the level of adjustment and change which affects the positions of junior mediums and are thus crucial for the on-going organisation of the cult. Furthermore, such mediums have important additional duties within the cult and in relation to other cults. As I indicated earlier, perpetual kinship links the Mutota region to a number of others (see Genealogy 1). Mediums of one cult enter into relationship with mediums of other cults in terms of these perpetual relations. Nebedza’s medium in addition to acting as head of the Mutota region during inter-regna also maintains relations with the Chikunda regional cult through his “son-in-law” the medium of Kanyemba, brother of the founder of the Chapoto dynasty. In the past Kagoro’s medium maintained formal relations with the Dzivaguru cult region as a “son-in-law” of Dzivaguru’s medium and all transactions with the cult were conducted through him. Similarly Nyahuina’s medium maintained relations with the chthonic cult region of Musuma.

Mediums occupying these key positions in a cult are ranked immediately below the regional medium and, with the exception of Nyahuma, are presented as his ‘children’. It is critical for the cult’s organisation that the regional medium maintains regular contact with them and exercises control over them. He does this by the regular transmission of symbolic messages, by receiving tribute from them and by issuing directives through them to be transmitted to local populations either through sub-ordinate mediums where these exist or, in their absence, directly to the elders of the land sh*rine communities. Such directives may include matters ranging from those of purely ritual concern such as the conduct of land shrine rituals or the behaviour of subordinate mediums, to those which, whilst of ritual significance, may impinge directly on the economic life of communities, such as the enforcement of rest days (chisi), restrictions on the use of fish poisons and prohibitions affecting diet and modes of cultivation. In addition, should one of these senior mediums commit some misdemeanour, he may be summoned to appear before the regional medium to account for his behaviour and he may even be fined.

Expansion and the Regional Medium

Since the twenties, there is no doubt that the influence of the cult of Mutota has increased considerably. Since that time, the subordination of the cult to that of Dzivaguru has ended, at least two ritual territories have been incorporated directly into Mutota’s region and its overall influence has been expanded and extended far beyond the domain of Korekore cults into the Zezuru cult of local spirits.

The region of Mutota, and indeed the cult to which it belongs, has been much influenced in its current form and extent by the most recent medium of Mutota, the late George Kupara. The three factors which have been most important for his influence are first, his personal qualities; second, the condition of linked cults and regions at his installation and during his period of office; and third, his connections with and recruitment from an inner circle of kin who are an elite in both local and territorial cults. I will consider each factor in turn. In 1963 when I lived with him, Kupara was about 65 and had been in office since his possession as a young man in 1919. He died in 1973. Physically, he had a striking appearance: he was tall, lean, with bright piercing eyes and an expressive face. Like other mediums, he dressed simply, but distinctively, in a black cloth worn, toga fashion, around his waist and drawn over his left shoulder. Around his neck he wore a thick, elaborate, plait of fine black beads and on his wrists, black bangles of horn. He had a sharp mind, an extensive knowledge of herbal remedies and a phenomenal ability to retain genealogical detail and the minutiae of diverse myths of origin. His obvious charismatic qualities were revealed particularly through the drama of the seance.

Seances were held in a spirit house (zumba; great house) situated in the bush about a hundred yards from the regional medium’s hamlet. Its only openings were two doors, one facing south which the medium used, the other facing north which the acolyte and supplicants used. Seances were conducted in total darkness beginning around midnight and continuing until dawn. After some preliminaries, supplicants were guided by the acolyte into the darkness of the spirit house and seated along its west wall facing the medium who lay stretched out on his stomach covered completely by a black cloth and a fur blanket. With no visual cues whatsoever, the drama of the seance depended entirely upon the medium’s vocal control and oratorical skills. Kupara was a supreme master of mamhuko, a rythmical, free-verse, speech form used by mediums. He would be fierce and terrifying with the sceptical or recalcitrant; gentle, playful, grandfatherly, with the awed and the afraid.

About half the matters raised with the medium at seances concerned personal and familial misfortune. ^XA supplicant might enquire about his illness, or why he could not keep a job, or a woman might ask why she had not borne children. Generally, the regional medium would deal with such matters routinely without a great deal of discussion, quickly identifying a witch, an ancestor or some other spirit as the cause of the misfortune. He would dispense herbs to deal with symptoms, charms to protect against witchcraft and prescribe standard rituals to placate ancestors or exorcise afflicting spirits. Frequently, after hearing a supplicant’s name, origin and praise names, Kupara was able to give a genealogy of the supplicant’s ancestors, a feat which, if the supplicant was from a distant area, appeared particularly impressive.

Supplicants who came to consult the senior medium paid a small fee (chitekwe) to call the spirit and were expected to make further gifts of black cloth and money during the course of the seance according to their means and the nature of their problem. At the end of the seance, Kupara would apportion a small amount of money to himself, to his acolyte, and to any other officer of the cult in attendance. Any remaining money was regarded as belonging to the spirit and was given into the care of the acolyte. This money was used to buy grain to feed the many supplicants and their companions who came to Kupara’s hamlet through the course of the year. Significantly, it is also used to make relatively substantial loans and gifts which seldom appear to be repaid even though a nominal interest may be charged. I recorded loans varying from £5 to £20. These appear to have been given to enhance the reputation of ‘Mutota’ for generosity and for not amassing wealth at the expense of ordinary people. Loans, however, do not create any kind of debt dependency and most appear to be quickly forgotten. What the senior medium retained as personal income was just enough to enable him not to work extensively in his gardens and to keep his family modestly clothed.[20] Over the year, taking into account receipts and payments, he more or less broke even. In this region, therefore, the principal officers do not acquire a great deal of personal wealth through the operation of the cult.

Whilst in seance the senior medium’s performances were consistently impressive, he reserved his most spectacular displays for visiting mediums and candidates for mediumship. With visiting mediums, clad in a black and white costume, with fur headress and false beard, he would participate in a competitive dancing display in which he would scoff at their dancing abilities, to the amusement of onlookers. With candidates for mediumship, he was much more serious. Such candidates were expected to become possessed during the seance, to identify their possessing spirit and to give genealogical and biographical information about it to validate their mediumship. Kupara reacted particularly sharply if he considered a candidate to be making a false claim to mediumship. This happened when two candidates, referred to him by the medium of the Zezuru Nehanda, claimed to represent Chaminhuka. They failed to become properly possessed and he burst out angrily, “Do you want to be delighted by being told that you have a mhondoro? I am not like an ngozi (adept) who rubs and shakes people’s heads to bring out the ngozi.[21] I really want a candidate to speak out everything I ask him. You come from very far away saying, ‘We have spirits’, but you do not speak. What am I to make of that?” He then directed criticism at Nehanda’s medium, through her acolyte who was present, saying that he wanted to know why she had sent him ‘eggs’ (ie, ‘speechless mediums’).[22] Was she trying to test Mutota’s ability to discover whht was in the ‘eggs’? Conversely, when candidates for mediumships gave particularly striking performances (such as were given by the child medium Gutsa and the female medium of Chikare, cases which will be considered again later) the medium would continue for weeks afterwards to uphold them to other candidates as examples of how a person truly possessed by a mhondoro should behave.

I have dwelt on Kupara’s behaviour in the context of the seance because it must be stressed that it is only in this context that a medium can establish his reputation, create relationships with supplicants and other mediums and so expand his vicinity and increase his influence. A medium who becomes old and fails to sustain his performance in seances rapidly loses his vicinity whilst a young medium, who succeeds to an established prestigious mediumship, may find that he has not the ability to sustain the vicinity of his predecessor (cf. Bourdillon 1972). The ritual authority associated with a medium’s office and the prestige which may come from an expanded vicinity are conceptualized as not attaching to the medium but to his possessing spirit. Outside of the context of the seance a medium is, in a real sense, powerless. Kupara, himself, like many mediums I observed, was a lonely man. He had few close friends or local kin. Of course, the disruptive move from Gota had detached him from local support within his spirit province but, even there, whilst he appears to have had a strong connection with a local headman, Mushoshoma, who acted on occasions as a ritual assistant, Kupara did not have connections which could be construed as providing him with any kind of local secular power base. Indeed, a medium must appear to be detached and avoid an obviously close association with one faction or another otherwise he runs the risk of being accused of faking possession and giving partial decisions. Kupara’s loneliness also stemmed from his ambiguous identity: was he to be treated as an ordinary mortal or as the ancestor, Mutota, personified? This uncertainty, which tended to pervade all his interactions with neighbours and locals, created a continuous tension. In addition, neighbours and locals would comment on the quality of his performances at seances and, in gossip — even occasionally to his face — would allege that he faked possession on certain occasions. It was difficult for Kupara to distance himself from those with whom he was in daily contact, and the tension in his relationship with locals was manifested in a formal joking relationship: he was addressed as aSekuru (Honoured Grandfather) which is a joking category. Visiting supplicants, by contrast, would address him very respectfully as aMvura (Lord Rain).

This tension also extended to Kupara’s relations with his staff, the acolyte (mutape), the ritual assistant (dunzvi) and, to a lesser degree, with ‘the wife of the spirit’ (mukaranga) who, as a young girl has been offered into the service of the spirit, but who had later become his wife. In 1963, Kupara’s acolyte was a man of about fifty. He was third in the present line of acolytes, established during the mediumship of Gavanga, Kupara’s predecessor (see Genealogy 3). He had succeeded his father some twenty years previously. He was head of the hamlet in which the senior medium lived and was a popular man with a reputation for uprightness and fair dealing. He was responsible for the day-to-day management of the seances and local affairs of the cult. Significantly, he addressed Kupara, even when they were alone, as ’aMvura and never as ’aSekuru. The two did not spend a great deal of time together, and all formal business pertaining to the cult was only discussed at seances. Like other mediums, Kupara, unpossessed, would not discuss matters of the spirit, claiming to have no knowledge of what had occurred during seances. In the context of the seance, whilst acting in a supportive role to Kupara, the acolyte was very much subservient to him and addressed the ‘spirit’ humbly and respectfully. This was true also of the ritual assistant, a very old man, who performed ad hoc rituals for the spirit of Mutota, and of the ‘wife of the spirit’ who attended the medium when he first becomes possessed and assisted in the treatment of women with reproductive disorders. Kupara’s staff, therefore, were very much under his control. They did not exercise any influence over his behaviour during seances nor did they meet or collude with him before a seance. It would appear, therefore, that what the mediumship of Mutota became during Kupara’s lifetime stemmed very much from the personal initiatives which he took in response to issues arising in the context of the seances. Of course, in many matters, particularly those with secular consequences, his actions were constrained by a watchful public and by the opinions of other mediums. In the cult’s external relations and in the expansion of his own vicinity, he had more freedom of action since the consequences of his actions had no immediate local effect. But even here, whilst there were certain policies which it seems clear he wished to pursue — such as the attachment of a medium of Chaminhuka to the cult — the opinion of other mediums and the public acted as a constraint. However, I do not consider that one can determine any overall long-term strategy underlying the whole of Kupara’s actions. As revealed through the forty seances which I observed, Kupara’s behaviour can best be understood in terms of four separate but related objectives: to uphold mediumship as a general value in a situation in which it was being assailed by the colonial administration, by missionary activity and by separatist sects (cf. Ranger 1970: 220); to maintain and increase the prestige of Mutota as a spirit in comparison with the spirits of other chthonic and Korekore cults; to expand his vicinity and hence increase his own prestige as a medium; to protect himself from the scepticism of neighbours and locals. Thus, overall, his actions can be interpreted in terms of a series of tactics in which one or more of these objectives was paramount. I shall return to a discussion of these crucial issues later.

The Decline of Chthonic Cults

Kupara came to office just before the chthonic Tavara cults of Dzivaguru and Karuva went into sharp decline. In 1922, after a ritual murder, the administration arrested and imprisoned all the officers of the Dzivaguru and Karuva cult. With the exception of the acolyte of Karuva and a junior medium of the Dzivaguru cult operating from Mozambique, they were never replaced (Bourdillon 1972a; 1972b). Whilst the sacred grove of Karuva at Nhenhene, near Choma (Map 2) is still tended and offerings are made there, this chthonic cult has never again attained the prominence which it clearly had at an earlier period when Korekore, through their mediums, would supplicate for rain.

At the end of nineteenth century, there appear to have been four cults of chthonic beings, mostly associated with the Tavara, of ritual significance to Korekore living in the Zambezi Valley and immediately adjacent areas of the plateau: those of Dzivaguru and his son Karuva, of Musuma and of Chimombe. Korekore in the region of Mutota tend to speak of them nowadays as if they were all part of one cult and, in myth, make Chimombe the guardian of Musuma who, in turn, is made a brother of Karuva (see Genealogy 1). They also replace Dzivaguru with Karuva and do not now distinguish their separate cults. Further east, near Choma, these connections are not made and all the chthonic cults are held to be quite distinct. It is difficult now to reconstruct how the cults operated in the past: there has been no medium of Dzivaguru for over fifty years whilst Musuma’s shrine has decayed and its ritual officers have died or dispersed. Chimombe’s cult virtually collapsed in the late fifties when its principal officers were among people moved by the colonial administration and resettled on the plateau.

From what can be gathered from early reports (see Bourdillon 1972b), legends, and accounts of elders, Korekore acknowledged the chthonic beings as “owners of the earth” and would supplicate them, through their mediums, for rain. Each cult, with the possible exception of that of Dzivaguru, had a centre, with a shrine, a medium and a staff. Wheiher each cult constituted a region with a bounded territory cannot now be established. Broadly, the region of the Chimombe cult appears to have extended west from the Angwa river, that of Musuma, east, from the Angwa to the Kadzi river, and that of Dzivaguru from the Kadzi eastwards, with intrusions to the north into Mozambique and south, well-beyond the Zambezi escarpment. Karuva’s region appears to have extended to include the chiefdom of Choma, at which his famous shrine is located, and the neighbouring Tavara chiefdoms of Chiswiti and Gosa (now in Mozambique) as well as the immediately adjacent region, to the south, of the Korekore ancestor, Nyabapa. In myth, Nyabapa is said to have overcome Karuva by a trick. He then disappeared into a pool in the sacred grove at Nhenhene, from where he is held to control the rainfall and the fertility of the earth over the whole of his region. Periodically, when the virgin (Nechiswa) who tended Karuva’s shrine at Nhenhene was held to have been defiled, her supposed defiler was burnt alive as a sacrifice to Karuva. It was such a sacrifice which caused the intervention of the colonial authorities in 1922. The Dzivaguru cult, to which the Karuva cult and all Korekore cults within its domain were, in some sense, subordinate, was clearly the most extensive and important of all the chthonic cults. However, at least in the late nineteenth century, it appears to have operated without a central shrine and headquarters. In the remote past, the medium of Dzivaguru was the “leading religious figure” at the court of the Mwene Mutapa(cf. Ranger 1963; 1970).

Outside of the region of the Mutota cult, the cults of Dzivaguru and Karuva have a continuing, if diminished significance. Both Bourdillon (1972b) and Gelfand (1962) have reported that despite the absence of mediums and a full staff of ritual officers, the chief of Choma, acting as acolyte, receives regular delegations, once or twice a year, from neighbouring and some more distant, easterly areas, who bring gifts of cloth to be offered to Karuva in the sacred grove. Whilst the officers of the Korekore cult of Nyabapa are ambivalent about the superordinate or subordinate status of the cults of Karuva and Dzivaguru (Bourdillon 1972b), the medium of Nyabapa still sends tribute to the sacred grove. This is quite unlike the position within the Mutota cult where all mediums and especially the regional medium, completely repudiate any notions of the cult’s subordination to that of Dzivaguru and, pointedly, do not send tribute. They treat the Dzivaguru cult as an equivalent, autonomous, regional cult, denying the efficacy of Dzivaguru’s power over rainfall and fertility within the boundaries of Mutota’s region. This appears to have been the position since at least the thirties when, as far as I could establish, the last tribute was sent from Mutota’s region, via Kagoro’s medium, to Choma.

The evidence concerning the relations between the Dzivaguru cult and Korekore regional cults would appear to indicate that, with the exception of the Mutota cult, the ritual significance of the Dzivaguru cult has continued despite its very weak organizational significance. By contrast, the mediums of the Mutota cult, deny the ritual and organizational significance of any of the chthonic cults. I suggest that this may be related to the rise in influence of the Mutota cult in a wider field of ritual relations in which it became important that its autonomy and supremacy were fully manifested.

Kupara’s Recruitment

As for the third factor in Kupara’s influence on the Mutota’s cult, his connection to, and recruitment from an elite: his family and lineage had a long association with important mediumships in both the local and territorial cults. By descent he was not Korekore. He claimed to be a Zezuru descended from Shayachimwe, the founder of Hwata, a chiefdom supposedly once powerful and still of some influence. (See Genealogy 3, for this and the following relationships). His predecessor as medium, who died before Kupara was born, was his own grandfather, Gavanga. Another close kinsman was also crucial. The man who tended him during the years of his possession and who helped install him in office was his father’s brother, Nhamuyebonde. Nhamuyebonde was himself a medium of an important Korekore ancestor (Chiurinyanga; Genealogy 2: I, 1) and because this close kinsman was involved with the succession to a number of Korekore chiefdoms at the time of Kupara’s installation and later, it appears likely that Kupara was able to gain information and learn how to deal with the tangles of mediumship. Kin in the local cult of Zezuru were in some respects even more important for Kupara’s part in regional expansion. The first medium of the Zezuru Nehanda to become attached to the Hwata dynasty was Chapo. She is said to have become the wife of Shayachimwe, the first Hwata. After her death, Charwe, a classificatory sister of Kupara’s grandfather, succeeded to the Nehandaship. She was the famous medium who was hanged for her prominent role in the uprising of 1896 and its aftermath (Ranger, 1970, 10). Some time after her death, probably in 1906 (see Ranger, 1970: 10), a classificatory sister of Kupara, Matiirira became possessed by Nehanda. She died sometime in the fifties. Thus, at various times the descendants of the once powerful chief Hwata held important mediumships in both cults and for a considerable part of his life Kupara had a close relative, Matiirira, in control of the Zezuru mediumship of Nehanda at Mrewa. This provided Kupara with an established and fruitful link for referrals from the local cult: the medium of the Zezuru Nehanda at Mrewa continued to be responsible for referring local mediums of the Zezuru cult to Mutota’s medium in order for them to be tested and for the most prominent of them to become formally attached to the territorial cult and Mutota’s region. Such formal attachment is achieved through what amounts to a cross-cult translation of belief: i.e., the mediums’ local spirits are transposed into Korekore ancestral spirits which are then accepted as children of Mutota.

To clarify this further in order to demonstrate how relationships are extended from one cult into the other, it is essential to consider the conceptual framework which Korekore themselves use to comprehend this situation. I refer to the genealogy of remote ancestors (Genealogy 2). Mutota’s genealogy has a distinctive pattern which calls for explanation. It is shaped like a “T” rather than, say, the inverted “V” so common in lineages. The first generation of descendants is a great band, somehow largely without issue. Generally, Mutota is credited with some 16 to 20 children, almost all of whom are not said to have had descendants.[23] Indeed, usually only one is credited with his own line, but even this extends in the manner of a king list, for ten generations, before there is any significant segmentation. The identity of many in the list of Mutota’s children is not surprising and only calls for brief comment. They are members of the Korekore pantheon of remote ancestors; their spirit provinces are in and around the Zambezi valley. As I indicated earlier, many of the ancestors at this genealogical level are credited with having had large spirit provinces which later were sub-divided. Some of the mediums representing them, I have argued, are significant in maintaining the enduring territorial and organizational basis of the cult and its external relations with other cults. Placed in the hierarchy at this level, these senior mediums are above the secular processes which affect the emergence of new mediums, the delineation of new spirit provinces and the segmentary processes which affect its lower levels. What does call for fuller comment, however, is that these Korekore ancestors are named in relationship with others quite unlike them. Two of these have spirit provinces but appear to have been added to the genealogy in recent times whilst others have no spirit provinces, and I infer from external evidence that they are part of the Zezuru cult and are local spirits.

The two ancestors with spirit provinces who appear to have been added to the genealogy are Madzomba (Genealogy 2: Bl5) and Chiobvu (B,l). I consider only Madzomba’s case since I have strong evidence on it and also because I consider it provides insight into the general cognitive model which is used to attach Zezuru local spirits to the Mutota cult. My evidence on Chiobvu’s attachment is weaker but is similar in its essentials.

Madzomba’s spirit province (Map 1: I,h) lies some 20 miles due north of the hamlet in which the medium of Mutota was living in 1963. I was told by the acolyte of Mutota’s medium and others that they had all been aware of the medium of Madzomba for many years. Indeed, five previous mediums are recalled and the mediumship appears to have been regularly occupied and long-established. In the past, the mediumship was firmly associated with the Tavara ancestral cult of Bamwomwe and Madzomba was held to have been his ‘sister’s son’ (muzukuru). In 1957 or 1958, shortly after the death of Bamwomwe’s medium, the medium of Madzomba suddenly appeared and asked to consult Mutota. At a seance he is said to have elaborated a myth in which Madzomba’s mother, supposedly a wife of Mutota, had been driven away by Mutota’s son, Nebedza, after having failed a withcraft ordeal (muteyo). Eventually, she was taken as a slave by Chief Dotito in the northeast. Later she and Madzomba escaped and sought sanctuary with Bamwomwe. Madzomba’s medium concluded by offering himself as a “long-lost son” of Mutota. This account was accepted by the regional medium and by other senior mediums. Madzomba’s name is now regularly included by all of them in their recitations of Mutota’s genealogy and Madzomba’s medium is firmly established within the region of the Mutota cult. As we shall observe, this device of establishing in myth that a spirit is a long-lost child of Mutota appears to be one which is cognitively acceptable and one which is used to attach Zezuru local spirits to the cult of Mutota. It is delightfully simple and yet it is reversible. A slight alteration to the story, making the woman pregnant after she leaves or runs away, makes the child illegitimate without any necessity to deny the whole myth. I have not heard this reversal used in the context of denying the claims of mediums, but it is frequently used to exclude descent groups from participating in the succession to a chiefship.

Examples of those cited as children of Mutota who are clearly identifiable as local spirits of Zezuru cults are Goronga, a well known local spirit (see Ranger 1967: 211) associated with the Lomagundi District, Nyamuswa (Gelfand 1959: 18); Chifwanya, who features early in the Korekore origin myth as leaving Mutota at the escarpment and returning to Mrewa; Chikare, whose case I consider later, and probably, Beperere. Somewhat in a conceptual category by herself is Nehanda, as befits the spirit of a medium with perhaps the most important role in cross-cult relations. She is what might be called a “transcult spirit” in that she is now both a local and a territorial spirit, rather than merely a local spirit assimilated through kinship into the genealogy of remote ancestors. Thus she has mediums both as the renowned local spirit among the Zezuru and as a daughter of Mutota and ‘mother’ of all Korekore. Korekore accept the existence of two mediums of Nehanda: the one operating in the Mrewa area, the other operating in the Zambezi Valley and temporarily living in the spirit province of her ‘brother’, as I have noted. Nevertheless, Korekore argue that there is only one Nehanda and that her spirit possesses mediums, at different times, in two places. In fact, in the sixties, three mediums of the Zezuru Nehanda were operating, one at Mrewa, one in the Mhondoro Tribal Trust Land and the other at Rwizi. As we shall observe, considerable rivalry existed between the mediums from Mrewa and Mhondoro. Already, in the late nineteenth century both a Korekore and a Zezuru medium were operating. This indicates the continuing importance not merely of the transcult spirit and her medium but even more of the extension of relationships across the cults.

The Linking of Local and Territorial Cults

Now we can appreciate more of the significance of the pattern of Mutota’s genealogy with its inclusion of unlike spirits together in relationships with the apical ancestor. I shall comment on the features that are most illuminating for my discussion of the linking of cults. These features allow for expansion along with the inclusion of non-hierarchical and hierarchical relations, in accord with the situation which the genealogy comprehends. Having numerous ‘children’ without issue (ie, the top of the ‘T’ and upper part of the genealogy) allows the assimilation of local spirits, and thus the incorporation of their riiediums to occur in an ambiguous way that fits their lack of hierarchy. So long as the birth order is not specified and the total number of children unknown or not fixed, more and more relationships can be extended to local mediums without them having to constitute a hierarchy among themselves. Another feature is the long line prior to segmentation in the genealogy. The long line provides a fixed set of reference points, which a candidate medium can cite for legitimacy. Moreover, the rest of the genealogy with its segmentation defines the hierarchical relations.

As I shall document through case material later, the senior medium plays a key role in the revision of the genealogy which is necessary for the formal attachment of additional mediums. I would stress here that a senior medium such as Mutota’s cannot simply alter the genealogy at his whim. There are a number of prominent mediums well versed in the ‘orthodox’ Korekore genealogy. Nevertheless, there is uncertainty among all of them as to exactly how many children Mutota was supposed to have had and what their birth order was and so forth. Adding a child or two, particularly when they have no descent line associated with them is not difficult but, and it is an important qualification, such amendments to be meaningful must receive wider acceptance and more validation. More complex revisions of the genealogy, which imply further relationships and perhaps special standing for some mediums, may be more difficult and less readily validated. To illustrate briefly from the case I discuss more fully later, Mutota’s senior medium acknowledged the claim of a visitor, a medium of the local spirit Chikare, that her spirit was the daughter of Mutota. The new relationship was established and the genealogy amended at a seance in my presence. Within a few months other prominent mediums in the Zambezi valley began to recount the tale of how Motota had found his long lost daughter, Chikare. The truth of her medium’s claim was accepted. However, the senior medium tried to elaborate the relationship and make a more complex amendment of the genealogy, unsuccessfully. He was fond of adding the name of that other prominent local spirit Chaminhuka to the genealogy as a son of Mutota. He went further and, at a public seance after the visit of the medium of Chikare, made her out to be a uterine sister of Chaminhuka; both were said to be children of one mother, Marumbi, a further genealogical amendment. This was not accepted by other mediums. Their unwillingness to accept Chaminhuka as a son of Mutota may well have stemmed from the fact that the Zezuru myths concerning Chaminhuka’s miraculous powers, his adventures and encounters with the Ndebele, were widely known not only by them but by ordinary Korekore. Indeed, they are often told, along with others of Korekore heroes, as “bed-time” stories to children. Korekore conceptualize Chaminhuka as a creator being, equivalent in power to Mutota. It would have taken a particularly convincing medium of Chaminhuka to have altered this well-established view. Whilst two candidates for the mediumship were referred to Mutota’s medium by the Zezuru Nehanda in 1964, both gave abysmal performances in seance. One, as I noted earlier, failed to speak; the other, although he attempted to speak, was confused and ill at ease. He was advised by Mutota’s medium that an alien spirit was intervening to prevent his full possession by Chaminhuka’s spirit and he was told to return to Nehanda’s medium for treatment and instruction. The senior medium once again expressed his anger at Nehanda’s medium for sending him someone whom she had not tested adequately herself. Significantly, however, the candidate, who came from a village, Mushake, in which I lived in 1958, told me that he was convinced that he was possessed by Chaminhuka but that the alien spirit and Mutota’s anger at Nehanda had prevented him from becoming properly possessed. He then said, “Truly, Chaminhuka is the son of Mutota, as Mutota says. I believe that he is because when I was sent here, Nehanda told me that I was to go to my father to be tested”. This statement strongly suggests that not only had Nehanda’s medium a similar interest to Mutota’s medium in wanting to incorporate Chaminhuka directly into the Mutota cult but also that she was suggesting the idiom in which candidates for certain important mediumships should present themselves when they were tested. This further suggests, that Nehanda’s medium had some special knowledge of Korekore mythology and of the Mutota cult and the case of the Chikare medium adds weight to this. It is possible that she acquired some of this knowledge from her husband who is Korekore, a descendant of Chuzu (Genealogy 2: Ml), and who also acts as her acolyte. I should stress that she and Mutota’s medium had only met once, in 1962, when she presented herself for testing by him and was accepted as a genuine successor to the former medium, his classificatory sister, Matiirira.

Of the fourteen mediums who came to be tested by Mutota’s medium during the seven months from October 1963 when I lived with him, all save one came from the Zezuru cult of local spirits.[24] Ten of them were referred to Mutota’s medium by the medium of Nehanda. Most were sent simply to establish that they were genuinely possessed: neither they nor Mutota’s medium made any attempt to translate their spirits into Korekore ancestors. Only in the case of prestigious mediumships such as those of Chikare and Chaminhuka was this attempted. But it is more than the prestige of the mediumship which is significant: all the Zezuru spirits accepted as children of Mutota are those whose influence is held to span a number of local groups and none are ancestral spirits with specific genealogical connection with the living. In the Zezuru cult of local spirits, mediurris possessed by spirits of this order, are those who validate the possession of lesser mediums and thus have a significant coordinating and organizational role. Thus, they are in some respects the equivalents of such senior mediums in the Korekore territorial cult as those who represent Nebedza, Kagoro, Nehanda and Nyahuma. Hence, the appropriateness of inserting the spirits they represent into the higher levels of the Mutota genealogy.

The connection between the territorial cult of Mutota and the Zezuru cult of local spirits has existed at least since the mediumship of Mutota was held by Gavanga, Kupara’s grandfather, at the end of the nineteenth century. This connection, was expanded and developed under the mediumship of Kupara, first through his classificatory sister, Matiirira, and then through her succesor to the Nehandaship. Nevertheless, there seems to have been a break in referrals to the Mutota cult of perhaps ten years after Matiirjra’s death inthe fifties before her successor made a successful bid to re-establish the link in 1962. Why this should have occurred I cannot fully determine. I never met any of the Zezuru mediums of Nehanda: they all lived and operated over a hundred miles from my field area. However, from my own knowledge of the Mrewa area in 1958, from Fry’s (1976) study in an adjacent area and from Chavanduka’s (1970) study of a Zezuru succession dispute, certain matter are clear. In the late fifties, the Mrewa and Mhondoro mediums of Nehanda came into competition over the succession of the Zezuru chiefship of Hwata. Briefly, two factions became involved in a lengthy succession dispute after the death of Chief Hwata XI in 1952, one supporting his son, the Acting Chief Henry, the other supporting another candidate, Samuel. The medium of the ancestor Gutsa, responsible for nominating a new chief was dead. Henry and his faction eventually brought forward a man whom they claimed was possessed by Gutsa. After divination had confirmed this, the medium nominated Henry as chief and the administration duly installed him as Chief Hwata XII. Meanwhile, Samuel and his faction went to inform the Mrewa medium of Nehanda that they were convinced that the Gutsa medium supporting Henry was false and that they had a youth who seemed to be possessed by Gutsa’s spirit. The Mrewa medium ordered all the parties to the dispute to assemble at the burial place of Hwata chiefs, at Mazoe, where she would establish which of the two mediums of Gutsa was genuine. Henry and his faction, alarmed, went off to consult the Mhondoro medium of Nehanda. She assured them that their medium of Gutsa was genuinely possessed and that Henry was rightfully chief. After this the two factions became bitterly opposed and eventually Henry, Chief Hwata XII, led his own following away from the Chiweshe Tribal Trust Land to settle in the Zambezi Valley in an area recently offered to him by the administration. I consider it extremely significant, in the light of this succession dispute and the rivalry between the Nehanda mediums, that within a year of the failure of the faction she supported to get the decision to appoint Henry as chief, reversed, the Mrewa medium of Nehanda should have presented herself for testing and acceptance by Mutota’s medium. In October, 1963, she then sent the child medium of Gutsa to have his possession validated by the senior medium.

The Mrewa medium’s rival at Mhondoro was a famous medium, extremely well-connected through her marriage to Chief Mashayamombe whose dynasty has long been associated with anti-colonial risings and political movements (see Ranger 1966). I suggest that the Mrewa Nehanda’s moves to re-activate the “dormant” link with the Mutota cult and also to refer to the senior medium candidates for key mediumships such as those of Chaminhuka and Chikare, were tactics in the competition with her formidable rival whose position, possibly, was so well established that she did not consider it necessary to seek any external connection or validation.

It is also clear, that through the late fifties and early sixties, there was a general recrudescence of the cult of local spirits among the Zezuru. I observed it partially in 1958 in the Mrewa District, and Fry (1976) has considered it at length. He has argued that the recrudescence was associated with a general cultural and political movement away from institutions associated with colonialism in a situation in which overt political action was severely restricted. Many new mediums appeared during this time, some claiming to represent spirits long forgotten or which had not ‘spoken’ for many years. There were many cases also in which there were multiple claims to the same mediumship. As a consequence of this flood of new mediums, ’it seems highly likely that mediums occupying the key mediumships, which had on the whole endured, had begun to sort out rival claims and to impose some order, albeit only partial, on an extremely fluid situation. Without some order, their own positions would become uncertain and threatened. One way of doing this, I suggest, was to turn to the Mutota cult which, with its high prestige and hierarchical order, could fix the positions of at least some of the key Zezuru mediumships in relation to each other and then validate the claims of lesser mediums referred by their incumbents.

To substantiate this argument and develop it further, I now consider the case of a woman, aged about 40 who came from Nyamweda in the Hartley District (about 130 miles to the south). She claimed to be possessed by the spirit of Chikare and to have been operating as a medium for some years. She said that she had been sent by the medium of the Zezuru Nehanda from Mrewa. Chikare is a famous local spirit in the Zezuru cult and, in 1896, her then medium, together with the medium of Goronga, was prominent in the uprising (Ranger 1967: 211; Gelfand 1959: 17).

The medium of Chikare arrived in February 1964 with a retinue of 17 people including drummers, mbira players and ritual officers. At a public seance held at the small dancing arena outside the spirit house the senior medium of Mutota asked the women, now possessed, who she was. “I am your daughter, Chikare”, she replied. “I was lost for many years but now I have found you. I come to greet you, father.” This response clearly staggered the senior medium. He became silent for a while, growled, then hummed a Korekore war song, he took snuff, gargled with water,[25] sang another song, then broke into mamhuko which occupied him for another half-hour: all devices which I had noted that he used when he wished to delay making a response. He then said abruptly, “I have only one daughter, Nehanda”. Chikare’s medium replied, “I am your daughter, Chikare. I am the elder sister of Nehanda. I was born after Nyajore. My elder brothers were Chiobvu, Sundo ...” And here she began to recite part of Mutota’s genealogy. She ended, “I was lost in the war. I was reared in another place. My younger sister Nehanda found me and she has sent me to greet you. I greet you, father”. Again, the senior medium hestitated. Suddenly, he said, “Yes! Yes! You are my daughter! Truly, you are my daughter! He then went on to describe how Korekore had fought many wars and how during the fighting some of his wives and children had been captured or fled. In fact, what emerged was a very similar story to the one which Madzomba’s medium was said to have presented when he was tested and formally incorporated into the Mutota cult. Thus, Chikare’s medium was accepted. Some months after she had departed a steady stream of supplicants and candidates for mediumship began to come from the Hartley District where she lived, directed by her to Mutota’s medium.

The evidence of this seance and the external evidence relating to the names given in the genealogy of Mutota demonstrate how certain spirits of the Zezuru cult become added to the Korekore pantheon of remote ancestors and how their mediums become formally incorporated into the cult of Mutota. Chikare’s mediumship was long established, of high prestige and was associated with a set of lesser mediumships I have indicated that all the mediumships incorporated into the Mutota cult were of this type. I have suggested that in periods of flux or when there is competition among rival claimants to a key mediumship, one of the ways in which some order may be imposed, albeit partial, within the Zezuru cult, is by reference to the Mutota cult.

Conclusion

In general terms, two processes appear to have been occurring in the cult of Zezuru local spirits in relation to the Mutota cult: one in which mediums were simply seeking external validation and the other in which certain key mediumships were fully incorporated into the Mutota cult. The first process can be related to the tendency within the Zezuru cult for a number to appear all claiming to be possessed by the same spirit. In part, as the Hwata succession dispute illustrates, this tendency can be related to secular processes occurring within chiefdoms. Factions seek validation for their actions from an ancestral spirit through a medium: each faction has, or produces, its own medium. This in turn necessitates factions seeking further validation for such mediums from other, more prestigious mediums. Here again, however, the same pattern appears to be repeated. Each faction seeks out from a number claiming to be possessed by the same spirit, one who will validate its own medium. In such situations, therefore, it seems that further validation is sought completely external to the Zezuru cult. No doubt, this tendency was exacerbated during the recrudescence of the Zezuru cult. The second process, however, is much more complex for it is not simply a matter of the external validation of a medium but involves his or her actual incorporation into the Mutota cult. The Mutota cult, with its high prestige and hierarchical order, validates and fixes at least some of the key Zezuru mediumships in relation to one another. But, for the Mutota cult, the actual incorporation of such mediumships within it is an aspect of its expansion into the domain of the Zezuru cult. The medium of Nehanda, in referring Chikare’s medium for testing and incorporation was, in part, using the Mutota cult in her competition with her rival. She appeared to be attempting to establish a set of senior mediums, with their subordinates, all acknowledging her as the one, true medium of the Zezuru Nehanda. But validation and incorporation are not simply cognitive phenomena, they have far reaching organizational implications. The act of incorporation brings mediums of both cults into one ritual field. Within this field, through the network of relationships among mediums, people, tribute and communications flow.

The ritual field has cognitive and organizational aspects or sub-fields: the first is relatively stable through time, the second, inherently unstable. Cognitively, the Zezuru and Korekore cults become partially interlocked by the translation of one set of spirits into the other: Zezuru local spirits become Korekore ancestors. Once this has been accomplished and buttressed by supporting myths, the cognitive field, as the prime example of the translation of the Nehanda spirit illustrates, appears to be stable and relatively enduring. It provides a model and an idiom which certain Zezuru mediums can use, from time to time, to attach themselves to the Mutota cult. But it must be clearly understood that the fact that a Zezuru spirit is translated into a Korekore ancestor does not mean that one Zezuru mediumship is permanently incorporated into the Korekore cult even if the Korekore mediums and particularly the senior medium of Mutota conceptualize it thus. Mediumship among Zezuru is radically different from mediumship among Korekore. Organizationally, there is no one fixed set of offices in the Zezuru cult arranged in an enduring hierarchy. There may be several mediums, each with considerable local support, claiming to represent the same spirit. The fact that one of these is recognized and incorporated into the Mutota cult does not permanently resolve the question of who is genuine and who is not. What it does do, I suggest, is to bring certain mediums into a common organizational frame, for limited spans of time. Nevertheless, much of significance within the Zezuru cult continues to occur outside of this frame.

Inherently, in its organizational aspect, the ritual field is unstable. This instability arises within the Zezuru cult because of the charismatic nature of mediumship, the related phenomenon of multiple possession and the resulting lack of a clearly defined organisational hierarchy. There is always a striving for coherence and stability within this field by certain mediums who attempt to bring other mediumships under their control but this very striving, since it involves competition and the denial of the claims of others, produces further instability. Incorporation within the Korekore territorial cult cannot permanently fix the relations of Zezuru mediums to one another. It can only impose a temporary, fragile, stability. This is because the mediumships incorporated into the Korekore cult are not offices in the strict sense with a single incumbent and regular succession: they depend greatly on the charismatic qualities of particular mediums and the extent to which they can expand their vicinities. By contrast, mediumship is an office within the Mutota cult with rules regulating its incumbency. Nevertheless, here, too, there is a charismatic element and I have stressed how the cult’s expansion depended upon the personal qualities of its senior medium, George Kupara.

I have examined the ritual field created by the interpenetration of the two cults largely from the perspective of Kupara. It is clear that under his leadership, the Mutota cult has grown to prominence since the twenties and now greatly overshadows the formerly powerful chthonic Tavara cults and other neighbouring Korekore cults. Uniquely placed, through family connection and having personal qualities which enhanced his formal role, Kupara maintained and expanded the connection between the Korekore and Zezuru cults. Nevertheless, it would be an error to view the ritual field formed by the interpenetration of the two cults as extensively controllable or manipulable by him. He was but one of a number of foci of ritual authority. The interpenetration of the two cults cannot be understood in terms of a grand strategy formulated by him or by Zezuru mediums, either acting separately, or in collusion. Only within limited local arenas is it meaningful to speak of strategy. I have touched on the Zezuru Nehanda’s rivalry with the Mhondoro Nehanda in relation to the Hwata succession and suggested that her desire to have herself accepted by the Mutota cult, together with other rqediums supporting her, was an aspect of this. Similarly, the senior medium of Mutota can be shown to have acted strategically in relation to the chthonic cults, certain Korekore succession disputes (see Garbett, 1966) and in his relations with other Korekore mediums. The appearance of strategy, when the total ritual field is viewed over time, arises not from the manipulations of particular mediums, coordinated in terms of some overall plan — the key actors rarely if ever meet and do not collude — but from the emergent properties of the ritual field itself. These arise from the dialectical relationship existing between the two cults. The order and relative stability of the region of the Mutota cult attracts Zezuru mediums as offering some solution to the indeterminancy of their positions within the Zezuru cult, this in turn leads to an increase in the prestige of both the Mutota cult and its leader. Thereby, its leader is able to establish the ascendancy of the cult, expand his own vicinity, and increase his ritual authority over his subordinates within the cult which further increases the cult’s order and stability. However, the order which attachment to the Mutota cult imposes on Zezuru mediums is only partial and transient. Of itself, it creates further instability, basically because order and hierarchy are fundamentally incompatible with the charismatic, non-hierarchical nature of Zezuru mediumship. Finally, the charismatic element in Korekore mediumship itself suggests that with the death of the senior medium of Mutota and the collapse of his vast vicinity, many of the organizational linkages between the Zezuru and Korekore cults will be ruptured. Only at the cognitive level will the links endure, carried forward and elaborated in myth, to form the basis for the dialectical generation of new patterns of organizational inter-connections between the two cults during the next incumbency.

References

© Kingsley Garbett 1977

4. Bush Negro Regional Cults: A Materialist Explanation[26]

by Bonno Thoden van Velzen

Introduction: Radical Economic Change and Cults

In the 188O’s, a centralised, hierarchical cult appealing to a High God— Gaan Tata[27] arose among a group of Bush Negroes in Surinam’s interior. Around 1890, the cult spread rapidly throughout the six Bush Negro tribes and across most of their numerous rivers, and became the established cult of the whole region. Covering great distances, the Bush Negroes made their pilgrimages to the cult’s central shrines on the Tapanahoni River, in the south east of the country (Spalburg 1896–1900). A subsidiary shrine established by Gaan Tata’s priests at Santigron, only fifteen kilometres from Surinam’s capital Paramaribo, was visited by an impressive number of Creoles from the coast (De Ziel 1973: 40). Thus for a period the cult’s clientele came from various ethnic groups, with widely different backgrounds. Then the cult declined. In some valleys of its former region it ceased to have any adherents at all. Elsewhere it became fractionized and fell apart into a number of local congregations. While the central oracle’s priests continued to wield considerable power in Gaan Tata’s heartland on the Tapanahoni, their authority over other Bush Negro areas was only nominal. First and foremost, it had been a regional cult of entrepreneurs, of successful boat owners who controlled river traffic when gold was discovered in the interior of Surinam and French Guiana. They brought fortunes back home, until the transport market went from boom to bust with the collapse of the gold and balata industries in the 1920’s. The flow of tribute towards the cult’s central shrines became a trickle, the regional authority of the Gaan Tata priests had vanished. On the Tapanahoni, a parochial Gaan Tata cult continued to dominate religious and political life until 1972. Then a new prophetic movement swept this parochial cult away.

This paper will explain the rise and fall of the Gaan Tata cult. The explanation suggests that the emergence of the cult and its later decline is a dependent variable of radical economic change. The alteration of a mode of production is responsible for the birth of the Gaan Tata cult and its first flourishing decades; later reduction in income and resources without any great changes in the mode of production caused the fractonization of the cult. In the 1960’s, a new pattern of resource distribution within a new political context invited reactions from a new type of religious leadership. Intervening variables, such as the expenditure of resources within the cult on death rites and the destruction of the property of witches, influenced the cult’s career also, as riches and regional surplus turned into scarcity and poverty. But these variables are secondary. What is primary is the economic change, coming from outside the cult region, which introduced and later destroyed its basis. With the economic impoverishment of the region, Gaan Tata as a regional cult ceased to be viable. Prophets with aspirations for regional leadership had to await the introduction of new resources into the area; they came at the end of the 1960’s. A new cult is now attempting to establish itself on a regional basis.

As for the Bush Negroes or Maroons themselves, they are descendents of slaves, brought to Surinam from West Africa. Their ancestors, escaped from the plantations, fought a successful guerilla war against the Dutch planters and their mercenaries, and managed to establish semi-independent tribes in the interior. The most populous of the six Bush Negro tribes are the Djuka (about 4,000 in 1900 and 15,000 in 1970). They have their villages both in the interior, to the south, along the Tapanahoni, Maroni and Lawa rivers, and in the coastal plain, to the north east, along the Cottica, Commewijne and lower Saramacca rivers (see Map 1). The other groups, in rough order of size from the next largest, are located as follows:

1. the Saramaka in the centre of the country, along the Suriname River and its tributaries;

2. the Matawai, a much smaller group, to the west along the upper reaches of the Saramacca River;

3. the Aluku or Boni to the east on the banks of the Lawa River;

4. the Paramaka, also in the east, along the Maroni River; and

5. the Kwinti, an offshoot of the Matawai, to the far west along the Tibiti and Coppename rivers.

In 1970, there were approximately 40,000 Bush Negroes in Surinam.

Heydays

It was in the second half of the nineteenth century that the hinterland of the Guianas became a focus of international economic interest with the discovery of gold. Although deposits were scattered widely over Surinam and French Guiana, they were mainly to be found in the interior which is covered by the dense South American rain forest. So transport for the gold-diggers had to follow the rivers flowing to the Atlantic from the mountaineous region on the border with Brazil. This form of traffic posed serious problems: about 40 to 70 kilometres from the ocean all rivers are blocked by natural barriers of rapids and falls: and worse still, these obstacles repeat themselves every 5 or 10 kilometres. On each trip, a boat crew had to look for a path through the rapids which was different from the one taken in preceding months: water levels were different from one month to the next as a result of variations in rainfall. There was only one group that could unlock the interior for the thousands of gold-diggers that came from all places of the earth in search of El Dorado: the Bush Negroes. They had the means of conveyance adapted to these shallow and treacherous waters: the dug-out canoe. Only they knew where to go in the continuously shifting meanders of rapids, and they had the nerve and the skill to do so. Their only competitors were the Amerindians, but their numbers were too small to influence conditions on the transport market. Quickly, the Bush Negroes gained a monopoly over river transport. They used it to exact high wages and high freight prices. The best way to bring out their strong economic position is to compare it with that of the elite among the Creole workers in Paramaribo, the capital. Creoles under contract with gold companies or balata bleeders considered themselves fortunate if they earned two florins (approximately $0.80) a day, well above what other labourers in Paramaribo could hope to get. Around 1900, it was impossible to employ a Bush Negro for such wages; employers were tucky if they could hire him for double that amount. The income of the Bush Negro boat owner was truly astonishing, in contemporary terms. Between 1880 and 1920, the average income of these entrepreneurs seldom fell below 2,500 florins ($1000) per annum, and for many years was well above 3,000 florins ($1200) (De Beet and Thoden van Velzen 1977).

An appreciation of the fortunes these entrepreneurs earned is one important step in the argument. But to understand the rise of the Gaan Tata movement, it is even more important to consider the radical changes that had taken place in the mode of production. Formerly, before 1880, Bush Negroes were predominantly lumber workers. In large numbers, they had left their villages on the Tapanahoni to settle in permanent villages or temporary bush camps along the banks of the Commewijne, Cottica and Courmotibo rivers. They either had fixed places of residence, or stayed long enough at one place, to be accompanied by their wives and children. Often, a number of matrilineal relatives made the move of residence to the Cottica region together. The lumber workers clung together because they regularly needed help for hauling tree trunks to the river banks. This was an ideal economic underpinning for the ideology of the corporate matrilineal group with its stress on mutual sharing and responsibility; it also meant a boost for the distributive mechanisms for the levelling of income differences which had been part of Bush Negro culture for more than 200 years.

River transport altered the organization of work considerably. Between Albina and St. Laurent at the mouth of the Maroni River on the one hand, and the main gold fields of the Lawa on the other, lie 270 kilometres of river; to the rich gold fields of the Inini the distance was even greater: 350 kilometres. A Bush Negro who set out to work as a river carrier did not need the help of many relatives: the services of one or two sons or sister’s sons would be quite enough. He could also team up with one adult kinsman or a friend and the boat crew would be completed. He did not any longer depend on the assistance of his matrilineal kin; quite the contrary, they had become a liability. Thus, the entrepreneurs could amass small fortunes, while there was hunger and deprivation in many villages of the Tapanahoni River.

The Gaan Tata cult was sparked off almost simultaneously with the first gold rush in the Lawa-Maroni basin (1885–1888). As we have seen, the thousands of gold-diggers that came from all over the world, were willing to pay incredibly high prices for transport to the Dorado.[28] Elders, who were no longer strong enough to carry heavy equipment through the rapids, and some spirit mediums, could siphon off some of this new wealth through the Gaan Tata oracles (Morssink 1934). As part of Gaan Tata’s commandments, the entrepreneurs were given a set of standards that proved their usefulness in contacts between Bush Negroes from various areas and with different cultural backgrounds; the cult, for example, gave religious sanction to peacefulness: even the shedding of a single drop of blood could arouse Gaan Tata’s wrath. The priests also guaranteed the moral stature of its believers by subjecting them to an ordeal. Thus, the residents of villages and bush camps along the ‘traffic lanes’ could invite the boatsmen to spend the night with them without any risk. But larger than the threat from the external enemy, loomed that from within the once close-knit kin group, the matrilineage. Gaan Tata priests assured the entrepreneurs that their God would take care of any envious relative who had turned his eyes on them. Significantly, any boatsman who established a good relation with the Gaan Tata oracles through timely presents and acts of deference—soliciting help in case of illness—could rest assured that his reputation would be protected. Any accusation of witchcraft or other malevolent acts directed at these ‘friends of the oracle’ would be declared null and void at these centres of communication, the only ones in the Bush Negro society.

Other clues for our understanding of the Gaan Tata cult come from looking at the wider political scene of which these Bush Negro communities formed part. The land to the southeast between the Lawa and Tapanahoni rivers had been contested by France and The Netherlands, until it was allocated by arbitration to the Dutch in 1891. Yet relationships between the two colonial administrations were to remain uneasy for a long time, both after all shared the Maroni-Lawa basin, rich with gold. The Dutch considered it vital that there be cooperation between them and the Djuka in particular, for control of areas where colonial officials seldom penetrated. Harmonious relations were dented somewhat by too obvious Djuka sabotage of mission schools and by their resistance to Dutch exploration of the borderland with Brazil in 1904 and 1907 (Franssen Herderschee 1905; De Goeje 1908). But before the great strike of the Bush Negro boatmen in 1921, the overall picture remained one of calculated but usually satisfactory partnership.[29] Founders and early leaders of the Gaan Tata cult, such as the Djuka tribal chief Oseisie, and to a lesser extent also Agumasakka, felt entitled to remuneration and sought, first of all, to secure favours from the colonial administration: they were taking care of Dutch interests in the interior. Indeed, on several occasions, cult leaders and headmen of the Djuka Bush Negroes had cooperated with the Dutch in attempts to curb the price rises of the freight carriers, which the gold companies considered excessive. Moreover, against the missions, the Gaan Tata priests were usually on the defensive, preferring a modus vivendi to open war.[30] Thus an external enemy, the colonial administration or the missions, was not the prime target of the cult. The cult was directed at eradicating an internal evil, witchcraft, that was felt to dog the steps of the nouveaux riches. Society should be purged of its hidden enemies, the witches or ‘traitors within the gates’ (Mayer 1970: 60). This was the matrix of economic and political conditions in which the Gaan Tata cult arose. We shall later see how deeply the theology and basic ethics of the cult were influenced by these conditions. But first we will present a brief account of the spread of the cult and its centralization.

The spread of a regional cult and its centralization

The new cult orginated around 1885 to the southeast in the interior among the Djuka in the valley of the Tapanahoni River, had its first successes in Djuka settlements in the coastal plain in 1891, but soon made inroads into the territory of other tribal groups as well. There was at least one western village of Saramaka which had accepted Gaan Tata worship before 1892 (Albitrouw 1892–1896; Albitrouw + 1915), but it seems likely that other Saramaka villages had also gone over to Gaan Tata at a very early stage.[31] The Gaan Tata priests had even more dramatic successes in 1892 and 1893 in the far west in the villages of the Matawai, who had earlier been the first to be converted to Christianity; and in 1895 also in the far west, even the small and remote group of Kwinti came under the spell of Gaan Tata (Kraag 1894–1896). Right from the beginning, a stronghold of Gaan Tata in the northwest was the village of Santigron, which had a mixed population of Djuka, Saramaka and a few Matawai. It was from here that Gaan Tata’s advocates travelled south to Matawai villages in 1892 and 1893. Its location near the coast and less than fifteen kilometres from the capital of Surinam, Paramaribo, made Santigron a place of pilgrimage for many Creoles in the city who heard about the new powerful god (De Ziel 1973: 40). Special expeditions were organized by Christian missions to preach against the ‘false God’ and undo the impact of Gaan Tata’s message on their Matawi following (BHW 1895: 12–42). But despite their efforts to wipe out the cult, they were only partly successful; a quarter of a century later Gaan Tata was still worshipped in secret in a few Matawai villages.[32]

In its heyday the cult had a regional organization which kept its leaders in contact with one another. In the 189O’s the leading priests at the regional centres on the Tapanahoni river kept in touch with believers on the Surinam River and its tributaries (NB 1904: 255) and sent messages to their junors, priests such as Brokohamaka, at the northeast’s main centre on the Cottica River. In their turn, these northeastern priests were in communication with others in the northwest at Santigron (NB 1892: 574) and in the west at Kriki-Pandasi on the Sara Creek (MBB 1893: 181). Even in the 1920’s, there were still lines of communication with Sara Creek villagers and with a Saramaka village on the Pikin Rio tributary of the Suriname (Junker 1923).

The organization of the cult was centralized and hierarchical. Bush Negroes from other rivers came to the centre on the Tapanahoni, not the other way round. The hierarchy was also asserted through the constant flow of tribute towards the Tapanahoni shrines. In 1894, every settlement on the Cottica and Commewijne rivers, which boasted of a village headman, had to pay a special tax of 128 florins ($51) to the Gaan Tata priests of the Tapanahoni (MBB 1895: 53). In 1917, a small Saramaka village paid 320 florins for the right to open a Gaan Tata shrine, an enormous sum of money at that time, more than what most labourers in Paramaribo would earn in a whole year (Junker 1925: 154). That money was sent directly to the Tapanahoni River and the Djuka village of Dritabiki, then the main place of worship for Gaan Tata. Significantly, not all that was owed Gaan Tata was paid, at least not immediately. It was reported to the colonial administration (Junker 1923) that at least 19 Djuka from Sara Creek villages were heavily in debt to Gaan Tata priests at Dritabiki. Even some Christians in the area were helping to collect the money so that the curse of Gaan Tata, then threatening their matrilineal group, might be bought off. Many of the debtors were living in poverty with not enough food to eat. In 1923, the most prestigious of the village headmen in the Sara Creek area was held responsible for the collection of these debts. The threat was: If payment did not arrive, Gaan Tata would kill him! (Ibid.). Even more important for our understanding of the dominant position of the centre, was the obligation to bring the legacies of witches to the Tapanahoni shrines (MITT. 1896: 76). Leerdam, writing in 1957, but probably referring to a much earlier period, noted that the effects of the dead who had been found guilty of witchcraft were with great difficulty transported from the Sara valley to Dritabiki. Instead of crossing the watershed between the Sara and Tapanahoni, which was too arduous with fully loaded boats, they followed the Sara and Suriname rivers downstream, then east along the Cottica, crossing the watershed to eastern Surinam, and finally upstream again along the Maroni and Tapanahoni rivers. And all the time they had to be on their guard not to lose or break a single object on their long journey, while hauling their boats through the rapids and’carrying the goods over slippery stones at difficult spots.

Although there were many Gaan Tata shrines and oracles left in the 1960’s, worship was on the wane among the Saramaka and had disappeared almost completely from the Matawai villages. Moreover, there was hardly any communication between the centre (Tapanahoni) and remaining outposts (Santigron and Totikampu on the Saramacca, Agiti-Ondro on the Cottica, Dangogo on the Pikin Rio and a few other places). The authority of the tribal chief who was also high priest of the Gaan Tata oracle, was recognized in theory but not in practice (cf. Kbbben 1976). The cult’s decline and fractionization was fundamental: no longer did the Djuka from the Sara Creek, Commewijne or Cottica rivers (i.e. from the central, north, and northeast areas) take pains to bring the possessions of the witches across numerous rapids and falls to lower these in front of the Gaan Tata priests at Dritabiki on the Tapanahoni River. In the central and northern sectors, around villages such as Santigron, Agiti-Ondro and probably also Dangogo, vigorous but parochial cults survived. Nowadays, among the Kwinti of the Coppename River, the process of fractionization has reached the point where the small group of Gaan Tata adherents even deny the Djuka origin of their parochial cult (Personal Communication, Chris de Beet). On the Tapanahoni religious life had all the appearances of normalcy with the Gaan Tata oracles at Dritibiki and Granbori unchallenged. That was the situation in the 1960’s to which I shall return at the end of this paper.

Theology

The corner stones of Bush Negro religion are four, five or sometimes six pantheons, each independent of the other. A supreme deity is recognized whose role is seen as so restricted that we may safely speak of a deus otiosus. at regular intervals throughout the history of the Bush Negroes, this ‘traditional’ cosmology is challenged by a Great Deity or High God cult. Central to this alternative world view is a supreme deity who is active in the world and does interfere in human affairs. Man’s aim in life, first and foremost, is to find favour in the eyes of this new deity, who will protect his faithful followers from other supernatural powers and dangers. His commands are rather specific, and when his demands are met, such as refraining from witchcraft, man can count on divine grace.

This view of man’s destiny is radically different from the traditional conception, which makes success and harmony hinge on a proper compromise that man has to find between the claims of diverse supernatural agencies. In various ways, this Great Deity (Gaan Gadu) cult suppresses or contravenes traditional forms of religious behaviour, such as spirit possession. The cult strikes against the vested interests of ‘traditional’ religion. Each wave of Great Deity fervour is accompanied by a radicalization of attitudes; feelings of inadequacy and remorse over past failures express themselves in acts of iconoclasm and neglect of daily economic and social routines.

The Gaan Tata cult is one of many Great Deity of Gaan Gadu movements in Bush Negro history. Gaan Tata, also known as Gaan Gadu, Bigi Gadu, Massa Jehovah, Gwang wella or Bakaa (European or Outsider)[33] Mitt 1895: 52), is the God that led the Djuka out of slavery and guided them in their battles with the colonial mercenaries. In the 188O’s, with the fears of witchcraft steadily rising, he was called upon to safeguard his people. But it was believed that he would only protect those who were pure in heart. They had to enter into a covenant with him by drinking a potion (Swell). This was the religion’s main sacrament and ordeal: those who survived proved themselves to be the faithful worthy of God’s protection; those who died were the witches. There was a marked preoccupation with the danger of witchcraft and most ritual was concerned with it.

However, the vindictiveness of this new deity was a familiar phenomenon in Bush Negro religion, and a characteristic he shared with many deities of ‘traditional’ cosmology such as avenging spirits (Hurault 1961: 221–235; Price 1973; Thoden van Velzen 1966b). More unfamiliar was the fact that this High God, like Jehovah, not only scrutinized their deeds but followed their thinking and feeling as well. He was the God “that looked down into their hearts” and from whom no evil thoughts could be hidden. Punishment would be meted out to those who nourished feelings of envy, hatred and resentment: these states of mind seduce the individual to employ witchcraft. Along with the commandment “Thou shalt not commit witchcraft”, there were other moral injunctions, prohibiting suicide, physical aggression and upholding the integrity of cultural institutions that markedly set off Bush Negro culture from those of other ethnic groups in Surinam. The emphasis the Gaan Tata priests placed on the rules of menstrual seclusion, is a case in point. At the same time, it should be stressed that the Gaan Tata cult acquired surprisingly few anti-white or even anti-missionary features. With great success, Gaan Tata’s priests succeeded in hiding.rites and holy places from the eyes of the outsiders, but they never preached militancy against the whites. This was a corollary to the exigencies of common interests between Djuka Bush Negroes and colonial administration, as indicated above.

These moral demands also aimed at underpinning peace and cooperation ii. a wide region and among larger groups than lineage and village. The good man no longer was merely a good kinsman or neighbour perhaps, a man who fulfilled all obligations and was loyal to his kith and kin. Virtue could now be defined according to a universally valid standard: the good individual is not corrupted by the evil powers of witchcraft. Successful submission to a poison ordeal, a drink from the sacred potion, proved his purity and he could now move freely in wider circles.

Theology and ethics remained basically unchanged in the central sector of the cult along the Tapanahoni river before the new parochial cult arose (see Thoden van Velzen and Van Wetering 1975) and even after the Gaan Tata cult declines and became fractionized into a number of local congregations. However, the edge of iconoclasm had worn off. The major moral norms, those prohibiting witchcraft, suicide and the infliction of serious physical harm, were upheld as vigorously as in the early years. But some transgressions no longer were counted as sins against Gaan Tata: adultery, for example, came to be treated as a purely secular offence. Apart from the mild disdain shown by Gaan Tata priests to other manifestations of religiosity, such as fits of spirit possession, Gaan Tata lore was welded into perfect unity with traditional beliefs. Nevertheless, to a high priest from the cult’s central sector, the Gaan Tata ritual that others performed elsewhere seemed to be a lapse into error. After one of the last of the high priests had occasion to visit the Cottica River and one of the cult’s few remaining outposts, he publicly complained about the many ways his colleagues had strayed from the true path of Gaan Tata worship. He mentioned some examples: offerings were made at the wrong shrines; liquor was used for libations at the most holy places while it should have been beer (“that biting, burning stuff at such a holy place!”). At the same time, he confessed that there was very little he could do about it: “People from the Cottica are headstrong, they don’t listen to me, later they will agree that I was right, but then it may be too late!” The process of fractionization had gone too far to reverse.

Allocation of resources and internal distribution

The boatmen brought about an enormous influx of resources into Djuka society. However, there were very few possibilities for investment open to these entrepreneurs. To acquire more land made little sense: land was an open resource, readily available, but the labour force that could be hired was small and, above all, the costs of exporting agricultural products over hundreds of miles, over rapids and falls, would be prohibitive. The only investment that made sense was the replacement, every few years, of the big dug-out canoe; however, the 200 guilders required for the purchase were but a small percentage of what could be earned within that period of time. To marry more women was another possible investment, but it was a risky policy because it meant incurring cumbersome obligations towards each wife’s matrilineage. This may have been an important inducement for many Bush Negro entrepreneurs to marry Creole women from the coast. What usually happened was that the entrepreneur saved his money by burying it in bottles in the ground, spent his money on luxury goods for ostentatious display or— less important in quantitative terms—bought presents for elders in positions of power and influence.

Given this evidence that the possibilities for entrepreneurial investments were extremely limited, we should now devote our attention to the societal mechanisms of resource distribution, to levelling practices and to groups that controlled the internal flow of resources. Two of these groups need to be discussed: the traditional association of the gravediggers and the priests of Gaan Tata. The questions posed are: to what extent and in what ways did these groups profit from the new affluence and what did this signify for the position of Gaan Tata as a regional cult?

One of the most important institutions of the Gaan Tata cult was the confiscation of the legacies of all those who had died at the behest of this wrathful deity. The traditional inquest into death was the basis on which the decision was taken whether or not the goods of the dead had to be cleansed and then confiscated as God’s cargoes. The core of the inquest was the ancient West African tradition of ‘carrying the corpse’ (Rattray 1927: 167–174), kept alive by Surinam’s Bush Negroes. The corpse was tied to a litter and carried on the heads of two bearers through the village. Questions were put to the deceased’s spirit by elders following the bier. The spirit replied through the bearers’ movements. The new element was that the Gaan Tata priests were not satisfied with answering the question of cause—which supernatural agency or what person killed the deceased; they also insisted on receiving a verdict on his moral character: each death was then classified as either a witch’s death (wist dede), a sinner’s death (misi dede) or a respectable death (jooka dede). It was only the property of the ‘respectable’ dead that did not have to become a God’s cargo. In 1962, as a result of these inquiries, seven out of every ten dead were stigmatized as witches (wisiman); of the remaining three, two were posthumously condemned as sinners against the laws of Gaan Tata, and only one allowed a ‘respectable death’.

The inquest that diagnosed the cause of death and determined the moral character of the deceased, was supervised by the association of gravediggers (oloman). As these posthumous verdicts were of obvious economic and social importance, we need to discuss the position of the gravediggers and their relation to the Gaan Tata priests in some detail.

During adolescence Djuka men have a choice between two associations: they can join the gravediggers or the coffin-makers (kisiman). The task of the last association was limited to what the name suggests. The coffinmakers had little work to do because ‘sinners’ and ‘witches’ were not entitled to a coffin; their corpses were left unburied at reserved spots in the jungle (witches) or put in a shallow grave (sinners). Thus, the coffin-makers convened only after the demise of a ‘respectable’ person. While making a coffin, gravediggers and coffin-makers indulged in horseplay at each other’s expense, and in sham attacks; and they attempted to get as large as possible a share from the communal meal at the end of a day’s work.

The gravediggers had many more responsibilities: they attended to the internment of the deceased and took charge of the ritual for the dead. From their ranks, they selected the bearers of the corpse and the interrogators of the deceased’s spirit. Immediately after a demise, gravediggers from the village where it occurred, and from neighbouring villages, would meet. Even gravediggers from far-away places who happened to be visiting or passing through were expected to participate. Within a couple of«hours after the demise, work would start. Usually, there would be two stints of carrying the corpse. Relatives of the deceased were invited to share in the tasks of carrying and interrogating. In this way, later charges of fraud were preempted by letting the relatives share in the responsibility of the verdict.

On the surface, relationships within the association of gravediggers appear strongly egalitarian, but a closer inspection shows that some held key positions. These leaders or bosses (basi), as they were called, assisted by a few elders selected from the relatives of the deceased, formed the committee that supervised the interrogation. This same body was also responsible for the decisive communique issued after the carrying of the corpse.

Within the arena of power, the bosses of the gravediggers were in no way competitors with the Gaan Tata priests. The committees could only convene for specific reasons — burial and death rites — often at unexpected moments, and on each occasion these bodies were differently constituted. Thus, although the decisions taken were important, the bosses could not develop strategies or even policies and therefore did not form a power group that could stand up to the Gaan Tata priests. Moreover, their political position was far from enviable. To prevent charges of fraud from the bereaved relatives, who could and often did accuse them of playing into the hands of the Gaan Tata priests, was one of their concerns. To prevent being blamed by the Gaan Tata oracle was equally pressing. In many cases, there was a chance that one or more of the bosses would either be a Gaan Tata priest or an elder sufficiently associated with Gaan Tata’s oracle to be partisan. The latter were apt to remember dubious statements, signs of hesitancy on the part of the bearers at certain questions, many months after the inquest had taken place. In order to escape the dilemma, the bosses of the gravediggers often clouded their statements with oblique phrasing, or, by clever wording, presented verdicts that allowed for later refutation by the Gaan Tata oracle (Thoden van Velzen 1975). For all these reasons, the association of gravediggers could not develop into a power group that would grow in importance as more and more resources flowed into Djuka society, nor were they to gain considerable emoluments from this position. Before turning to the only other group that controlled the internal flow of resources, the Gaan Tata priests, we first need to discuss some levelling practices that have been part of Bush Negro religion for centuries, i.e. the costly death rites. This is necessary, because they affected the life chances of the regional cult considerably.

There is a sequence of death rites, which are held at fixed stages after the death of any adult. The burial ritual which often lasts for five or six days, or even longer, is only a first stage. Then the gravediggers and coffin-makers have to be offered meals and gifts; and the dancers at nightly ceremonial occasions are regaled with refreshments. Moreover, the deceased should be honoured with clothes that are pushed into the roomy coffin. Three days after the burial there is another ritual, which again is repeated three months later. Both are accompanied by libations, and the ceremonial distribution of drink and food for the community. The relatives of the bereaved carry the brunt of it. Most expensive is the celebration closing the mourning period at some twelve or eighteen months after death. At this anniversary the feasts may last for more than a week and include a banquet for the ghosts of the ancestors, a large-scale distribution of food and daily libations. Anniversary expenditures run into hundreds of guilders. At the turn of the century, Spalburg (1899: 41) estimated the expenses as ‘usually above 300 florins’ ($120). Furthermore, there are always reasons for performing an extra ritual to mollify the spirit of the deceased: perhaps a lack of piety shown by the surviving spouse or a remarriage without proper ritual care. Whatever the case may be, the conclusion is easily arrived at that a great wrong had been done to the deceased during his lifetime so that a conciliatory ritual to placate the ghost is now called for. All told, enormous amounts of resources, in time and in money, are spent on these death rites.

The Gaan Tata cult introduced additions and changes to these rites which had.far-reaching consequences. Not only were the death rites for ‘witches’ and ‘sinners’ made less elaborate and less costly. Their deaths ceased to be occasions for economic transactions exclusively within local and kinship relations. Their deaths provided occasions for the removal of goods from these relations, rather than for the consumption of food and the redistribution of durables primarily among relatives. The goods became regional in their significance. The relatives of a witch or sinner were obliged to bring his effects as God’s cargoes to the cult’s regional shrines situated in the village of Dritabiki on the Tapanahoni River. During one year in the later period of the cult, in 1962, 35 of such cargoes were received by the Gaan Tata priests. At first sight, this series of rites appears to be a levelling institution par excellence. The obvious course for the priests would be to try and lay their hands on the effects of the wealthiest among the deceased, thereby levelling most economic differences but creating another rift within Djuka society— the priesthood versus the laymen. But is this what actually happened; were these the consequences for Djuka society? An examination of the daily routines of the Gaan Tata priesthood shows that the answer should be a qualified ‘no’.

First, part of the God’s cargo was returned to the relatives, approximately 30 to 40%. Of the remainder, many valuables and all cash were kept by the priests. But an astounding amount of goods, part of it valuable, was left to rot at Santi Goon, Gaan Tata’s secret shrine in the bush. This clearly demonstrates that they had not taken the God’s cargoes solely for personal enrichment or an income for themselves, their favourites or the rest of the cult’s staff. They clearly had set aside a large portion for Gaan Tata. Purchased goods—household utensils, Victorian bric-a-brac bought by the early entrepreneurs, mantlepiece decorations, gold chains, shoes, plastic wares—were abandoned. Great heaps of goods in varying states of decay remained from the beginning till the last days of the cult. In 1972, adherents of the new prophetic cult salvaged some of the durables and brought them back in boat loads for redistribution.

Secondly, we should not ignore the fact that the relatives of the bereaved and the gravediggers often surrendered the effects voluntarily, even thrust them upon the Gaan Tata priests so that these would discharge them of the dangerous legacy of ‘the traitor within the gate’ and its polluting effects. Admittedly the evidence for the priests’ greed is unequivocal. In many instances I know of, they brought pressure to bear upon both gravediggers and elders of the village concerned to change the verdict of the inquest away from ‘a respectable death’. But many villages were too far away to be effectively controlled; political opposition to the religious establishment was strong at times; and the priests of Gaan Tata also had a reputation to protect. In brief, not all God’s cargoes that arrived at Dritabiki came as a result of applied political pressure or even in anticipation of it.

Thirdly, and perhaps most significantly, is the information on the social categories of persons that the cargoes came from. Van Wetering (1973): 59) provides some quantitative facts. Of the 116 deaths (59 men, 57 women) that occurred between 1960 and 1970, and on which information is available, only 39 posthumously obtained the verdict of respectability. And of these, only 13 were women, which makes the relation between ‘woman’ and ‘witch’ statistically significant (Van Wetering 1973: 59, 286). As women are generally much poorer than men, these data should caution against any quick conclusion of ‘levelling’. Moreover, it has been my impression that village headmen and wealthy elders usually did not meet with divine disgrace upon their deaths. But to predict at the death of an old woman what the verdict would be was not a hard task: a witch.

Thus, the priests by not being able to touch some of the bigger legacies, had no prominent levelling role. In fact, they sustained the entrepreneurs by protecting their reputation at the oracles and condemning poorer relatives for any gossip or slander aimed at these boatmen. In this way, they shielded them effectively from levelling coalitions within the village. Later, the new prophetic cult of Akalali would render the same services to the labour elite in pfesent-day Surinam. In conclusion, we can say that Gaan Tata religion did little to bridge the gap between rich and poor; it preserved and sustained divisions within their society.

Gaan Tata’s Bulwark on the Tapanahoni

In the 1960’s the parochial Gaan Tata cult of the Tapanahoni seemed little affected by the great changes that had taken place in Surinam’s interior after 1925. There were many oracles in the 30 villages of the Tapanahoni with their 6,000 inhabitants, but the Gaan Tata oracles of Granbori and Dritabiki held sway. Certainly in Dritabiki at least, things did appear ‘as of old’.[34] A group of about ten priests gave direction to the operation of the Gaan Tata oracle and they regularly convened for that purpose. During 1962, the oracle was consulted on approximately 125 days regarding a total of 424 cases. The clientele of the Dritabiki oracle flocked from most Tapanahoni villages, a sprinkling came from other Djuka areas and some were Saramaka or Boni. Most were ill, and wished to learn the reasons for their misfortune and what could be done about it. Although usually an illness was attributed to supernatural causes, the priests often managed to link these causes to lack of responsibility or even complicity of living persons. Sometimes, this sequence was reversed in that a person was ordered to appear before the oracle, where soon the priests proclaimed that their client’s life was threatened by evil forces almost impossible to ward off at this late moment. The client concerned could do little but beg the priests to help him survive. This assistance was never withheld and the honorarium demanded seemed trifling in comparison to the dangers averted, or at least such was the official view. But let us first examine the workings of the oracle.

In the Gaan Tata cult as in other Bush Negro cults, the deity made his wishes known through oracles. Just as a deceased’s spirit replied to questions at an inquest through the movements of bearers carrying his corpse tied to a litter, so too Gaan Tata replied to questions through the movements of the bearers of his oracle. This was a small bundle tied to a plank which two Gaan Tata priests carried. The bundle presumably contained sacred objects such as hair and nail parings of famous ancestors, and it was revered as the tabernacle of Gaan Tata. It was hidden from view by long draperies which touched the ground when the ends of the plank rested on the heads of two bearers. Other priests standing in front of the bearers put questions to the oracle. Gaan Tata replied through the movements of the bearers in this way, a forward move signified an affirmative, a backward or sideways one a negative answer. Wild, chaotic movements indicated the High God’s vivid displeasure. This ‘dialogue’ between god and priests always took place at some distance from the laymen so that they could not overhear what was being discussed. A high priest was in charge of the oracle: after receiving a whispered report from his subordinate priests, he issued a final communique to the meeting of elders. As a rule, the oracle sessions were held in public in front of the temple, an ordinary but bigger hut where sacred objects were stored. At the main oracle of Dritabiki, the elders of that village, and those of a few surrounding villages, were expected to attend these sessions.

Although an oracular session usually started with a patient soliciting for assistance, opportunities for political manoeuvres often presented themselves. Take, for instance, the case of a village headman disliked by the chief for his independent stance in many political confrontations. His wife had fallen ill, and although the case did not look very serious, he had consented to escort her to the oracle at Dritabiki. There the priests had a surprise in store for him. Those charged with interpreting the movements of the bearers, reported that the illness of the headman’s wife was of the gravest nature; a diagnosis that had to be taken seriously although all present at the oracle could see the woman walking cheerfully through the village. Subsequently, the village headman was blamed for being the source of all suffering: his irresponsible neglect of his wife’s ancestors had caused their wrath, after which the ghosts in retaliation had visited sickness upon the woman. The priest .charged the headman a stiff fee for treatment. The position of the village headman was undermined and he was punished for not being sufficiently compliant.

Roles of critical importance at the oracle, such as front bearer of the holy bundle, were often given to a high priest’s personal dependents or fugitive clients. Every high priest was surrounded by such people who had left their native villages in order to settle with a protector. Some stayed a few years, others prolonged their stay indefinitely. In some cases elders forced to flee their villages as witches were, a few years later, in a position to judge the actions of their former prosecutors: they, the fugitives, were carrying the tabernacle of Gaan Tata. But most of the clients had a much more humble task to perform. The female clients kept the high priest’s official residence clean, emptied chamber pots and did other household chores. Most other clients prepared and cultivated his gardens. If the high priest needed volunteers for a corvee at short notice, he usually recruited them from among his clients. The high priest kept this small, but loyal labour force till the middle of 1972, when the new cult had taken over. Fugitives then fled to Njoen Konde, the centre of the new movement and Akalali’s village.

Besides their share in the God’s cargoes, priests enjoyed other forms of income, such as fees. These were seldom exorbitant and for religious and medical treatment amounted to one bottle of rum, four pieces of cloth and twelve bottles of beer. The retail price of these goods in the Taponahoni region was about 25 Surinam florins (approximately $13). Occasionally, instead of payment in kind, money was demanded; sometimes as much as 32 Surinam guilders (approximately $17). In 1962, 62 fees were imposed but only 30 actually collected. There was no pressure to pay straightaway: a debt paid two or three years later was just as welcome. But if some one returned to the oracle after a couple of years with an unpaid debt for previous treatment, it was virtually certain that the outstanding debt would be indicated as the cause of this new affliction. In serious cases, the priests usually demanded an extra payment along with the usual fee for a consultation. Four litres of rum was the normal fee which had to be paid before a consultation could be held. But it did happen that the priests settled for less or even waived the session costs altogether.

One may wonder what the priests did with the large quantities of rum they received. Some of it went to the storage room in Gaan Tata’s temple. A part was consumed immediately: the priests gave one glass of rum to each of those present at the oracular sessions, but only when the last case for that day had been dealt with. The elders regarded this as a small compensation for the time spent at the meetings. Moreover, the high priest regularly presented his collaborators with a few litres of rum to enable them to entertain in a proper manner the numerous elders and patients who called on them. Additionally, a great quantity of rum was earmarked for the corvees organized for dignitaries such as the tribal chief and a few of the prominent priests.

Special taxes were sometimes imposed. In 1962, for example, the priests informed the faithful that they felt Gaan Tata punished his people too severely. The priests were quick to add that they would not object to a harsh and merciless treatment of witches. Their concern was with the respectable citizen who had suffered an occasional lapse and was likely to have his life taken away by their vindictive God. The high priest intended to mollify the Great Deity. Therefore, he proposed to the village headmen that each of the thirteen matrilineal clans of the Djuka on the Tapanahoni should donate 32 Surinam guilders ($18) for this purpose. The money was collected at the end of 1962 and handed over to the high priest who pocketed it and, I suspect, used it for his own convenience.

Gaan Tata’s Decline

Some of the forces that had destroyed the fabric of the Gaan Tata movement as a regional cult, were also at work undermining the parochial cult of the 1960’s. Let us examine these more closely.

First, the resource basis of the Djuka and other Bush Negroes, the total of all forms of income taken together, rapidly melted away in the 1920’s. The gold rush was over at the beginning and the balata boom at the end of that decade. As a consequence, there was less money to be tapped by medicinemen, spirit mediums or the priesthood of Gaan Tata. All who had lived off these entrepreneurs were then in trouble. Fewer patients came to the oracle-, they brought less money; gone also were the rich friends of the oracle who had bought them such handsome presents. As a consequence, the priests no longer had the wherewithal to add force to their arguments or even to send delegations and messengers to the Gaan Tata outposts.

As the impoverishment of the region grew, centrifugal tendencies became stronger. The question “why pay tribute to the far-away Tapanahoni, what can we expect to gain from them?” had been asked before, but was now posed with more force. Sometime in the early 193O’s, the Djuka of the northeastern area, the Cottica, stopped bringing God’s cargoes to the Tapanahoni. Other areas rapidly followed suit and soon developed their own organisation for the cleansing, redistribution and dumping of the God’s cargoes.

Although many Djuka and other Bush Negroes continued to work as freight carriers, the decline of river transport had set in around 1925. In 1921 (Koloniaal Verslag), it was mentioned for the first time that Bush Negroes were willing again to resume their old occupation of lumber work. This trend must have accelerated during the late 1920’s and the 1930’s with the consequence that the numbers of those who relied on the norms and ethics of the Gaan Tata cult for free movement among other Bush Negroes, were dwindling. General impoverishment tended to have the same influence. As the rift between rich and poor grew less significant, witchcraft fears allayed. With poverty spreading, the number of Bush Negroes who lived in terror of the malevolence of the envious kinsman or neighbour, declined. Now, there was less need for an anti-witchcraft movement.

After 1930, only a trickle of money and goods reached the main Gaan Tata oracles. Throughout the cult’s old region there was a drastic reduction in income. Moreover, where there was heavy expenditure on ritual costs, this went for a community’s death rites rather than for the specialized services of the Gaan Tata priests.

The priests adjusted to this unknown austerity by lowering their sights, for example, by demanding smaller fees. But-they also had recourse to measures that were profitable in the short run but detrimental when seen over a longer time period. The priests began reducing the number of their staff and they also paid less to those who stayed on. This new generation of priests proved to be less tractable than their well-paid predecessors had been. For example, it happened occasionally that a front bearer of the holy bundle refused to perform his duties. This put the high priest in a difficult position because front bearers were considered specialists. In 1962, there were only two of them and refusal to work occurred when one of them had left for a short period, with the high priest’s permission. More common but equally costly, were the repeated absences of periods of leave that a high priest could not refuse to grant. This created a vicious circle. Not being able to pay the priests fully meant fewer working days, this led to fewer patients and thus less income for the priests.

In order to secure higher incomes, the Gaan Tata priests had recourse to an even more hazardous strategy. As we have seen, the gravediggers were responsible for the inquest into the cause of death and for the verdict on the moral character of the deceased. In the 1960’s, the priests put pressure on the gravediggers to give as many verdicts of a sinner’s or witch’s death as they could, and to resist demands of the relatives that the deceased be recognized as a respectable citizen. In the cluster of villages around the main shrine at Dritabiki it was not particularly difficult to do this. One or more of the gravediggers at the death ritual would either be a Gaan Tata priest or a close associate as an elder at Dritabiki. It was often suspected that these notables acted as spies. Any uncertainty about the reputation of the deceased would be reported back to Dritabiki and this could then lead to a change of verdict. After the gravediggers had arrived at a judgement, the verdict had to be communicated to Dritabiki. For Dritabiki and neighbouring villages the priests of the oracle knew the proposed verdict even before the relatives were informed and could—and often did—alter it. In 1962, 21 out of 30 deceased were considered witches, 6 were condemned as ‘sinners’ and only three qualified as ‘respectable citizens’. In this way, 27 out of 30 legacies fell to the Dritabiki oracle (Van Wetering 1973: 56). For the whole decade from 1961 to 1970, the number of ‘respectable deaths’ was much higher: 34% against only 10% in 1962. The longer period, however, includes the years aftgr 1964, when the positions of high priest and tribal chief (gaanman) were not held by one man; considerable confusion and rivalry then disrupted the workings of the oracle. Yet, the percentage of legacies confiscated remained amazingly high.

A new regional cult?

During the last days of the Gaan Tata cult discontent had spread. Whereas during the early 1960’s it came to the surface in private conversations, malpractices were openly debated in the early 1970’s. In 1972, a prophet, Akalali, launched a new cult. It should not have come as a surprise that he took the institution of the God’s cargoes as his prime target: he pointed this out as the signal abuse of the Gaan Tata priests, the epitomy of corruption. With a few daring manoeuvres, and supported by a considerable following, Akalali put an end to the work of the Gaan Tata priests; he then invited the people of the Tapanahoni to rescue whatever they could from Santi Goon, the cult’s storage place in the bush. The history of this take-over, the reasons for its stunning success, need not detain us here (cf. Thoden van Velzen and Van Wetering 1975). What is pertinent to our theme is not the fact that the parochial cult of Gaan Tata was replaced,[35] but that it was supplanted by a parochial cult with regional ambitions and even some degree of regional success. What prompted the new religious leadership to reach out for the whole of Surinam’s interior, instead of resting on the laurels of its Tapanahoni victories? The reasons, I would submit, are essentially the same as those that caused the rise of the Gaan Tata cult: a massive injection of new resources and new sharp divisions within Bush Negro society which generate anxiety.

Two developments seem relevant. Around 1955, the government of Surinam launched an ambitious programme for the exploration of the interior: air strips were constructed; roads were laid out far into the rain forest to pry it open for lumbering companies; geologists embarked on a systematic survey of the hinterland; a hydro-electric dam was built on the Suriname River. With support of the World Health Organization, a campaign to eradicate malaria in Surinam’s interior was launched. Like the gold-diggers of old, the medical doctors of the anti-malaria campaign, the geologists, the engineers all were dependent on the Bush Negroes for transport. Top officials of the departments concerned employed the tribal chiefs as brokers. In this way, many jobs fell to the young relatives of the chief and the Gaan Tata priests at Dritabiki.

Then, in 1963, an even more important development set in; for the first time in Surinam’s history the Bush Negroes were given the right to vote in a national election. It soon appeared that the few seats Bush Negroes held in the Staten, Surinam’s parliament, were of strategic importance. Their representatives could virtually decide who would rule Surinam: the Hindostani, descendents of contract labourers brought over from India, or the Creoles. Skin colour, or a common cultural heritage, were not enough for the Bush Negroes to decide whom they would choose. More was needed, and it was delivered. New jobs were created; many an elder was promoted to village headman or village assistant, both pf these paid positions. In 1973, the one village of Dritabiki, which has less than 500 inhabitants, had a tribal chief and no fewer than 14 headmen and assistants to guard over its welfare. For many of the younger men at Dritabiki, jobs, had been furnished. On top of this, during the months preceding the elections, representatives from both main parties came to deliver boat loads full of food, liquor and corrugated iron at the landing stages of many Tapanahoni villages.

These goods were not evenly spread over the Bush Negro population. Much of it was pumped through old channels, the political middlemen at Dritabiki: the tribal chief and the priests of Gaan Tata. The flag of the main Creole political party was hoisted over Gaan Tata’s oracle. Obesity, cases of high blood pressure, began to take their toll among the well-to-do at Dritabiki. For the people of neighbouring villages, the privileged position of the tribal chief, priests and their relatives was a stone of offence, an open invitation for a take-over bid. Akalali came from a neighbouring village that had taken a leading part in opposition to Dritabiki before. After the take-over, Surinamese officials, quick to accommodate to the political realities, promoted the prophet to village headman. After the 1973 elections, when the Tapanahoni seat in parliament was lost by the Hindostani coalition, the new Creole government made Akalali a village headman of the first rank (Hoofdkapitein).

We should try to get the wider perspective in view: the new demand for Bush Negro river transport and the massive political patronage were a Surinam-wide phenomenon. Paradoxically, the opening up of the interior had started off its depopulation. The men left for the labour-recruiting centres, in particular to Paramaribo. As a result, a village such as Dritabiki had lost two-thirds of its population. A’change in attitude, a readiness to adjust to coastal ways of life, became notable (Thoden van Velzen 1966: 21). However, a new period of full employment, like that of the gold-rush days, was not to come. The majority of Bush Negro migrants had swelled the ranks of the unemployed proletariat. A new division arose among the Bush Negroes: the employed versus the unemployed and those who went from one part-time job to another. In many ways, the employed occupied the position of a labour aristocracy. This would even hold on a national level; in the 1960’s Djuka boatmen working for government services earned twice as much as the Javanese sugar-cane cutters of the coast. The privileged workers jealously guarded their position and were apprehensive about the possible consequences of this new economic rift. It should therefore not come as a surprise that the Akalali cult of the 1970’s was again an anti-witchcraft movement. Equally, in line with the earlier position of the Gaan Tata priests, the young employed men who came to the Tapanahoni for the New Year celebrations of 1974, had business to do with Akalali, sporting colourful woollen sweaters, they came in boats driven by outboard motors, a few even in chartered planes, to obtain a certificate that they themselves were not witches and also for protection against witches. Both wishes were satisfied. In the first days of 1974, Akalali screened approximately 200 persons for any signs of witchcraft: none of the afluent young labour migrants who deposited such substantial cash gifts in his coffers, was pointed out as a witch. From the 200, about 15 were set apart as witches, and most of these were old women; none of the male witches had a job, they were destitute people and often ill as well.

Akalali had to cater for the demands of a ‘national’ clientele, Bush Negroes from all tribes in Surinam who were anxious to prove that they were ‘clean’ and who were in want of emotional security and protection against their envious brothers. The traditional social building blocks of society, the matrilineages, were again breaking up. Many Bush Negroes in Paramaribo had shaken off the obligations imposed by these groups, but were still in need of protection and security; Akalali appeared to take care of this and thus the privileged flocked to the new shrines of the Tapanahoni in such great numbers. The great majority of these were Djuka, but invitations soon reached Akalali from other areas in Surinam. In 1973, he spent a number of weeks in Paramaribo’ where hundreds of migrants visited him. In 1974, he travelled to the Djuka of the Cottica area and to the Paramaka to conduct large-scale witch cleansing rites. Early in 1975, he was invited by some Matawai to cleanse the village of their tribal chief, a Christian bulwark, from witchcraft. In 1975, he performed anti-witchcraft rites in some Saramaka villages in the central part of the country. Although these travels outside his Tapanahoni stronghold were not consolidated by the setting up of local branches, the recurring demand for Akalali’s services from various areas made his cult more than a parochial one.

Summary

The first stage in the history of the Gaan Tata cult (1885–1925) can be directly linked with the alteration in the mode of production. The Djuka, as most other Bush Negroes, left lumber work for more profitable employment in river transport. This entailed a loss of functional utility for the corporate matrilineal kin group which had proved to be so instrumental in the organisation of work in the lumber camp. As the proximity and assistance of relatives no longer were necessary, kinship obligations became a burden. The new unit of production, the boat crew, consisted of only three men; there were no phases in the production process when larger numbers of workmen were required: smaller work units thus, but not isolated atoms. As the crews moved over vast areas in the river basin, their common interests were peaceful relations with other residents, easy movement among strangers and unstrained relations with the District Commissioner and his assistants. The Gaan Tata movement honoured these instrumental needs by stressing the values of peacefulness, honesty towards employers and rejection of any parochial ideology stressing family, lineage or even tribal interests.

Equally significant, the Gaan Tata priests showed themselves partial to the privileged socio-economic group of that time, the freight carriers. While the gap between wealthy and poor was widening, Gaan Tata’s priests saw to the emotional needs of the boatmen by affording them protection against witches. As the Gaan Tata priests controlled the only centres of information on a supra-village level, they could intervene on behalf of their wealthy clientele by upholding their reputation.

For the second phase (1925–1955), the argument gets much more complex. The period began with a steep decline in freight prices, from fl.25 per barrel for the Albina-Lawa trip in 1909 to fl.7 in 1930. More gradually, a diversification in the mode of production took place: many Bush Negroes returned to lumber work, others worked off and on as freight carriers, depending on the sluggish demand. The impoverishment of the freight carriers is the main independent variable for our explanations of developments during the second phase. The diversification,came to reinforce the first factor, but in order to gain a clear view of the main lines of cause and effect, I will ignore it.

The impoverishment of the boatmen had two consequences: (1) the gap between wealthy and poor narrowed as the entrepreneurs saw their income reduced to a fraction of what it had been before; there was now less to fear from the witches as they possessed fewer things people could be envious about. Hence the demand for an anti-witchcraft cult became less pressing. (2) As an immediate result of the dwindling fortunes of the entrepreneurs, the priests saw their income sharply reduced. The high cost of the death rites further squeezed the available resources. The massive dumping of property at the secret bush shrine of Santi Goon, was only the most spectacular way of taking resources out of circulation. The cycle of death rites with the many libations and the huge communal meals also tended to dissipate reserves.

The dependent variable then is the breaking up of the regional cult of Gaan Tata in a number of local congregations. There was less demand for an anti-witchcraft cult, less ‘push’ from the various areas. There was also less ‘pull’ as the centre on the Tapanahoni no longer had the power assets to steer taxes, tribute and clients towards its oracles. As we have seen, they had been forced to abdicate the right to collect and redistribute the God’s cargoes from other Djuka areas, meaning a lower level of income. Very few normal lines of contact between centre and periphery could be maintained. The Gaan Tata priests in the centre were painfully aware of the many deviations from what were considered the standard ritual proceedings in the Tapanahoni area. They did not condone these, but there was nothing they could do to redress the situation. The process of disintegration had gone too far; the economic basis for a healthy regional cult no longer existed.

The third phase (195 5-present) began with a new demand for the specialized services of Djuka boatmen. But full employment, as in the days of the gold rushes, was out of the question; a few hundred Bush Negroes could man all the boats of the various government services working in the interior. The labour demand of the lumbering and bauxite companies of Surinam could only absorb some of the Bush Negroes. The result was the emergence of an unemployed Bush Negro proletariat in Paramaribo; men who settled in the same neighbourhoods as their more fortunate kinsmen, waiting for a job.

The cleavage between wealthy and poor grew wider when the main national parties began vying for the favour of the Bush Negro electorate: new patronage jobs, new village headmen, boat loads full of gifts arrived, but the glut tended to concentrate at certain points in society.

The two basic conditions that gave rise to the Gaan Tata cult are also evident in the new cult of Akalali. The massive migration of Bush Negroes from their old tribal areas to Paramaribo blurred cultural differences: again Bush Negroes had to work and live among strangers. The ethos of the cult had to bridge cultural differences instead of stressing them. But it is partial to one segment in the new socio-economic division that emerged among Bush Negroes: just like the Gaan Tata cult, the Akalali cult supports the successful. The vitality of both cults derived from a clear choice in favour of the privileged.

References

abbreviations of the names of missionary journals

BHW: Berichten uit de Heiden-Wereld uitgegeven door het Zendingsgenootschap der Broedergemeente te Zeist
MBB: Missions-Blatt aus der Briidergemeine
MITT: Mitteilungen aus der Brudergemeine
NB: Nachrichten aus der Brudergemeine

© Bonno Thoden van Velzen 1977

5. A Continuing Dialogue: Prophets and Local Shrines among the Tonga of Zambia by Elizabeth Colson

Acknowledgements: This paper is based on field research among Plateau Tonga of Mazabuka District (now divided among Mazabuka, Monze and Choma Districts) and Gwembe or Valley Tonga of Gwembe District. I worked in Mazabuka District, August 1946—September 1947, June 1948— August 1950, and September 1968, and have paid brief visits to friends there over the years. My last visit was in August 1973. My field work among Gwembe Tonga was carried out July 1949, September 1956—September 1957, January 1960, September 1962—September 1963, July—August 1965, 3 days September 1968, July—December 1972, and June—August 1973. I have drawn on the field notes of Thayer Scudder, my colleague in a long-term study of Gwembe District. His notes date from September 1956— September 1957, September 1962—September 1963, 1967, 1970, May-October 1972, and June—September 1973. Over the years we have discussed the Gwembe data many times until it is now impossible to be sure which of us is responsible for orginating any particular interpretation. Paul Syanemba recorded a diary for us in 1972—73 on which I have drawn. Scudder and Richard Werbner commented on early drafts of this paper.

The research has been supported by various agencies: the Rhodes-Livingstone Institute (now the Institute for African Studies in the University of Zambia), the Joint Committee on Africa of the Social Science Research Council and the American Council of Learned Societies, National Science Foundation Grant No. GS—3295, the Small Grants Committee of the National Institute of Mental Health, the Kafue Basin Committee, the African Committee of the University of California, the Research Committee of the University of California (Berkeley), and the California Institute of Technology. The California Institute of Technology gave me the rare privilege of serving as a Fairchild Fellow during 1975—76 and so gave the leisure to return to old field notes and to write this paper. The map is at the end, page 136.

Cults and their publics

No cult which primarily functions to serve the particular interests of a territorial community on a regular basis can serve a general public unless it radically alters its practice and its constituency: a cult serving a wider constituency on an ad hoc non-exclusive basis may well serve the special purposes of local communities. The two can co-exist and reinforce each other. The cult with universal claims then serves as a further court of appeals when a local cult does not satisfy its adherents. An appeal to such higher authority is subject to public review or at the behest of local officials, which implies that delegations go for consultation with an oracle or other universalistic cult as messengers of a local community and representatives of its power structure. The oracle will be asked to confirm the continuities upon which the local community depends in regulating its life rather than requested to provide guidance in some radical departure from accepted practice. This was true in ancient Greece where each small city state had its own cult, associated with its fate as a human community, but delegations went to consult the Oracle of Delphi, or perhaps Dodona, which served the entire classical world for many centuries.

The same interplay between local cult and oracle or prophet cult appears today among the Tonga of Zambia. The two cults appeal to the same basic human desires for health and prosperity. They incorporate comparable beliefs about how humans, the natural world, and divinity are interrelated. They share a common vocabulary and much symbolism. By and large they deal in the same kinds of messages. Both reflect the nature of the land the Tonga occupy and their concepts of social order. Yet the two cults remain distinct because they are seen as serving different purposes: each local cult under its custodian is a symbol of the continuity of a particular community, while the oracles and their prophets provide reassurance and guidance in emergency or dilemma to a fluctuating public. Today the economy of central Africa is diversifying, even in rural areas. The prophet cult is responsive to the new interests and many prophets are redefining their role so that their primary mission is to serve individuals rather than communities. The local cult is less able to respond to such diversification. Many local communities no longer have a local shrine or rally to support community rituals linked to agriculture or any other common interest. But in some communities local shrines still exist and elders still gather around them to invoke the associated spirits and assure them of continued respect and support in return for their care of the common interest.

The Setting and the Cults

Plateau and Gwembe Tonga occupy much of Zambia’s Southern Province, while Gwembe Tonga also live beyond the Zambezi River and man-made Lake Kariba in northwestern Zimbabwe (Colson 1958, 1960, 1962, 1971; Scudder 1962, 1971). Their terrain varies from the undulating reasonably well-watered plains of the Zambian plateau, through the rugged hills of the Zambezi escarpment, and drops to what was once the floodplain of the Zambezi River and is now Lake Kariba, before rising again through the Zimbabwe escarpment. Altitude varies from between 4,500 and 1,300 feet. Rainfall is highly variable, both locally and from year to year. The highest average rainfall is recorded at the edge of the escarpments which receive something like 38 inches annually: this drops to 31 inches on the Zambian plateau and to 25 inches in the Zambezi floodplain. As one goes east or west from the escarpment edge, rainfall declines and also becomes more uncertain. In the Zambezi floodplain an annual variation of from less than 15 to more than 30 inches has been recorded. But this was before Lake Kariba altered the microclimate and probably raised rainfall about its shores (Scudder, 1962, p.9).

Language links the Tonga with other communities in Southern and Central Provinces. In the Zimbabwe escarpment, on the other hand, they meet with Shona-speakers. The archaeological record indicates that iron-age settlers arrived in the first millenium with .their crops and livestock (Fagan, 1967). The Tonga may be descended from them although they have also incorporated more recent comers including Shona-speaking migrants from Zimbabwe. Old trade routes which crossed their country once linked them to the Rozwi, Karanga and Korekore states of the Zimbabwe highlands and lower Zambezi.

Whatever the earlier political relationships of Tonga communities with these kingdoms or with the Lozi kingdom of the upper Zambezi, by late nineteenth century they lived in independent polities, most of these no larger than a single neighbourhood inhabited by from 400 to 1,000 people. Each small polity was usually under the aegis of a shrine or shrines (malende or kokalia), whose custodian, usually called a sikatongo (‘person of the old settlement place’), represented the matrilineal descent group of the founder of the shrine and the founder himself. Henceforth I shall refer to such a shrine as a local shrine although elsewhere I have sometimes termed them “earth shrines” because they relate people to the earth they occupy, or “rain shrines” because Tonga emphasize their importance in appeals for rain. I shall call the sikatongo associated with such a shrine a custodian, though again sometimes I have used the term “earth priest” in earlier publications. Frequently the shrine was conceived of as the embodiment of the dwelling of the first couple who by settling in the neighborhood had transformed it from a wilderness and introduced the human routines associated with the building of homesteads and the cultivation of the soil. The shades of the couple then were held to continue to watch over the community they had founded, and in this case the matrilineal descent group of the wife also had responsibilities in the rituals. Less commonly a local shrine was said to have been initiated by a prophet, but that at his or her death custodianship passed to a matrilineal kinsman or kinswoman rather than to another prophet. In either case the spirit associated with a local shrine usually was called basangu, used also for spirits which speak through a prophet and sometimes also for any spirit. Unlike other such spirits, however, those regarded as originating from first settlers were held to have power only in the specific locality known to them as living men and women. There they were said to “know the country” and to be effective in dealing with it. Otherwise they were merely mizimu, ancestral shades, with power over their own descendants and members of their descent groups.

Local shrine communities were small, reflecting the highly localized nature of the rainfall pattern. Those living only a few miles apart can have very different weather and regulate their planting, weeding, and harvesting by different schedules; and their harvests differ accordingly from year to year. Prophets, on the other hand, are not so confined. Some received delegations from communities as much as a hundred miles from their own location. Prophets were involved in the unbounded social network which linked all parts of Tonga-country and ultimately those in Tonga-country with other parts of Central Africa. Political boundaries provided no convenient grid for organizing social relationships and the spirits associated with prophets did not represent any regional alliances or association with a defined geographical space. People traded, visited, and once raided over many miles, but movement in general was from the Zambezi to the plateaux along an east-west axis since this associated those in different ecological zones, brought access to different resources, and gave the greatest possibility of supplementing local food shortages. By the end of the nineteenth century, men were also going out to work along the same routes which continued to channel the flow of traffic from the Zambezi to the plateaux. That all known territory was subject to some of the same forces was evident from the fact that in some years drought or good rains were general. But if rain fell anywhere it was in the escarpment edges where rose most of the rivers and streams flowing across both the plateau and the Gwembe Valley. Significantly enough, the most important prophets seem to have lived in the high rainfall zones of the escarpments, either in Zambia or Zimbabwe, though minor prophets were scattered everywhere and probably most neighbourhoods contained a prophet or two. The vast majority had purely neighbourhood reputations and therefore served the same population as the local shrine of the neighbourhood whatever their claims to a wider mission.

Shrine Custodians

Custodians had official duties associated with the annual cycle. When officiating in the ritual, they were addressed as basangu although they were not expected to be subject to possession. This distinguished them from prophets who spoke from basangu and sometimes as basangu.

Custodians were expected only to continue routines which had proved effective in dealing with the earth and its forces. They did not exercise political rule or represent their neighbourhoods to other communities. In only a few places was anything comparable to the office of chief in existence. More commonly someone known as ulanyika (“warden of the country“), who might also be a custodian had some relationship to hunting and might claim rights in animals killed in the neighbourhood (Scudder, 1962 (pp. 111–129, 196–197). The de facto political leaders were usually vigorous men who could mobilize a following to dominate a neighbourhood or a number of neighbourhoods, but they created no office and held no ritual role in the community. If no one succeeded in dominating, the political community was a republic of independent homestead heads who acknowledged the ritual guidance of the local shrine custodian and the discipline of knowing that they were all vulnerable from actions that angered the shrine spirits. The Tonga stress the independence of each adult—or at least each adult male—and the autonomy of individual homesteads rather than a hierarchy of authority.

The matrilineal descent groups which regulated marriage, inheritance and succession, and which also provided for self-help, were usually egalitarian in organization. They rarely numbered over forty adults and these were dispersed among a number of neighbourhoods and dependent upon those with whom they intermarried and associated. Each descent group made offerings to its own ancestral shades, but this cult of the shades was distinct from the local shrine cult.

Prophets and the Greater Community

Prophets spoke as the vehicles of basangu and could be addressed as basangu or baami ba imvula (“lords of the rain”) in identification with their possessing spirits. Henceforth I shall use basangu specifically for the spirits that “visit” or “enter” prophets. These, unlike the local shrine spirits, were identified as spirits of the wild, or as shades of long dead prophets, or as foreign spirits — thus they were identified with no known social community and could affect any. Indeed, they were spoken of as “visitors,” emphasizing their transitory connection with their prophets and their prophets’ locale. Some prophets say that their basangu manifests itself as a snake which crawls into the body through the mouth and they forbid the killing of a snake within their homestead. Others visualize their basangu as tiny lions or buffaloes. To the best of my knowledge Gwembe prophets do not claim possession by spirits associated with local shrines though they may contact such spirits through the intermediation of their own basangu and so learn their grievances and how to pacify their anger: I am less certain that this is true on the plateau.

Especially in Gwembe District, people identify basangu as a class, or a particular basangu, with the mhondoro spirits of the Shona (Garbett 1966: Gelfand 1956; 11–34; Daneel 1970; Ranger 1966). Thus the spirit of Chibwe Handabo, claimed by various prophets on the Zambian Pleateau and in Gwembe Valley, is said to come from Shona country and to be the same as mundolo, the Tonga pronunciation of mhondoro. But the Tonga do not seem to personalize the basangu or to be interested in their histories. The basangu have names, sometimes places of origin, but that is about all. For that matter, most Tonga do not seem to be concerned with which basangu is present and many do not know the names even of basangu associated with prophets in their neighbourhood. The various known basangu are not ranked in terms of spiritual power or linked into a system by using a genealogy or other method of stating relationship. Here the Tonga differ markedly from the Shona who try to draw the various oracle spirits into a common system reflecting the hierarchical relationships of Shona society. The Tonga think of their basangu as free to behave in the same independent egalitarian fashion in which they themselves behave.

Anyone may become a prophet. It is assumed that basangu choose their vehicles at their own pleasure from any descent group or locality, and there is no reason why a basangu should continue to visit a locality where it has had a vehicle. Basangu enter, men, women, and children, but children become active prophets only on maturity. Significantly, prophets need not be married; it is the shrine custodian who represents ordered society who must have a spouse. In recent years more women than men have become prophets, but there is no good evidence on how long this has been true and some reason to think the situation may be changing. Any number of prophets may be active at the same time; indeed, several basangu may announce themselves within the same neighbourhood in the same brief time span. Various prophets also claim possession by the same basangu. The names of a few basangu are widely known,* and prophets all over Tonga country claim to be their vehicle. Such claimants work independently of each other and do not form a cultic corporation to advance the influence of their common basangu. But basangu are thought to be in contact with one another. They are thought to spend much time in the bush, in hollow trees, standing pools, springs, rock outcrops; but they also visit each other. They are said to withhold the rain or send epidemics or other troubles to alert communities to their demands or as a punishment on all those in the countryside for an infringement of the respect due their prophets and their chosen sites. But they are not all powerful, and in the last analysis it is God who sends the rains and orders the world.

Basangu cannot be invoked as agents of individual ends. They are not appealed to for vengeance against some other community-or faction within a community or personal enemy. They are thus neutral in human disputes. Presumably they symbolize the general relationships between the forces of nature and human society rather than the division of that society into particular communities, whether these be based on neighbourhoods or descent groups.

If prophets have no formal organization, they ought to be in harmony and all basangu are assumed to have common interests. Quarrels between their vehicles anger the basangu, and drought may be attributed to antagonism between prophets. In fact, however, prophets vie with each other for recognition and their influence fluctuates enormously even within a short period of time. Each neighbourhood, and indeed each homestead, appears to have the option of choosing which prophet it will patronize on any particular occasion, although I have heard a prophet chide consultants for coming in small groups rather than in full strength. Only prophets who are also diviners receive individual clients, i.e., those asking about a concern such as illness specific to an individual or household. Basangu concern themselves only about public matters.

Most prophets have local reputations and are consulted by delegations only from their own neighbourhood or at most from within their immediate vicinity, a radius of no more than five or six miles. The reality of their claim to possession is also treated with scepticism. The few famous prophets who have inspired great respect draw occasional delegations from many parts of the country, but are probably consulted regularly only by their own neighbourhood or by other neighbourhoods in their immediate vicinity. When an appeal is made to such a prophet for rain, the rain is expected to accompany the delegation back to its home territory. These prophets also warn of trouble coming unless people throughout the countryside heed their message, receive and wear a particular charm, or comply with some new or half-forgotten ruling. They are therefore able to introduce innovations in ritual, including local ritual. In southern Gwembe in the 195O’s, Thursday was observed as a rest day on the orders of such a prophet: Sunday was the chosen day of another.

In the late nineteenth century the prophet now called Monze, the name of his basangu, received delegations from many Tonga, Ila, and Sala communities, some coming from Zambezi neighbourhoods as much as 100 miles away. Monze lived on the Zambian plateau not far from the escarpment edge in the high rainfall zone (Livingstone 1857: 554–555; Syaamusonde and Shilling 1947: 47–56). His successor was recognised as a major chief during the colonial period. Rain rituals were carried out at the shrine associated with the two men at least as late as 1972 and the shrine served more than a single local community. The basangu Monze is still claimed as a possessing spirit by contemporary • prophets. In the midtwentieth century the most powerful prophet among Zambian Tonga was Chibwe, who lived not far from where Livingstone first met Monze. Her basangu was Chibwe Handabo. She was occasionally consulted by delegations from Mazabuka, Gwembe and Namwala districts. Lesser prophets of midcentury usually visited Chibwe soon after first announcing themselves and cited her recognition of them as fellow prophets as proof of their own calling. Two I talked with, one on the plateau in the 1940’s and one in Gwembe in the 1970’s, claimed that Chibwe had said the people who lived near them should come to them rather than make the long journey to her homestead. After her death in the 1960’s, her shrine appears to have become dormant, but new prophets are still claiming possession by Chibwe Handabo.

In the 1950’s some Gwembe neighbourhoods sent delegations to the shrine of Monze even though it no longer had a resident prophet, others sent to the prophet Chibwe, and still others sent to mhondoro prophets in the Zimbabwe escarpment. Of these last, the prophet near Gokwe seemed the most popular. Adjacent communities might send their delegations in different directions. The delegations were sent only in a year of drought and after consultations with prophets of the vicinity had brought no rain. »Some communities had no memory of having consulted prophets outside their immediate ken. Scudder and I have never been able to determine what influenced the choice, though we note that if Gwembe neighbourhoods consult distant prophets, then these live in the’highlands where the rainfall is greatest and where the storm clouds can be seen to gather. We have no record of any Plateau Tonga delegations visiting Gwembe or Shona highlands in search of rain or other assistance or of Zimbabwe highland delegations coming to Zambezi prophets. The choice of prophets certainly is not affected by considerations of ethnic identification, and basangu are not thought to have ethnic or territorial loyalties. Thus people are free to consult whichever prophets they regard as most likely to be effective and other than this probably follow their own convenience and familiarity with the intervening country.

Freedom of choice is also encouraged by the fact that all prophets seem to use a modicum of symbolism common to central Africa. They wear wristlets of black or dark blue beads, colours associated with rain clouds. Those who approach them for rain take black or blue beads. Those who wish an end to continuous rain take white beads because white is associated with clouds which carry no rain. Red beads or anything red in colour is taboo during consultation because red symbolizes blood. An appropriate gift is a hoe blade, symbol of cultivation (and of femaleness). Spears and axes, symbols of maleness and the hunt and war, do not appear to be used as gifts to either prophets or local shrines, though we have one instance of a novice prophet receiving a hoe and a spear from an established prophet. Often petitioners take seed to a prophet and receive in turn blessed seed to mix with their own seed for planting. If a prophet announces that the basangu want an offering carried out at a local shrine, this usually involves slaughter of a black chicken, goat, or ox—any trace of white being forbidden since this would bring lightning with the rain. The animal should be held so that it cannot struggle, since the struggle symbolizes wind which blows grain about. The meat should be roasted without salt. Informants have not been able, or willing, to make the symbolism here involved explicit, but salt is “hot” and the spirits are asked to be cool “as water is cool.” Delegations appealing to a prophet proceed in much the same fashion throughout Tonga contry, adopting the same posture (sitting with legs straight before them on the ground), clapping with the same rhythm, and using a common repertoire of rain songs. This is also the way they present themselves when appealing for rain at a local shrine or in the homestead of a shrine custodian.

Neither Scudder nor I have beem present when a Tonga delegation has approached a Shona mhondoro, nor have we visited a mhondoro, but we are told that at Gokwe and elsewhere they are coached in a more formal ceremonial and must deal with subordinates instead of going directly to the prophet as they do in Tonga country. No Tonga prophet of the old styles seems to have an entourage or maintain an establishment. Both Livingstone and Selous (1893: 211–212) commented on the lack of such an entourage for Monze, then the most famous Tonga prophet. Today at most one helper, usually a spouse or resident kinsman or kinswoman, assists in seances and interprets pronouncements. A prophet’s homestead is usually no more elaborate than any other and sometimes is poorer than most. It is marked, if it is marked, only by the presence of a small hut shrine standing near the dwelling of the prophet. The prophet except during consultation is addressed and treated as an ordinary mortal although each prophet has special rules with respect to sleeping, food, or other behavior which he or she observes at all times. It is only in possession, or when a delegation requests a prophet to seek possession, that the prophet is treated with respect and regarded as having authority.

The Diversification of Society

Before the mid twentieth century, Tonga prophets were primarily consulted about agricultural hazards, pestilence, and cattle plagues. Until the last decade of the nineteenth century the Tonga were subsistence farmers, though hunting, fishing, and gathering provided a good deal of the diet and much of the zest of living. Thereafter, the need for a cash income drew them.into other activities. They turned first to labour migration.The Plateau Tonga were able to shift to cash crop maize farming in the 1920s and 1930s and raise cattle for the market,but Gwembe Tonga continued as labour migrants until the late 1950s and even in the 1970s many are away at work intermittently. In recent years the economy of both plateau and valley has been diversified and both have shared in the advances made by Zambia since independence in 1964. Most villages are now accessible by road. Some Tonga have cars and tractors. People move easily from rural areas to the towns and cities to visit kin, seek work, shop, or consult government agencies. Townspeople in turn visit rural kin and invest savings in rural areas. Many children are in secondary school and look forward to professional jobs and to associating with fellow professionals. They take their beliefs and values in part from alien sources (Colson and Scudder 1976).

Work opportunities, and increasingly the policy of the Zambian Government, bring immigrants from other parts of Zambia and central Africa to live in Tonga neighbourhoods. Administrative and technical personnel associate with educated Tonga men and women who are similarly employed or occupied in large-scale farming. Others, who come as labourers, fishermen, or small traders, associate with the mass of villagers. If they stay, they learn ciTonga, as did earlier immigrants, most of whom ultimately merged into the Tonga population where diversity of origin is quickly forgotton (Colson 1966).

Present day Southern Province communities, even rural ones, are heterogeneous. Their populations differ in education, experience, occupation, and in the demands they make on life. Many, including the most prosperous farmers, business men and civil servants, dislike the system of matrilineal inheritance and wish to emphasize the nuclear family rather than the matrilineal descent group. Many think of themselves primarily as citizens of Zambia rather than as citizens of a neighbourhood. Those regularly employed are less concerned with rain and local harvests. Missions have been active since early in the twentieth century and by the 1970’s churches had been built in many neighbourhoods. Any who went to school before the mid 1960’s spent some years under Christian teaching. Some Tonga are Christians. Some are agnostics. Some fear witches and medicines but deny that ancestral shades or such spirits as basangu can affect the living. Still others firmly believe in what they regard as the tested wisdom left to them by their ancestors. In communities where the majority support these beliefs, as is true particularly in many neighbourhoods in Gwembe, even the agnostic may find it expedient to make offerings and take part in other rituals. Conformity then encourages belief.

In our study of Gwembe District, Scudder and I have observed a progressive change among many who as young men and school boys in 1956 proclaimed themselves opponents of the ancestral cult and local shrines and held that prophets and diviners only cheated the people. They have become increasingly involved in all three cults as the years pass and they have become the responsible householders of the 1960’s and 1970’s with children, stock, and crops at risk.

Everywhere the cults have been affected by the increasing mobility and diversified interests of the population. In the 1970’s, the various cults received little official support. Officials and other rural elite do not seem to take part in offerings to the shades or appeals to basangu or rituals at local shrines unless they are chiefs who belong to custodial lineages. The latter continue to sponsor rituals associated with their local shrine since it is the link with the past which, forms the nexus of their continuing position and the reason for their recognition by government. In general, however, we have been impressed by the tenacity with which both local shrines and the prophet cult maintain themselves.

Continuity and Change in the Cults

On the plateau, people were shifting cultivators. Until land shortage finally stabilized the fields, homesteads moved gradually across the land. Local shrine communities seemed tenuously linked to particular localities and the social units risked dispersal and the outmoding of their local cults as homesteads entered into new combinations or responded to different expectations of rain. The flexibility of the cult of the basangu and the prophets’ right to make pronouncements calling for the institution of new local rituals seemed to cater to this situation. An overlap in terminology, symbolism, and ritual behaviour between the local shrine cult and the prophet cult meant that the line between them was easily blurred.Theoretically, these circumstances should make it easy for people to turn to prophets who moved with them or were encountered in the shift to provide them with a new shrine and a new line of shrine custodians to serve the emerging community. Originally I thought this must have happened frequently in the past and I attributed the short histories and abbreviated genealogies of Plateau Tonga shrines and shrine custodians to the brevity of life of shrines associated with communities of shifting cultivators (Colson 1948). Longer experience now makes me doubt this interpretation since it does not take into account the general disinterest in the past which has been characteristic of so many Tonga. Among egalitarian people such as the Tonga of the nineteenth and early twentieth century, history is unlikely to be a major preoccupation; genealogies will be truncated; time usually is telescoped into a span that covers perhaps two or three generations beyond senior people. This span is enough to root institutions in the past and justify them to the future.

On the Zambezi River, Gwembe Tonga lived in communities stabilized by access to alluvial soils and good water. In the mid 1950’s they were no more able than Plateau Tonga in the late 1940’s of providing lengthy histories for their local shrines or lengthy genealogies covering several generations of shrine custodians. Like the shifting cultivators of the plateau, they began the histories towards the end of the nineteenth century, in the time of their grandparents and great grandparents. Some, indeed, said that they themselves remembered the first shrine custodian though they placed the founding of the local shrine itself before their memory began. They also spoke only of recent prophets. None of those they mentioned seemed to predate the prophet Monze who was visited by Livingstone in 1854 and by Selous in 1888.

To my knowledge no archaeological work has been done at any shrine sites and so there is no independent check on age. However, Tonga informants say that local shrines may be abandoned if they are perceived to be ineffective or if their communities die out or disperse, though they also say some member of the custodial descent group ought to stay in the old locality. If the spirits want the shrines reestablished, they will let their wishes be known soon enough. Some neighbourhoods have two or more shrines, each under a custodian of a different descent group. The shrines are then said to serve different functions or one is said to have been largely superseded when a basangu speaking through a prophet directed people to institute a new shrine.

Nevertheless, increasing data, mostly from Gwembe District, indicate that local shrines are remarkably viable and that prophets rarely institute shrines which outlast their own lifetime or effective period as prophet. In the nineteen years we have followed events in Gwembe District, we have heard of no new local shrines. I have heard of none being initiated among Plateau Tonga since 1946. In these same years prophets have proliferated and with them their personal shrines. The more important cater for more than a local clientele and are multifunctional. If they are associated with rain, as some are, they are also associated with curing. Their owners deal in medicines for success in business and other enterprises and provide protection against witches and other dangers. Their very eclecticism and appeal to wide-ranging interests of a mobile population stand in the way of their coming to represent the continuities of small agricultural communities. But even those prophets who have stayed closer to the traditional role of basangu prophets have not initiated a local shrine cult or inserted themselves or their successors into the office of sikatongo associated with such a shrine.

This failure is particularly marked given the fact that the period of our observation coincides with the massive resettlement of Gwembe Tonga which wiped out the familiar territory of most Gwembe neighbourhoods and any claim to legitimacy based on first settlement. %

The Gwembe Resettlement: A Test Case

In 1958 some 60,000 Gwembe Tonga were forced to move when 2,000 square miles was given to Lake Kariba. In preparation for the filling of the lake and the establishment of fishing grounds, their old terrain was cleared. Bulldozers which smashed through villages also razed local shrines and their associated trees. People said they were being thrown away into the bush where they would become like animals since neither their local shrine spirits nor their shrine custodians knew the controlling forces. In fact they were moved either to uninhabited bush with which no one had ritual associations or they were given part of the territory of settled neighbourhoods which had their own shrines and custodians. The majority of those within Zambia were resettled on higher land above the lake shore along the tributary rivers whose deltas they had cutlivated. Several thousand were shifted to the plateau. Another six thousand from Chipepo Chieftaincy in central Gwembe were moved one hundred miles downstream to Lusitu, below the dam site. Here they were settled in the territory of Chief Sigongo, many of whose people were descendents of Korekore immigrants from Rhodesia or of Chikunda from further down the Zambezi (Lancaster 1974).

Immediately following resettlement it appeared only too evident that shrine custodians of resettled communities had lost their office along with their shrines. Immigrants either joined the established community rituals of their new neighbours, or they tried to cope without ritual protection other than that provided by their own ancestral shades. Or they invested in protective medicine. In Lusitu the incoming settlers were told both by Chief Sigongo and by the custodian of the major shrine in Lusitu that they were strangers and must not build either local shrines or spirit gate shrines (cilyango). The latter are associated with descent group rituals. They were warned that spirits already in the area would be angry at a challenge to their authority. They gave more credence to this warning when many died during their first years in Lusitu. They attributed the death to ingozi, a medicine said to have been left in the ground by ancient inhabitants to destroy those who settled where they themselves had died. People also said the land did not know them or the spirits associated with their communities and it was pointless to build local shrines.

Elsewhere neither established custodians nor political officials seem to have actually forbidden immigrants to build their shrines, but the immigrants usually did not build. We heard of a few instances where they did. In one case, in 1961, the shrine was struck by lightning. This confirmed the belief that local shrine spirits already entrenched in the area claimed exclusive jurisdiction, and even those who had moved into previously uninhabited bush did not try to build.

The prophets moved into the vacuum left by the loss of local shrines and the custodial office. Prophets active before resettlement indeed at first denied being subject to possession after the move saying their basangu had abandoned them. This may reflect their dependence on local cues in predicting rainfall, but we think it also indicates a temporary loss of faith in basangu which had shown themselves powerless in preventing the loss of the homeland. We have no evidence that any Gwembe neighbourhood sent out a delegation to any prophet in the escarpment during these years of stress though the escarpment remained unchanged. Nevertheless, within two years of resettlement, prophets again became active. First new prophets announced themselves. These usually claimed possession by some new kind of spirit or by some widely known basangu or at least by a basangu previously known in their new locale. Sometimes the new prophet emerged under tutelage of a prophet already working in the region. In Lusitu, the first prophets became active early in 1960. They called themselves mangelo (‘angels’) rather than basangu and seem to have been influenced by contact with one of the separatist sects of Zimbabwe or Zambia since they adpoted a white gown, used invocations including the name of Jesus, and might display the bible during consultations. Moreover they could be consulted at any time of the year whereas basangu prophets usually are active only during the season associated with the rains crop. The new prophets made their major impact in dealing with the immigrants’ fear of the strange land which they believed was hostile to them (Scudder 1968). Not all Tonga, however, were prepared to give respect to mangelo. Rather they equated mangelo with the spirits associated with the possession dances which sweep central Africa (Colson 1969). By 1965, some of the old basangu prophets resettled in Lusitu were again intimating that they were subject to possession and others were claiming possession by basangu. People were coming to clap before them for rain. By 1972, several had built shrines in their own door-yards and even Chief Sigongo’s people had begun to clap before them and ask for seed for use in planting, protection from insects, and adequate rain.

In central Gwembe, where most Chipepo people were resettled, new prophets began to make pronouncements in late 1960, though here they called themselves basangu and concerned themselves primarily with rain rather than illness. When mangelo possession arrived in 1963, it was treated as a possession dance or equated with doctoring. But in central Gwembe, people were resettled in their old hinterland and had no such fear of their new locale as had the Lusitu settlers. By 1962 some who had been prophets before resettlement again claimed the presence of their basangu and were being consulted. Our records for southern Gwembe, in Mwemba chieftaincy, are more ambiguous, but we think the first new prophets began to function in late 1960 or early 1961. Even so many of those resettled preferred to consult prophets they found already established in the hills.

For some years, prophets made different demands from those we heard in 1956 when we attended consultations with prophets in Chipepo. Then they typically attributed drought to a failure of shrine custodians to follow rules left by shrine founders, or to people ignoring the custodian’s precedence in initiating agricultural routines, or to encroachment on territory associated with a shrine. To appease the angry spirits and bring rain, people were told to repair the shrine and carry out the appropriate rituals in conjunction with the custodian. Prophets active in the early 1960’s in resettled communities attributed drought to sorcery or to a lack of respect for their basangu. We have instances of prophets sending messengers to call people from surrounding neighbourhoods to greet a visitor (i.e., a basangu) and receive blessed seed for planting. Thereby, of course, they gave assurance from the spirits that the soils could produce and so alleviated one of the major fears of the resettled. In any event, the prophets ignored the former custodians and made no suggestion that their behavior or the absence of their shrines affected the general welfare. Former custodians continued to make offerings to ancestral shades, who had once also been local shrine spirits, but they did so only in kinship offerings, either in the house doorway or at a spirit gate, the sites associated with ancestral shades.

In 1962 and 1965, resettled custodians and members of their descent groups sometimes said they expected to rebuild the shrines and claimed still to be guardians of their communities. Others laughed at this. The spirits of the shrines were sometimes said to have remained behind in flooded country, although ancestral shades and the basangu of the prophets were regarded as able to follow people in their migrations. Even the death of a former custodian in 1963 was attributed to witchcraft and not to his neglect of ritual duty. His grave site did not become a new local shrine. By 1965 divination on the death of members of custodial descent groups had begun to point to the anger of shrine spirits at being given no home, but the general public regarded such divinations with scepticism. When we asked if by living and dying in the resettlement area, a former shrine custodian might re-establish katongo (‘right of old settlement’), most informants maintained that katongo came only from long ago and could not be established now in this generation. Or, they said, if it could be established, it was probable that it would be attached to the descent group of sofne prophet whose grave or homestead site might become the focus of a new local shrine and whose successor would become sikatongo, shrine custodian. There were those, however, who said that one good thing given them by resettlement was the demise of the old agricultural rituals and the freeing of each cultivator to proceed at his or her own pace (Colson 1971: 226–232).

We predicted that if local shrines were reestablished this would be through the prophets, who certainly saw themselves as free to innovate. So far this has not happened. Since 1962 various prophets have died and been buried in the new localities and they have begun to merge with the ancient dead. Their own shades are invoked in the ancestral cults of their descent groups and their children, but their shrines have disintegrated and their basangu said to have gone to other vehicles or be tarrying for a while in the bush. But local shrines are being reestablished in a few neighbourhoods and these are the shrines known at the Zambezi. Those custodians who are still alive are being held to acocunt again and where custodians have died, successors are being recognized.

Our best information comes from CHEZIA neighbourhood, in Chipepo Chieftaincy in central Gwembe. This is one of the four Gwembe neighbourhoods we have been following intensively since 1956. It then had two local shrines, each under a custodian of a different descent group. Three villages were resettled in their old hinterland. Those who built close to homesteads in KOMA neighbourhood began to associate themselves with the KOMA shrine, while the rest who built homesteads and cleared fields in an area not known to have ever been previously cultivated no longer associated themselves with a shrine. One of the two former local shrine custodians stayed in the resettled community, still known as CHEZIA, but was not active as a custodian from 1958 until about 1971 when he built a spirit-gate in anticipation of a sign that he should build the local shrine associated with his line. The other custodian left CHEZIA and then died. His shade was inherited by a member of his descent group. In 1966 the inheritor responded to various divinations on family and lineage illnesses by building a shrine. Apparently resident prophets had not called for any such building. The inheritor dies in turn and again a successor was chosen. The shrine was rebuilt in 1970, this time with various members of the CHEZIA community contributing their labour in the old fashion. In 1972–73 the rains failed. Members of the community first consulted two women prophets resident in the neighbourhood. One had practised at the Zambezi but not for some years after the move; the other was first possessed in 1960. They reported that the basangu and the local shrine spirits wished the community to join in rituals, some of which had last been performed in 1957 before resettlement. In the following days this was done, the two custodians playing their appropriate roles. The ritual, as recorded in the diary kept for us by a member of the CHEZIA community who had been only a boy in 1957, is very comparable to what we recorded in 1957 although over fifteen years had elapsed and there had been drastic changes in the personnel carrying out the rites. Belief in the efficacy of shrines and ritual was reinforced when heavy rains fell. But some weeks later when again the rains had stopped, the CHEZIA people consulted a third prophet living in the neighbourhood. He had been possessed sometime after 1965. Most people considered him subject to mangelo or some other spirit associated with a possession dance and denied that he was a basangu prophet. Now he extorted a general acknowledgement that he was indeed a basangu prophet, though he faced the delegation with a bible open upon his knees. He demanded and received a hoe of ancient manufacture, the gift appropriate to a basangu. He then announced that the basangu was angry because one of the custodians had disrespectfully called it a mangelo and the same custodian had improperly performed his share of the ritual. The prophet ordered the custodian to make recompense both to his own spirit-gate and to the local shrine of the other custodian.

I In Lusitu, at least one local shrine was re-established under its old custodian prior to 1971–72. But in 1973 most Lusitu informants said that they still depended either on the local shrines of Sigongo people or on their own basangu prophets. In southern Gwembe, people of the Mwemba neighbourhood we are following also continue to rely upon prophets. The grave of a basangu prophet who died in 1971 has not become a shrine. Instead people consult a woman of another descent group who first claimed possession in 1971 and confirmed her status by a visit to an established prophet in the escarpment in 1972.

Prophets and Custodians

The nature of the prophet’s role and the political realities of the neighbourhood life work against the usurping of the custodianship by the prophet.

Prophets are affected by the general egalitarian character of Tonga social organization. If they had a corps of helpers there would be those with a vested interest in the perpetuation of a prophet’s own cult after the prophet’s death, but they do not build a bureaucratic order that could institutionalize their charisma. The basangu, for which the prophet is only a vehicle, is thought to be free to move t.o new vehicles unassociated with the old. It is only loosely associated with the shrine built for it by the prophet even during the prophet’s life. The shrine is but a temporary abode in which the spirit is only occasionally resident. Local shrines, and their immediate environs, are always in some sense sacred and should not be disturbed. The prophet’s shrine, in contrast, stands in the dooryard where children play about it. It does not inspire respect in itself. The very multiplicity of prophets also works against any one acquiring a following with exclusive allegiance.

The demise of the prophet then usually means the demise of the associated shrine and the dispersal of such clientele as the prophet has. No surviving staff exists with an interest in reminding people of the dead prophet’s power, and people are unlikely to associate any new difficulties they face with neglect of the prophet’s grave or shrine given that they never had any obligation to respect one prophet over another. They would first consult a living prophet. Even one from the same lineage as the dead prophet must speak as the vehicle of a basangu to receive a hearing and this means stressing his or her own relationship with the spirits. There is no reason to invoke the now dead prophet.

Pronouncements almost invariably involve some failure in the relationship between humans and divinity, but those who consult prophets evaluate the message transmitted to them in terms of probabilities. If a message is too alien to their experience or calls for something outside their power to comply, they are likely to reject it. There is a sameness about the pronouncements of prophets, perhaps because the questioning directed to them by representatives of the consulting community more or less dictate the acceptable answers.

Prophets who have a largely local clientele have an incentive to attribute misfortune to behaviour about which local people are uneasy but need to direct attention from possible delinquencies in their own behaviour as basangu vehicles. In Gwembe failure to rebuild the old shrines must have been a source of concern since people accepted this diagnosis of current troubles once the resettlement had receded into the past and was no longer the focus of all grievances. Continuing divinations put pressure on the prophets to revise their messages to conform to the growing consensus. When they did so, they also assured people of the continuing viability of their social unit. The explanation, however, also absolved the prophets of any responsibility as prophets for drought or other trouble—it was the community as a community which had done wrong and its own representative, the custodian, who was most flagrantly at fault. In some instances, personal rivalries were also in issue. Tension does exist between resident prophets and shrine custodians who base their claims to office on different sources of legitimacy. A prophet can humiliate a custodian but only by accusing him of negligence in office and this means giving him legitimacy. To invoke other explanations of community distress is to miss a golden opportunity to pay off old scores. The community is relieved to accept the explanation because it can put pressure on the custodian and check compliance by its own knowledge of appropriate routine.

Distant prophets who are not involved in local rivalries still look to the community order in framing their messages, since they can assume the organisation of a neighbourhood. It should be remembered that prophets speak to communities and the only community with the power to exact atonement from a culprit or general cooperation in ritual or observance of taboo is the neighbourhood community. A delegation also wants to receive an answer to its problem which lies within the area of its own competence. Even the most influential prophets cannot improvise the machinery which would coordinate the populations of larger regions for regular observance of complex intermeshed ceremonial. The more universal the message, the more likely it is that people will be asked to observe some single easily performed ritual, such as the wearing of a charm or the observance of a day of rest. Formerly, political divisions did not exist above the neighbourhood to provide a framework of authority to which prophets could attach themselves and so compliance with any message was left to general consensus or the power structure of a neighbourhood. The chieftaincies and districts instituted by the colonial government extended political boundaries, but few chiefs were able to make good their claim to ritual primacy within their own districts. People for the most part regarded their chiefs as administrative officers of government rather than their own hereditary representatives. Even the idea of a social unit of all Tonga is a recent creation and is still likely to be invoked principally in the national political arena, though the continued importance of the shrine of Monze may have political overtones of which I am aware.

At the present time the most ambitious prophets largely ignore the underpinning of community and treat more commonly with individual than with communal affairs. They continue to claim association with universal powers and hope to attract clients from anywhere in Zambia. They also receive clients throughout the year which means that their role is filled more than intermittently and periodically. They expect to receive payment in cash or livestock in return for service and so receive far more than the customary offerings of grain, beads, cloth and hoes which are appropriate for basangu. Most claim possession by a number of different kinds of spirits, of which basangu is only one. Some claim to be vehicles of God or angels, forces seen as more in keeping with the expanding universe in which people now live. They build shrines where appeal may be made for rain, but they see themselves as primarily healers who treat spiritual and physical ailments or as providers of protection against the dangers of contemporary life. They appeal to the diversified population of contemporary Zambia which consults them about the problems encountered as people move between town and country, shift their occupations, encounter new associates, cope with new uncertainties. Significantly, the most powerful set up their headquarters on the railway and bus routes rather than in the high rainfall zone of the escarpment edge, and their clients come from both cities and countryside.

Concern for the affairs of the land is left largely to lesser prophets who serve communities of which they themselves are members. Though they claim a power that stems from beyond the community, and is independent of it, they are subject to the community controls. They link their messages to the local order. That needs at best only one or two shrines to cater to its sense of harmony with the land it occupies. These prophets are therefore among the principal upholders of community, the land shrines, and the existing custodianships.

References

© Elizabeth Colson 1977

6. Regional and Non-regional Cults of Affliction in Western Zambia[36]

by Wim M.J. Van Binsbergen

1. Introduction

A generic relation holds between two types of cult of affliction in Zambia’s Western Province (formerly Barotseland). Cults of one type, the non-regional, form a substratum out of which cults of the other type, the regional, may spring forth under certain conditions, and into which they submerge again under different conditions. To understand this process and the conditions that govern it, I shall first describe the non-regional cults of affliction and then define the general characteristics by which the others set themselves apart from the substratum. Finally, I shall compare in some detail the development of two regional cults in order to make clear the importance of two series of variables: each distinct cult’s characteristics of idiom and internal organizational structure, and the structural characteristics of the geographical area which the cult transforms into a cultic region.

Several working definitions are useful in order to establish the contrast, briefly, between the types of cult within the general class of cults of affliction. The first type covers a few regional cults which have three main characteristics. Each of these regional cults has a specific idiom of its own. This idiom is pursued by a number of local congregations spread over an area of thousands of square kilometres. Thirdly, and most importantly, an interlocal formal organization binds these dispersed congregations through the interactions of the cult’s officials. In this way the geographical area over which the cult spreads is transformed into a region. What structures a cult’s region is thus the processes of interlocal communication, interaction and distribution which the cult gives rise to. By contrast, the second type covers cults which I shall call non-regional. A cult of this type, too, has a specific cult idiom which is pursued by a number of congregations. However, although it may have spread over a vast area, it has not yet transformed the area into a region of its own, or it has ceased to do so and a former region has become merely a non-regional area.

As for the general class of cults of affliction, in terms of cult idiom individual affliction invariably stands out as a major concern. This is central in Turner’s classic studies (1957; 1962; 1968) and in the earlier works (Kuntz 1932; White 1949; Gluckman 1943, 1951; McCulloch 1951: 72f; Turner 1952) which he developed. Turner coined the phrase, ‘cult of affliction’, to denote a cult (religious subsystem) characterised by two elements: (a) the cultural interpretation of misfortune (bodily disorders, bad luck) in terms of exceptionally strong domination by a specific non-human agent; (b) the attempt to remove the misfortune by having the afflicted join the cult venerating that specific agent. The major ritual forms of this class of cults consist of divinatory ritual in order to identify the agent, and initiation ritual through which the agent’s domination of the afflicted is Emphatically recognised before an audience. In the standard local interpretation, the invisible agent inflicts misfortune as a manifest sign of his hitherto hidden relationship with the afflicted. The purpose of the ritual is to acknowledge the agent’s presence and to pay him formal respect (by such conventional means as drumming, singing, clapping of hands, offering of beer, beads, white cloth and money). After this the misfortune is supposed to cease. The afflicted lives on as a member of that agent’s specific cult; he participates in cult sessions to reinforce his good relations with the agent and to assist others, similarly afflicted, to be initiated into the same cult.

Cults of affliction represent a domihant class of cults in present-day Western Zambia. However, in addition there are ancestral cults, individual specialists’ cults, chiefly cults and various types of Christian cults.[37] Of these, Christian cults and the major chiefly cults are regional (Van Binsbergen 1976d, 1977a).

2. Non-Regional Cults of Affliction

2.1. Main Characteristics

Non-regional cults of affliction occur in scores of versions throughout Western Zambia and surrounding areas (Carter 1972; Van Binsbergen 1972b, 1976d). These various cults have much in common. They differ from each other mainly in the following respects: the associated paraphernalia (e.g. type of adept’s shrine, musical instruments, vegetal medicine, ritual objects and bodily adornment used); the specific misfortunes for which each cult caters; and the name and alleged nature of the invisible agent venerated in each cult. Each cult is locally known by the name of this specific agent. Some particularly successful cults are encountered over large areas and are mentioned repeatedly in the literature; these include such cults as mayimbwe, muba, songo, viyaya, liyala and bindele.

We are indebted to Turner for his masterly description and analysis of some cults of affliction among the Ndembu. He shows the role of these cults in the succession of “social dramas” that make up the micro-history of villages over a time-span of several decades. However, he does not analyse the historical dimension of these cults themselves (cf. Shorter 1972: 141). In Turner’s work these cults are viewed as given institutions that are manipulated within the local social process, rather than as the dynamic outcomes of a process of religious change that in itself needs elucidation.

White (1949, 1961) was the first to explore the cults’ historical dimension. He pointed to a major conceptual shift that has affected cults of affliction in Luvale country, since the 1930’s. The shift was away from cults that claimed local ancestors (particularly diviners and hunters) as their afflicting agents. It was towards new cults that attributed affliction to abstract, scarcely personalized agents whose names were reminiscent of contacts with distant alien groups, African and European. Some of these new cults were songo (after the Songo people of Angola), ndeke (airplane), bindele (white people or people clad in white: either Europeans or alien African traders). Turner’s (1957: 296f) and my own fieldwork suggest, moreover, that the conceptual shift was accompanied by a shift in the recruitment pattern of the cults’ congregations. This structural shift was from a ritual congregation that was rather coterminous with the local community and that focussed on communal symbols (village shrine, ancestors), to a congregation whose members, individually drawn from a series of adjacent local communities, would associate for ad hoc ritual purposes in a pattern that cuts across, rather than reinforces the structure of distinct local communities. However, as I shall show, in contrast to Turner’s observations, my data point to the significance of local factors in the cults (see below, 2.4).

White suggested that the conceptual shift may be attributed to a complex of interrelated changes. These included increased interlocal contacts (through 19^th century long-distance trade and 20^th century labour migration), the breakdown of the village community, and the emergence of new causal models for the interpretation of misfortune. White interpreted these new models (featuring new, impersonal, non-ancestral affliction agents) as the conceptual response to the widening of the social horizon beyond the area where the ancestors could be thought to hold sway.

Elsewhere I have developed White’s stimulating ideas in the light of more comprehensive data on religious change in this part of Central Africa (Van Binsbergen 1976d). There I argue that the emergence of these non-regional cults of affliction represents just one specific outcome in an overall process of religious transformation. This process has extended over centuries and, as other outcomes, has produced such religious forms as chiefly cults, prophetism concentrating on “ecological” concern for the land and its produce, eschatological prophetism, and sorcery eradication movements. Throughout a succession of emerging religious forms, the overall process produced systematic changes in religious idiom. These changes can be conveniently mapped along five dimensions: the perception of time; the degree of “ecological” concern; conceptions concerning the nature of supernatural beings; the degree of individual-centredness; and the interpretation of evil. Ultimately, major political and economic changes can be seen as the motor behind the overall religious transformation. In this context I am thinking primarily of the increase of interlocal structures of political and economic power, distribution, and movement of people due to precolonial and colonial state-formation, raiding, long distance trade, and labour migration. These extra-religious changes called for new religious forms (including non-regional cults of affliction) that were capable of legitimating the new structures and that could meet the existential and interpretative needs of the people involved.

My field data mainly derive from the Nkoya, an ethnic group inhabiting the wooded plateau of central Western Zambia (Clay 1945; McCulloch 1951; Symon 1959; Van Binsbergen 1975, 1976a, 1976b, 1976c). Among the Nkoya, ancestral cults of affliction have continued to exist until today, but the great majority of affliction cases is now treated within the more recent, non-ancestral cults, particularly those of the non-regional type to which the present section is devoted. People recognize the recent nature of these cults, and can state the period in which a particular cult first reached their area. The non-ancestral, non-regional cults of affliction spread like fashions or epidemics. For many cults occurring in an area it is known from which direction they arrived; they became locally associated with ethnic groups living in that direction, and are sometimes held, incorrectly, to represent those groups’ ancestral ritual.[38] Intervals of a few years would pass between a new cult’s arrival and its being forced out of fashion by the arrival of yet another one. As an area is hit by a succession of cult fashions, many people come to be initiated into more than one cult. As a fashion wears out, the ritual for that cult will be staged less frequently and fresh cases of affliction will be attributed mainly to more recently arrived cults.

However, cult leaders keep staging the rituals of non-ancestral cults of somewhat older vintage. In this way they revive their ritual links with adepts whom they initiated some years before and who may since have taken not only to other cults but also to other leaders.

The spread and contemporary distribution of non-ancestral non-regional cults of affliction show that Central African ethnic groups (“tribes”) are not rigidly bounded, either geographically or culturally (cf. Van Binsbergen) 1975). The cults spread easily from one ethnic group to another, despite such great language differences as between Luvale and Lenje, Tonga and Lozi. Moreover, immigrant adepts from a distant ethnic group often proved acceptable as local cult leaders (a point emphasized by Symon 1959). Cult songs have sometimes been translated into a local language, but have often been retained in the original language of the ethnic group from which this cult was borrowed locally. Sometimes the original ethnic group even gave its name to the cult (songo, chimbandu). All this bears out the fact that, among other aspects, the non-ancestral cults of affliction represent an attempt to come to terms (both conceptually and interactionally) with the reality of extensive, inter-ethnic, interlocal contacts.

2.2 Ritual

Usually the differences between the non-regional, non-ancestral cults of affliction which occur in one area, are slight. There is somewhat of a tendency—it is not invariable—to associate certain paraphernalia (a peculiar type of headdress or scapular, miniature axes and hoes, fly-switches etc.) and certain bodily movements with particular cults. The main way, however, to tell one cult from another is from the contents of the songs that are chanted during the sessions. Below are some typical song texts:

1 Bakeshabe iyale yomana yowelile mama bamikupe iyale yomama iyale yomama lyiyo mama yowelile (text used alternatively for kasheba, bamikupe and mushelu) 2 Tu ba kombele biyaya bi neza tu ba kombele biyaya mama (text for biyaya) 3 Ba ka lenda mawulu mu pange mu ni yoyise lelo moyo mu pange mu ni yoyise bi nakunisisayela lelo moyo ba nganga na ba nganga kumpanga bafako (text for kalenda-mawulu)

From kasheba,really, mother, I am ill, mother from bamikupe, really mother, I am ill, mother o mother, I am ill We pray for biyaya has come we pray for biyaya, mother For this foot disease, treat me to make me alive today the life treat me to make me alive they don’t restore my life today these cult leaders or those cult leaders they don’t treat me

These songs are sung by a small chorus of adepts, instructed and led by a cult’s local leader, usually a woman. The adepts (usually women) stand in a semi-circle which is closed by a row of (male) musicians. Within this enclosure the cult leader and one or more novices dance, and enter into a trance. Before the session starts the leader has undertaken to treat the novices. She exhorts the latter, the chorus and the musicians so that the novices’ movements may catch up and gradually harmonize with the music. Thus the novices may reach ecstasy. In the course of most sessions vegetal medicine is prepared from a selection of plant species which varies with each cult. Although each session tends to be dominated by the songs and ritual for one particular cult (the one in which the leader excels), often elements of other cults are used in abundance. This is particularly the case when a patient cannot be brought to respond properly to a session’s main cult. Then one cult after another has to be tried in order to arrive at the correct diagnosis and treatment. Sessions last through the night and end at dawn with the final distribution of medicine to novices and adepts and occasionally to non-participant onlookers. For in contrast with the situation in the extreme north-west of Zambia, as described by Turner, in this area a small audience of non-adepts is present at every session, to share in the beer, to help out with the music, and to witness the patients’ struggle for healing.

2.3 Morality

Part of the night’s proceedings is, however, private: the therapeutic conversations which the leader and some senior adepts have with each novice. Their aim is to ascertain the specific personal conditions under which the affliction, dormant until then, has become manifest. Such conditions include a recent death, ill-will and suspected sorcery in the patient’s environment, or the occurrence of a major healer, diviner or hunter among the patient’s deceased kin.

The information obtained in the “anamnesis”[39], however, is not fed into the patient’s social environment. The information only serves to identify, to the officiants’ satisfaction, the occasion at which the affliction agent chose to manifest itself. Even if this occasion highlights interpersonal conflict and illicit aggression (in the way of sorcery), moral redress, rehabilitation or prosecution never become the cult’s concern. Once the diagnosis in terms of the cult’s etiology has been pronounced, the cult tries to curb the invisible agent’s harmful effects, but does not try to take away any group-dynamical causes of the misfortune. The non-ancestral, non-regional cults of affliction aim at individual readjustment. They fight symptoms of maladjustment but do not try to expose or resolve any underlying social tension. Nor are they meant to do so. These cults are concerned with suffering, and not with morality. Their frame of reference features patients afflicted by essentially unpredictable, non-human agents—and not victims injured by human evil-doers or by the effects (through ancestral revenge) of their own sins.

The non-ancestral non-regional cults, however, exist side by side with rival and alternative approaches to misfortune, which do emphasize morality and guilt, reparation and retaliation. Among the contemporary Nkoya, these rival approaches are frequently applied. They include: divinatory and redressive ritual in the idiom of ancestral intervention; similar ritual in the idiom of sorcery and counter-sorcery; litigation in more or less formal courts of law (Van Binsbergen 1976c); and finally regional cults of affliction, whose relation to non-regional cults with regard to morality I shall discuss below in section 3. A number of venues for redress are thus open to an individual facing misfortune. The choice is largely his own. He may apply to a diviner/healer known for diagnosis in terms of a cult of affliction, rather than to one who habitually propounds interpretations in terms of ancestors or sorcery; or he may apply to a court. The actor’s underlying decision model is complex and can only be analysed on the basis of detailed description of the various institutions involved, and of specific extended cases. However, if one pursues the option of cults of affliction, this can be safely said to indicate one’s reluctance to publicly present one’s misfortune as caused by human conflict (sorcery, court) or guilty neglect in one’s dealings with kin (ancestors). Such reluctance usually reflects lack of power, material resources and social support; the victim can neither answer a challenge through counter-sorcery and litigation, nor face such public disapproval as offences punished by ancestors inspire. Non-ancestral cults of affliction, however, interpret misfortune as an entirely individual condition which (in contrast with sorcery, ancestral intervention, or litigation) is not primarily determined by the victim’s interaction with other people—although such interaction may form the occasion for the hidden affliction to manifest itself for the first time. Amongst the institutions dealing with misfortune, these cults have, therefore, a marked competitive advantage among certain people: those who are dependent, relatively powerless, and short of social credit. Little wonder, then, that most of the patients in the non-ancestral cults of affliction are women and youths. Mature and elderly men, as a rule, pursue other venues for redress. And if they involve themselves at all in such cults of affliction, it is not as patients, but (occasionally) as cult leaders, as musicians, and particularly as sponsors for their womenfolk.

This implies that at least for women, as patients, the cults of affliction do not provide an idiom of internalized powerlessness, but rather an alternative way to demand, and receive., male support—albeit outside the arenas where mature men compete with each other. This, of course, does not preclude such competition between leaders and sponsors (see Werbner, 1971–1972: 235).

2.4 Roles, Personnel and Organization

For a successful session all the following conditions have to be met. The cult leader must be generally recognized and accomplished. The chorus (of a minimal size of four or five adepts) must know the basic song texts and must be prepared to assist in the proceedings. Musicians are needed who know or can quickly pick up the music peculiar to the various cults, and who are prepared to follow the directions of the leader. Musical instruments, beer and firewood are required. And finally, somebody has to organize and meet its expenses.

In a rural society where firewood is becoming less abundant, where beer has become a marketable commodity,and where drums and other instruments are scarce and privately owned, the logistics of the sessions often turn out to be quite complicated and problematic. Each session necessarily involves the sponsor, and the other participants in the session, in a network of transactions and obligations that both reflect, and bear upon, the extrareligious social process in the community. Good musicians cannot always be recruited from amongst a sponsor’s close kin, and even if they can, they often demand a payment roughly equivalent to a day’s wages. The adepts who constitute the chorus, do not receive any remuneration beyond a share in the beer that the sponsor furnishes for the cult leader. Besides this beer, and occasional gifts of white cloth, bottled fizzy drinks, bottled beer and other luxuries, the leader receives a small sum of money for the diagnosis and treatment during the session. This sum is again equivalent to a day’s wages. Even more important, the leader establishes a firm claim (actionable in court) against a much larger sum, easily a month’s wages. This large fee is payable any time after the patient has made significant improvement—no matter what additional treatments are pursued elsewhere.

From the sponsor’s side, therefore, the session involves immediate costs that are considerable, both in social and financial terms and, if the treatment is successful, heavy debts in the future. For the leader, the session means a significant gain immediately, and very likely a large financial claim in the future. At the same time, it also means a public test both of ritual skill and of such management power as is required for the mobilization of the chorus in the first place (it is the sponsor’s responsibility to recruit the musicians), and for the leader’s control over adepts and musicians during the many hours of the session. The adepts in the chorus clearly hold a key position. They are indispensible to the leader’s success but hardly share in the latter’s ‘ benefits.

The basic organizational form of non-regional cults of affliction is a smdll faction of loyal adepts around a leader, all living within walking or cycling distance from each other: within a radius not exceeding 10 kilometres, and usually much less. The leader can control the chorus’ adepts in two cases, which often coincide: when the adepts are themselves still under a course of treatment with the same leader (and treatment may, partly for this reason, be extended over a period of years); and when the adepts are very close kin of the leader. Kinship ties among the Nkoya are, however, not automatically effective in bringing about and maintaining solidary social relationships. In order to be effective they need to be backed up by frequent interaction, e.g. through co-residence. Attachments between kin are readily disrupted as people change their residence in order to seek their economic and local-political fortune elsewhere, among a different set of real or putative kin. Such individual intra-rural geographical mobility is extremely frequent and, in fact, constitutes a major structural theme in Nkoya society. Likewise, the curative link between leader and adept is often threatened, both by rival leaders who compete for adepts (as well as for new patients), and by the senior adept’s own aspiration to set up as a leader herself. Therefore, within (and between) the cultic factions tensions and conflicts are rife; and these tend to come to the fore during sessions, when the leader is most dependent upon the adepts. I have never attended a session that was not repeatedly interrupted by heated argument between the officiants. After a session there tend to be shifts in the faction’s internal structure of relationships, and even in its membership.

The leader’s status is rather insecure. Non-regional cults of affliction define the roles of leader, adept and novice, but they do not stipulate in detail the requirements by which one gets access to these statuses. These cults do not have a formal organization with fixed rules of eligibiity to office. Instead, recognized ritual leadership is a matter of public opinion. Whoever can persuade others to let themselves be treated by her, and can mobilize the necessary assistance, is a cult leader. On the other hand, a leader whose curative effectiveness begins to be doubted by many (partly as a result of her being forced out of competition by rivals) loses her patients and adepts.

How does one get launched as a leader? Long-term apprenticeship as an adept of an established leader is one way. In the take-off period as a leader, other standard methods to woo public opinion include self-appointed curative capabilities which are claimed to derive from dreams, from accession to the name of a deceased relative who was a well-known healer, or from a serious psycho-physical disturbance now overcome by a recovery— so that one is now acquainted, better than most people, with the experience of illness and death. Gradually, new*songs (often featuring the name of the leader herself), an original choice of paraphernalia, a notebook in which the names of treated patients may be recorded, fake or real licenses issued by local authorities, can all help to lend credibility and identity to the new leader. In this search for a personal “trade-mark”, considerable expense and creativity is invested; even though the result is usually a permutation of the same limited set of elements used by all leaders and all non-regional cults of affliction.

The main condition, however, to make the grade as a cult leader is the construction of a local network of loyal adepts who can assist in the treatment of new patients.

My use of the term “faction” for the basic local unit of officiants, has suggested the problematic internal dynamics, the shifting membership of these units, and (as a major cause of this) the intense competition between units. In the immediate social environment of each afflicted individual, a number of leaders with their temporarily loyal adepts are active. Leaders are in competition with each other both for the senior adepts on whom their practice is dependent, and for new novices who will boost the leader’s public status and bring in large fees.

The fees enhance a leader’s power in two ways. Once received, this wealth is largely redistributed in non-ritual transactions with kin and neighbours. But since most fees cannot be paid immediately, they also tie the patients to the leader with heavy debts. Both ways, the leader’s power must have considerable effect on the extra-religious social process. Moreover, the fees are so high and cash is so scarce in the villages, that the modern cults can be properly said to constitute a major redistributive economic institution. This holds true not only between villages, but also between village and town. Rural-based leaders treat afflicted urban migrants either by receiving them on short visits in the rural area, or by travelling to town at irregular intervals.

Both within and outside the religious domain, leaders hold enviable positions which they are constantly defending against the encroachment of others. As adepts and leaders have usually been initiated into more than one non-regional cult of affliction, and as the basic idiom of these cults is constant and well-known, an adept’s shift from one leader to another (which often means: from one cult to another) presents no great difficulties in terms of knowledge and skills. Leaders greatly resent it when such a shift diminishes their effective following. For this reason, leaders tend to avoid each other and, when they operate within the same social field, are often at daggers drawn because of having snatched each other’s followers in the past. Sometimes, however, leaders combine locally, stage sessions together and visit each other’s sessions as guests of honour. On such occasions, emphasis is on manifestations of mutual professional respect, symbolized by the ostentatious furnishing of a chair for each leader[40]. But even then each leader is on her guard lest her colleagues make too favourable an impression on the audience, or attempt to lure adepts and patients away. Only when such professional contacts are backed up by residence in each other’s proximity along with close kinship ties and good extra-religious relations in general, can they be seen to develop into stable, prolonged co-operation. In such a case leaders are no longer afraid of mutual poaching; they share patients and adepts.

These local ritual factions with their occasional co-operation between leaders, constitute all the rudimentary organizational structure that the non-regional cults of affliction possess. Leaders are predominantly ritual entrepreneurs, who exploit a local market on the basis of the population s general adherence to the cult of affliction model. Even between leaders in different localities who profess that they specialise in the same named cult (e.g. Viyaya), there is normally no contact—and often downright avoidance. Each may have learned this cult’s ritual in a different place, may have added her own innovations, and would have her own local clientele to consider. The differences between the ritual performances of two such leaders of the same cult may be greater than those between leaders who avowedly specialize in different cults. Each elaborates on the general idiom of the cults of affliction in her own personal way, without any interlocal formal organization binding local leaders or dictating any orthodoxy. Public opinion is prepared to accept a leader’s reputation and following as proofs of expertise, and it favours any innovation that does not greatly run counter to local canons of decency. But such public opinion concerning the cults is in itself mainly a response to the actions and pronouncements of the cult leaders. Thus it is by virtue of a widespread cultural model that each non-regional cult of affliction exists: each cult has its unstructured series of small factions which are disconnected, local and rather ephemeral. The widespread cultural model is continuously reinforced, at the grass-roots level, by the uncoordinated activities of hundreds of leaders in search of wealth, power, ritual prestige, and self-expression.

2.5 Ritual Leadership as a Callin

While exploring the political and economic dimensions of the non-regional cults of affliction, we should avoid reducing them to these and nothing more. Leaders, adepts, patients and outsiders are keenly aware of the economic and local-political aspects of the cults. Yet all of them Consider the cults’ healing efficacy their overriding justification.

The leaders, mainly elderly women, are not cynical operators and no more. They are gifted and passionate manipulators of symbols: song texts, paraphernalia, dramatic effects—and cash. Through an existential crisis earlier in life they have often paid the price (not just financially) towards becoming specialists in human suffering and some of its remedies. Their life-histories tell of prolonged illness, periods of insanity, ridicule by fellowvillagers confronted with the aspiring leader’s first, clumsy attempts at being a healer, and more recently the treachery of once loyal patients who have at last set up as leaders themselves and no longer pay respects to one who, in her own view, fought for their very lives. A leader’s strikingly intense and agile performance during the long nocturnal sessions may be partly motivated by a drive for money and power. But it has as much to do with art, while the prime concern is to heal a fellowman. If the ritual does not appear to succeed, it is not only a threat to professional prestige, but also compassion that forces the leader to exert herself for hours at a stretch, finally to bring the patient to join in the rhythm—thus making recovery possible.

Better than anyone else, the leader understands and believes in the idiom of the cults of affliction. She is the mediator of a conceptual system which is commensurate with the changing social order in which she and her patients find themselves. Offering deliverance from suffering is a professional calling for which the easy life, food, marriage, all have to be sacrificed. The leader Kashikashika vividly describes how her calling affects her life:

“At home I have no time to eat. I eat with one hand, keeping one hand on the head of a patient. (Laying-on of hands is a minor form of treatment in the cults in which she specialises.) I have no time to dress, but instead walk about in my short petticoats. I have no time to sleep. In the middle of the night people come and knock at my door for treatment.

“Years ago, a man fell in love with me and wanted marriage. I told him: ‘Before I go and live with you, realize how my calling affects my life. I shall often have to go out in the night. Sometime I shall have to stay away from you for three months at a stretch, to attend to my patients!’ But he did not want to listen and went ahead with the marriage. After a month he started complaining, and we soon divorced. Now I am married again, with one of my patients. He says: ‘Look, she has cured me. Should I complain when she goes out and stays away to cure others like me?...’.”

The husband is present and amply confirms that this is indeed his view of the matter.

Underneath this testimony of professional sacrifice there is, unmistakably, a note of pride and female power quite characteristic of elderly Nkoya women, and of cult leaders a fortiori. But at any rate, calling obliges; and many leaders have strict personal standards as to the size of fees, the requirements for proper diagnosis, and the ethics of dealing with uncooperative debtors amongst their patients.[41] Without exception, leaders have themselves suffered, in the past, from the afflictions they treat, and usually they are still in frequent silent communication with their own affliction agent, inside. Allegedly, the agent advises on the leader’s wellbeing, but he does not take mockery and is sure to punish a cheating leader of his cult. Adepts and patients are aware of this security device, too, and as a rule expect fair play from the cult leaders they deal with.

The high revenues and other benefits in terms of power and prestige associated with cult leadership, can hardly fail to appeal to the ambitious and calculating sides of the incumbents’ character. Yet I would maintain that the requirements in terms of artistic and emotional dedication, and in terms of self-confidence vis-a-vis the symbols manipulated in the cults, are such as to make the sham leader, who stages for money’s sake a ritual whose efficacy she secretly denies, a rare occurrence.

3. Regional Cults of Affliction in Western Zambia: General Characteristics

Due to a succession of non-regional cults of affliction that had been spreading over Western Zambia for several decades, the general idiom of such cults had become established by the 193O’s. By this time the new type of regional cult of affliction emerged. So far, three cults of this type have been studied: Nzila (also known as the Twelve Society), Bituma, and Moya. It is likely, however, that Western Zambia has witnessed the rise of several more, which subsequent research may throw light upon. The first description of Nzila is by Reynolds (1963: 133f), on the basis of an administrative report by I.H. Whethey in 1957.[42] In 1972, Miss M. Muntemba carried out research into this cult, mainly in the Livingstone area (Muntemba 1972); I am greatly indebted to her for the lengthy conversations we have had on the subject. The Bituma cult played a major role in my Nkoya fieldwork (cf. Van Binsbergen 1972a for a preliminary account of the cult). Ranger (1972) gives lengthy excerpts from both Muntemba’s and my own descriptions. The Moya cult is briefly discussed by Ikacana (1971: 33); although this author mentions the healer Liminanganga as this cult’s main leader in Kaoma (then Mankoya) district, thus suggesting a certain centralization that is characteristic of a regional cult, his further discussion gives the impression that by the 1940’s Moya was a non-regional cult. During my fieldwork in the early 1970’s, however, a regional cult of the same name was active in the eastern part of the district, where, headed by a prophet called Moya after his cult, it had penetrated only a few years previously—but perhaps not for the first time. My limited data on this cult do not enable me to decide whether it became a regional cult only recently, or was already one in Ikacana’s time.

Like the factional, non-regional cults described above, these three regional cults are cults of affliction. They interpret disease and misfortune by reference to a non-human agent, and attempt to redress the disorder through divination and subsequent initiation into the cult. As cults of affliction deriving from the same cultural area, the regional and non-regional cults have on the surface much in common. However, there are very significant differences between the two types.

The most striking aspect of Nzila, Bituma and Moya is their regional organization. These cults each have a central place, which is the seat of the cult’s supreme leader. Cult officers appointed by the leader are responsible for sections of the total area over which the cult has spread. They communicate regularly between local cult congregations and headquarters. In close cooperation with the supreme leader, they recruit new local adepts and leaders, enforce observance of the cult’s regulations (e.g. concerning the size of fees for treatment, and the forwarding of a portion of the fees to headquarters), supervise the ritual locally, and guard against undue autonomy and ritual deviation of local leaders. In contrast with the non-regional cults, effective leadership of a local congregation depends not so much on the manipulation of public opinion and the control of a loyal local ritual following, but primarily on admission and promotion within the cult’s hierarchy, subject to approval from headquarters. Literacy, in its simplest forms, plays a role; one may keep a record of the number of patients healed, of their names, and of the fixed fees for each type of treatment. The regional cults are more or less formal organizations that enforce, among local cult leaders, ritual conformity, compliance with the cult’s authority structure, and the forwarding of funds, over areas of many thousand square kilometres.

We have seen how leaders in non-regional cults of affliction tend to make considerable innovations upon the cultic material they use. Also, these leaders have usually gone through deep personal crises before establishing themselves as cult leaders. In these respects they have much in common with the founders of Nzila, Bituma and Moya. However, in the case of the non-regional cults a leader’s specific innovations and personal history are played down as irrelevant. The leader is viewed not as the inventor but as the mere transmitter of cultic forms which, allegedly, have already existed elsewhere and which could have been made available locally by any other leader. The regional cults, by contrast, are quite sharply considered to be founded by a particular prophet, whose name, visions, life history, the process in the course of which he shaped his cult, are all held to be very important and are often known even among non-adepts. In contrast with the leaders of non-regional cults of affliction, Chana (founder of Nzila), Simbinga (founder of Bituma), and Moya (founder of the Moya cult) all started out as charismatic leaders in the Weberian sense (Weber 1969: 358f). Each displayed what I have elsewhere (Van Binsbergen 1976d: 71) called the “standard biographical pattern for Central-African religious innovators”. Each reached prophethood through the same stages. Falling victim to a chronic disease or defect is a first stage. Chana was a leper and suffered from the bindele affliction which, however, the non-regional bindele cult proved unable to cure. (Later the difference between Chana’s affliction and bindele was acknowledged by renaming the former nzila: “path”.) Simbinga and Moya likewise suffered from afflictions of a nature until then unknown. In the next stage the affliction would develop to a crisis (which in the cases of Simbinga and Moya allegedly involved temporary death, in Simbinga’s case even burial). In this crisis the prophet allegedly received very specific messages and directions from the supernatural agent claiming responsibility for the affliction. Following the indications received in the messages, the prophet would be able to recover. He would then have the power and the calling to apply his new insights to those in his social environment who were similarly afflicted. These first patients (twelve in Chana’s case, seven in Simbinga’s) would form the original core of the cult, soon to be augmented through the combined efforts of prophet and initial followers.

The regional cults do not merely share the affliction-cult idiom with the non-regional cults. The former actually emerged upon the substratum of these earlier, non-regional cults. All three started as an individual’s response to an affliction that could have been cured within the established non-regional cults of affliction. Instead, the three patients devised their own cults.

Now why did these three prospective prophets fail to get healed within the existing non-regional cults of affliction? For all three, contemporary informants claim that they had tried such treatment; in Chana’s and Moya’s case, modern medical services were consulted as well. The standard course of their crises suggests that, in addition to any bodily disorders, a profound mental struggle was involved, bearing on existential problems such as the meaning of suffering and the interpretation of the universe. The then current, non-regional cults revolved around affliction-causing agents of an abstract and hardly personalized nature, mainly conveying associations with alienness and neighbouring ethnic groups. The solutions the three prospective prophets were seeking were to be—to judge from what they came up with—of a less particularistic and human scope. In the accounts of the founders’ original visions and in their current interpretations of affliction cases and their treatment, the three regional cults propound strikingly similar, new conceptions concerning the nature of the supernatural agents involved in affliction. Chana, Simbinga and Moya attributed .their visions, their miraculous recovery and their subsequent healing power to entities-closely associated with or, identical to, the High God. In Chana’s case this was a spirit who in dreams and visions manifested itself as a white being, exhorting him to acknowledge God, invoke God’s help, and preach about God to his patients. Simbinga had similar visions, which he interpreted in terms of an Angel (Angero) from God. Moya’s interpretations concentrate on the Life-Spirit (Moya), source of life and capable of dispensing health and illness at will. These are concepts of a much greater universality and with much more specifically theistic connotations than the affliction-causing principles of the non-regional cults.

But while this conceptual shift may have primarily sprung from the specific existential problems of the three prophets, the latter were at the same time yielding to a general tendency which since the beginning of this century has been encountered in quite a number of religious innovations in Western Zambian. Most of these movements had no specific concern with healing. They include a number of Ila prophets, amongst whom is the great prophet Mupumani (1913) whose movement had an impact throughout Western Zambia and the surrounding countries; minor eschatological prophets in Mamwala, Kalabo and Mwinilunga; the numerous Watchtower preachers and dippers moving through the area in the late 1920’s and the 1930’s; and even the Christian missionaries. All these religious innovators displayed the same conceptual shift towards ever greater prominence of the High God (Van Binsbergen 1976d).

A final major difference between the regional and the non-regional cults of affliction is that the former do display a moral element which, as we have seen, is lacking in the latter. Chana was instructed by the Spirit:

“to teach the people what to do and what not” (Muntemba 1972: 2).

He was given

“rules and instructions he was to teach his people at the Sunday afternoon gatherings.(...) the rules centre around purity of mind and spirit as the key to a healthy body. At the Livingstone branch service about one hour and thirty minutes are given to instructing the people, in the form of question and answer, in how to attain purity. (...) Members are exhorted to keep clean thoughts and clean minds by not bearing malice against their husband or wife, their brothers, sisters, in-laws and neighbours but instead to love. They are warned against bearing false witness, to avoid quarrelling and drink, drink which often leads to quarrels and other irresponsible acts. They are admonished against eating certain foods (...) which are considered unclean and which would therefore foul the body and soul of man” (Muntemba 1972: 5f).

Reynolds gives a similar picture and emphasizes the patients’ surrender of sorcery apparatus in this context (1963: 135, 138).

In the teachings of Simbinga and Moya the same emphasis on purity versus pollution is encountered. For them, the whole range of local medicine represents evil and pollution in its fundamental form. When Simbinga was called to prophethood in the early 1930’s, he immediately launched a fierce attack on local herbalists, diviners, and leaders of non-regional cults of affliction. He wanted them to give up their manipulations of material apparatuses, paraphernalia and medicine, and instead accept his own interpretation of misfortune (which by this time, however, was still blurred and inconsistent), and his treatment by means of laying-on of hands, prayer and dancing.

This is a moral issue for various reasons. First, much local medicine was (and is) locally applied for purposes which, however common in the local society, were considered immoral acts of sorcery: to attain excessive power and success; to harm or kill rivals and enemies. Moreover, especially the most powerful medicine derived from sinful manipulations of human bodies that were killed or desecrated for the purpose. And finally, on the basis of his visions Simbinga, like the two other prophets, presented an alternative theory of causation of affliction, whose universalist and theistic overtones implied a moral rejection of previous, medicine-centred interpretations.

Simbinga was at first unsuccessful in his attack. The local practitioners proved disinclined to accept his views. They mobilized public opinion against Simbinga and after a relapse of his affliction he had to leave his area. He travelled to the north-west (to the Balovale—now Zambezi—area). After a few years he returned with a revised version of his cult. This time material paraphernalia, divining apparatus, and vegetal medicine had been given a prominent place in the Bituma idiom. However, the cult has retained its abhorrence of all other forms of medicine and material apparatus, and still considers envy, rivalry and sorcery the main occasions upsetting the harmony between a man and his Angel—interpreting the Bituma affliction as the manifestation of such disruption.

The Moya cult has adopted a rather similar position. But whereas Bituma, though conscious of the dangers involved, does not absolutely forbid its adepts the possession and manipulation of medicine, Moya does—including hunting or fishing medicine, or products from modern dispensaries. In both cults, adepts and particularly leaders are continually aware of the dangers of pollution that threaten them everywhere. For this reason they may avoid certain persons and activities. However, beyond the reluctance to apply sorcery medicine to further their own interests, the purity-orientated personal morality as derived from the regional cults appears to have little impact on everyday life.

For even if the regional cults differ from the non-regional ones with regard to moral connotations, they are quite similar in a more fundamental respect. Both types of cults are based on an individual-centred interpretation of misfortune: misfortune is considered solely a matter of the relationship between victim and an unpredictable supernatural agent. These cults do not link misfortune to the small-scale social proceses in which the victims are involved, and therefore cannot have a direct impact on such processes.

There is an intimate relation between these three outstanding features of the regional cults: their founders’ charisma, the appeal to more universal, theistic supernatural agents, and the introduction of a moral element. In the idiom of spirit possession, association with alien or neighbouring ethnic groups might well provide a basis for personal charisma. But somehow this seems less effective in a society where through interlocal contacts across ethnic boundaries, such association has become commonplace over the past century, and where a considerable variety of religious innovatory movements have already effected an overall tendency towards greater prominence of the High God. Under these conditions, visionary experiences involving supernatural entities of a much wider scope are more likely to endow exceptional individuals with the charismatic qualities that could lead to effective mobilization over a vast area. In other words, I suggest that the founders’ charisma, and their propounding of universalist, theistic supernatural agents, imply each other.

Although universalist, theistic concepts were circulating in the area well outside the sphere of missionary Christianity, the three prophets started out under the latter s direct influence. Prior to their appearance as prophets, both Chana and Simbinga had been thoroughly exposed to missionary Christianity. Chana’s parents had been converts of the Seventh Day Adventist Church, and Chana belonged to this church before his prophetic calling. He even continued to adhere to this church while heading the Nzila cult (Muntemba 1972: 7f. Reynolds 1963: 133f suggests the contrary). Simbinga was among the first evangelists of the fundamentalist South Africa General Mission at Luampa, and appears to have received further Christian teaching as a labour migrant in South Africa. Although very little is known yet of Moya’s biography, the preponderance of Christian elements in the ritual of his cult (bible copies, church bells, and a supernatural entity whose name is the current local translation of the Christian concept of the “Holy Ghost”) reveals a similar background.

If the universalist, theistic elements endowed the prophets with just the amount of charisma they required to set up regional cults, their propounding of moral elements fits well into the same picture. As Weber (1969: 361) wrote, “From a substantive point of view, every charismatic authority would have to subscribe to the proposition, ‘It is written..., but I say unto you...’ The genuine prophet, like the genuine military leader and every true leader in this sense, preaches, creates, or demands new obligations”.

However, the development of a regional cult organization is an aspect not of charismatic leadership itself, but of its subsequent routinization (Weber 1969: 363f). Given the charisma, universalist and theistic agents, and moral emphasis by which the regional cults differ from the substratum of non-regional cults of affliction from which they sprang, the development of more or less formal interlocal organization does by no means follow automatically. Instead, such development represents a major analytical problem. Apart from the ill-documented Moya cult, we are fortunate in having, in the Nzila and Bituma cults, two reasonably well-known cases whose development into regional cults displays such divergence, that a detailed comparison is likely to yield some insights into the general dynamics of regional cults.

4. The Development of Nzila and Bituma as Regional Cults

4.1 Nzila

After a spell of labour migration in Zimbabwe, Chana returned to his village in Kalabo district. He developed his affliction, cured himself through the methods shown to him in his visions, and appeared as a prophet in the early 1940’s.[43] His methods involved the erection of a shrine, daily ablutions in this shrine with water medicated by selected herbs, and nocturnal ritual dances. Having recovered, Chana successfully applied this method to his first twelve patients, thus obtaining his first followers: the twelve first Nzila “doctors”. Subsequent visions instructed him to put faith in God, to organize weekly services for Him on Saturday afternoons and to extend his activities to moral teachings. Thus a formal ritual routine was worked out, quite reminiscent of a Christian church service; the influence, in particular, of the Seventh Day Adventist Church is unmistakable (Muntemba 1972: passim; Reynolds 1963: 137). Assisted by his doctors, Chana treated a rapidly increasing number of patients in the Kalabo, Mongu, Senanga and Kaoma districts. A cult centre was erected near Mongu, the capital of the then Barotse Province (now Western Province). The growth of the cult was such that already in 1952 a larger building had to be constructed. At the cult centre, Chana would personally diagnose the patients’ affliction as his own. After the diagnosis, the patient would return to his place of residence, where under the guidance of one of the doctors he would construct a windbreak in which he was to wash every morning, using medicated water. This would be continued until the patient appeared to be cured—which could take up to five months. Patient and attending doctor would then apply again to the cult centre, where Chana would subject the former to a series of tests in order to ascertain if he was truly cured. Some of these tests would be public and within sight of the weekly gatherings at the centre, whereas the final test would take place in Chana’s sacred enclosure (Reynolds 1963: 136f). This finished, the patient would be told what he had to pay:

‘Payment is normally in the form of livestock; a cow, a few pigs or goats, but frequently, a fairly substantial sum of money has to be paid. He must also bring a beast to slaughter for the feast that night’ (Reynolds 1963:137).

During this feast the cured patient would be installed as a new doctor of the Nzila cult. This gave him the right to build his own sacred enclosure at his own residence. He would have authority to attend on new patients during their months of ritual washing, but could not use his own sacred enclosure to diagnose or pronounce final recovery: this would have to be carried out at the cult centre, by Chana personally.

In the mid-1950’s the Mongu centre under Chana’s direct leadership catered for the Mongu, Kalabo, and Kaoma district; a subsidiary branch, led by one of the twelve doctors, was then already in operation in the Senanga district. Branches soon began to proliferate. Muntemba (1972: 3) reports that

“By the late 1950’s it (the cult) had spread to most areas in Zambia that had strong Western Province influences and also to Bulawayo and parts of Botswana. In the 1960’s branches were started in most towns along the line of rail including the Copperbelt. (...) People not only from Western Province but from other societies became Nzila members as well. The movement was registered as the Zambia Nzila Sect in 1966. Its membership stood at 80,000 people ihen. On 30^th March, 1972, the membership figure stood at 96,872.”

Thus in a quarter of a century the Nzila cult developed from a core of a handful of followers around a charismatic leader, into a fully-fledged interlocal formal organization with government recognition, branches and property (in the form of buildings for worship) in most major towns of Zambia. The supreme leader’s personal control over the crucial stages in the healing of each patient became, of course, no longer practicable. Branches and their leaders now enjoy a certain autonomy, which is however effectively checked by an interlocal organizational structure ensuring overall conformity of both belief and ritual, central administration of membership figures, and centralised control over the recruitment and performance of officers. This control is effected by the Annual Convention at the cult’s headquarters, by the distribution of essential paraphernalia,[44] and by formal examinations concerning the cult’s beliefs and regulations.

Here a creative process must be appreciated in order to see the general significance of this development of a regional organization in the Nzila cult. Imitation is not the crux of this development; Nzila is far more than a mere copy on the model of a Christian church. A failure to appreciate this leads Wilson (1975) to a radical misreading of the cult, and a wrong generalization about the primacy of imitation in the organizational development of sects. From Reynolds’ account of Nzila, Wilson (1975: 92f) wrongly infers that Nzila developed no other congregations than the Mongu one over which Chana presided. Wilson goes on to speculate that

“the abilities and facilities necessary for a more elaborate structure, however, were probably not yet available in Barotseland. Only where an organizational model of another type of movement can be imitated should we expect thaumaturgical responses (such as witchcraft-eradication movements and cults of affliction—WVB) to take on these structural characteristics which we associate with sects in advanced society. Such imitation, combining central direction with branch churches over a wide area, first occurred in this region in the Lumpa church.” (1975: 93f).

Applied to the Lumpa church of Alice Lenshina the explanation in terms of an imitation of the mission as organizational model (Wilson 1975: 94f) is gross and superficial—particularly in view of Roberts’ authoritative statement that “there is very little information of any kind on the internal organization of the Lumpa church—a most important subject which perhaps will never properly be elucidated” (Roberts 1972: 3; see also Ranger 1973, for a critical review of Wilson’s approach, and Van Binsbergen 1977b).

Wilson’s errors are plain. Nzila did evolve branches over a very wide area, and even Reynolds mentions the early Senanga branch. Therefore, the structural requirements for a regional organization were obviously met in Barotseland in the 1940’s. The Bituma and Moya cults, as I will discuss below, show that Nzila was not even the only cult to develop such regional organization. Admittedly, all three cults borrowed part of their idiom from Christian churches (this is clearly what Wilson means by “another type of movement”). Moreover, the fact that Nzila was embedded in a social environment dominated by formal, bureaucratic organizations (including Christian churches) appears to have favoured Nzila’s expansion (see section 4.4 below). Nevertheless, it is a gross error to explain the development of regional organization as merely an imitation of the Christian church model. Such an explanation wholly ignores the creative process by which new organizational solutions have been gradually worked out in Nzila and similar cults. These solutions were primarily determined not by the desire to imitate an outside organizational model, but, as I demonstrate in the following discussion, by the founders’ personalities, their relations with their early followers and the structural characteristics of the cults’ regions.

4.2 Bituma

The development of Bituma is much less of a success story. Simbinga’s first attempt to found a cult failed. Public opinion rejected his claims of prophethood. His first activities (which included the propagation of unusual healing methods and the free distribution of his herd of cattle) were considered the acts of a madman. His early attack on what he considered the impurity of established healers and diviners, including his cousin Kapata, contributed to his public rejection. He left the area. However, when he returned in the late 1930’s, he was in possession of an elaborate collection of paraphernalia,[45] and much more developed views concerning the nature of the afflictioncausing Angel. The affliction henceforth became known as Bituma (“sent him”) or Chituma (“message”), referring to Simbinga’s prophetic commission or (according to other present-day informants) to the Angel, who sends affliction and is himself sent by the High God. Simbinga shifted the area of his activities towards the east: the central and eastern parts of Kaoma district. There he treated his first seven patients, including Kapata. Kapata, and several others among these first followers, only joined Bituma after a career of intense personal crises involving previous initiation into such non-regional cults of affliction as kayongo and mayimbwe. In this respect they were Simbinga’s junior colleagues, rather than his disciples; they had access to experiences similar to those on which Simbinga’s charisma was based.

Simbinga built for himself a small, hut-like shrine (called kreki, “church”) and made his initial followers do the same at their own places of residence. At these shrines healing sessions would be held at irregular intervals, according to the demand from patients. The session would start and conclude with prayer, and be built around songs featuring Simbinga and crucial signs of his prophetic calling: his communication with heaven, and his rising from death:

1 Tukuyako, tukuyako We are going, we are going ku ngonda, ku ngonda to the moon, to the moon (allegedly the first song Simbinga composed when he rose from death, claiming that he had been at the moon). 2 Thangwe ngoma Start beating the drum ngwa Shimbinga here comes Simbinga (a song Simbinga used during ritual dances in the first years). 3 Kombelela nganga wayamukalunga Shimbanda-mwane nganga yetu yele kombelele nganga wayamukalunga Shimbanda-mwane Shimbinga-mwane kombelelelanga wayamukalunga mwane Kapata yele kombelele nganga wayamukalunga mwane (a major song of the later period).

We are going, we are going to the moon, to the moon Start beating the drum here comes Simbinga Clap hands for the cult leader you were in the grave, Great Healer, you are out cult leader, indeed clap hands for the cult leader you were in the grave, Great Healer, we are continuously clapping hands for Simbinga the Great Great one, who were in the grave Kapata indeed we clap hands for the cult leader Great One, you were in the grave

However, despite such distinctive elements as the regional organization centring around a charismatic leader, the theistic interpretation of affliction, and the purity-centred morality, the Bituma sessions in themselves remained, in their outside appearance, very close to the general idiom of the non-regional cults of affliction described above.

The cult took on well in Kaoma district, where in the 1940’s—1960’s a few thousand people were initiated. A more precise number cannot be given since in contrast with the Nzila cult no central records of membership were kept. (Individual leaders did keep records of their own patients, along with a price-list of various treatments available.)

In addition to the seven initial followers, about a dozen more leaders were appointed during Simbinga’s lifetime. These leaders would meet each New Year’s Day at the cult’s Annual Convention around Simbinga’s shrine, where they would hand over part of their patients’ fees and would enjoy a collective meal of white chickens.

In the 1950’s several leaders settled in Lusaka and other line-of-rail towns, where they held sessions for urban migrants from Kaoma district.

Meanwhile old tensions between Simbinga and Kapata (going back to the time of Simbinga’s first appearance as a prophet) became manifest again. They concentrated on the cult’s leadership and on the use of the cult’s funds. Simbinga had married a woman from the eastern part of Kaoma district and enjoyed the loyalty of the Bituma leaders there, several of whom were his close affines. Kapata had settled in the central part of the district, and other Bituma leaders there sided with him. Simbinga’s sudden death during a hunting party (1960) was the occasion for the cult to fall apart. The leaders around Kapata (including those in town who hailed from the central part of the district) accepted his succession to the cult’s leadership. They continued to visit the Annual Convention, now at Kapata’s shrine, and to forward part of their fees to him. Those in the eastern part of the district (and urban leaders hailing from there) have severed their connection with the cult’s organization and have since practised the Bituma ritual as independent, individual cult leaders, without any organization being maintained between them. For them and their patients, Bituma has lost its regional character and has become merely another non-regional cult of affliction among the many prevailing locally. These eastern leaders retain a reminiscence of Simbinga’s original, theistic and universal visions. However, for those of their patients that have been initiated only recently, these associations are virtually lost. Surviving only in the central part of the district and among a few urban leaders, Bituma has elsewhere returned to the substratum of the factional, non-regional cults from which it sprang.

Thus the development of Nzila and Bituma has been very different indeed. We shall presently try to identify some of the factors responsible for this difference.

4.3 Bituma and Nzila compared from within

Let us first look at both cults from within: their internal structure and specific idiom.

As Reynolds rightly observed (1963: 137), during the first two decades of Nzila, Chana succeeded in maintaining control over crucial phases of the ritual. He kept most of his doctors in the position of mere assistants, without autonomous ritual powers. Only Chana’s heir (nominated by Chana himself), and presumably his Senanga deputy, were allowed to exercise ritual powers on their own behalf. As the cults spread further, effective devices were worked out to safeguard the cult’s unity and perpetuation even when it ramified into a number of geographically distant branches. A body of centralized administration, the Annual Convention, explicit regulations, indispensible paraphernalia whose distribution was strictly controlled, all in combination constituted an adequate structure through which the founder’s charisma could be routinized and channelled without becoming dissipated or usurped. These devices constitute clearly recognized principles of legitimation, controlled from headquarters, outside which no cult leader pursuing the Nzila idiom could claim ritual efficacy. The gradual adoption of these devices rendered an ever-increasing specific identity to the Nzila cult. It more and more diverted from the substratum of non-regional cults. Nzila’s interpretational idiom and cultic procedures became so specific that a relapse into the non-regional substratum became increasingly unlikely.

This may have been due not only to the cult’s internal authority structure. Through the very large number of adepts and the participation of nonadepts in the mass services (Reynolds 1963: 136), the Nzila cult must have become something of a generally recognized social institution in the society of Western Zambia. Nzila’s definition and perpetuaton came to be based not just on concepts and actions of leaders and adepts within the cult, but also on public opinion. The latter’s influence is suggested by the one case of a break-away from Nzila cited by Muntemba. There the protagonist had to adopt an idiom explicitly different from Nzila; and even so, Nzila turned out to be so well established that after a temporary decline the population again made a mass demand on its services:

“Two years ago Nzila was thought to be on the decline in the Mulobezi area when a former Nzila healer decided to leave the sect to concentrate on pure faith healing which he thought (and Christian leaders accepted) was more Christian. People went to him from all over Livingstone area and his activities met their needs. Yet now this man’s authority is on the decline as more people turn back to Nzila.” (Muntemba 1972:Ilf)

A very specific ideology and ritual, and a well-defined structure of organization and authority, externally supported by expectations among the general population, produced for Nzila effective checks against the threat of fragmentation and kept the cult from returning to the non-regional phase.

By contrast, Simbinga failed to assert himself as a continually indispensible source of charisma vis-a-vis the other Bituma leaders. Once initiated, the latter could all act independently. Moreover, at least some of them had had experiences similar to those from which Simbinga derived his charisma. Simbinga’s personal status in Bituma was far less exalted than Chana’s in Nzila. The great benefits in terms of prestige, power and income attached to independent leadership along the non-regional model, produced a strong tendency towards fragmentation. Little could counteract this tendency but the leaders’ personal emotional attachment to Simbinga, which ceased to have organizational relevance when he died. Moreover, in order to possess ritual efficacy in the eyes of adepts and of the general population, Bituma leaders were by no means dependent on paraphernalia that could only be obtained from the cult’s headquarters and that, once obtained, would only be regarded as legitimating leadership as long as their owner continued to belong to the cult’s organization. Instead, Bituma leaders freely obtained and augmented their paraphernalia, which were similar not only to Simbinga’s but also to those in use in the non-regional cults of affliction: white clothes, scapulars, fly-switches, bells, etc. And although the cult certainly became an established institution in Kaoma district, in its public manifestations it remained so close to the well-established non-regional cults of affliction, that any external check (in the way of expectations from the general population, concerning ritual and organization) against relapse in the direction of these non-regional cults, could not have been strong.

As a regional cult, Bituma was far less powerfully designed than Nzila. Not so much Bituma’s decline, but its part survival around Kapata seems to pose a problem. Personal ties based on past treatment supported the relationship between Kapata and his loyal leaders after Simbinga’s death. In addition, however, there appear to be more systematic reasons for the survival of Bituma in the central part of Kaoma district, which will become clear as we now turn to a discussion of the structural characteristics of the regions of Bituma and Nzila.

Let us sum up the position. Nzila and Bituma sprang from the same substratum of non-regional cults of affliction, and both were triggered by very similar personal experiences of their founders. What then explains their their very different development? Part of an answer lies in personality differences: Chana was clearly a more creative innovator and organizer than Simbinga. Another part lies in the nature of social relationships between the leaders and their initial followers: in contrast with Chana’s selfconfident and unchallenged control, Simbinga was much more of a primus inter pares and had to compromise with the cultic backgrounds and ambitions of his initial followers. The differences between the cults’ organizational structures and ritual idioms as they were gradually worked out, provide, as a third part of an answer, such explanations as I have just attempted to give in the present section. However, it would appear that these organizational differences in themselves largely derived from the differences in structural characteristics between the Bituma and Nzila regions.

4.4 Nzila’s and Bituma’s Regions Compared

Although the regions of both cults extend into towns outside Western Zambia, the processes that have shaped the cults in the first two decades after the founders’ first appearance as prophets have taken place within Western Zambia. I feel therefore justified in concentrating on that area, and I refrain from discussing the cults’ wider ramifications.

Map 2 outlines the geographical location of the regions of both cults in this area, and indicated the main direction of spread. By comparison with Map I and with the geographical material contained in Davies (1971), a number of striking differences between these two regions can be observed.

Demographically the differences are considerable (Davies 1971:42–47). The Nzila region has, by and large, a higher density of population. In addition to small administrative centres at Kalabo, Senanga and Sesheke (which are perfectly comparable to Kaoma in the centre of the Bituma region), the Nzila region includes the only two significant towns in this area: Mongu and Livingstone.[46] Nzila’s headquarters are near Mongu, whilst the Livingstone area (extending into the commercial timber concessions around Mulobezi) appears in Muntemba’s description as a major growth area of Nzila. By contrast Bituma has been a rural affair, whose present headquarters are located at about 25 kilometres from Kaoma. Admittedly, Bituma does have a few hundred urban adepts: but these are almost exclusively lower-class migrants from Kaoma district who, unable to become stabilized urbanites, keep circulating between town and village, and in many cases go through the Bituma ritual not in town but in the village. The occurrence, in the Mongu and Livingstone-Mulobezi area, of male/female sex ratios of more than 100 (as compared to under 70 in the Bituma region) further bears out the demographic difference.[47]

The demographic situation has important implications for the regional cults. Where population density is low (in the eastern part of Kaoma district it drops under 7 inhabitants per km²), the creation of local congregations and the maintenance of interlocal ties between them poses serious problems in terms of the number of people available, transport, and communications in general. The factional nature of the non-regional cults of affliction is partly explicable in this light. Similar problems play a much smaller role in densely populated areas, and a fortiori in towns.

More important perhaps, the demographic differences reveal a social-structural difference between the situation of the typical Bituma adept in comparison with the typical Nzila adept. Bituma caters for a stagnated peasantry whose experience with urban living may be considerable (due to labour migration), but whose prime identification and source of economic and social security is the village. Modern institutions, and particularly formal, bureaucratic organizations in the political and economic sphere do penetrate from the metropolitan and national centres into the peripheral village scene, and determine the villagers’ lives to an ever increasing extent. Bui these institutions and organizations, with physical outlets at the district centres mainly, are seldom visible at the village scene. Their impact on rural life does not imply that the villagers frequently participate in them or identify with them—quite the contrary. Debarred from substantial economic and political participation in the wider national context, the people in the Bituma region see themselves primarily as part of an economic and political order whose major concerns and transactions refer to local, largely pre-capitalist, historical structures. Bureaucratic organizations set, from a distance, the narrow confines for the local social process but do not play a prominent role in the participants’ subjective life-world. And it is primarily from the latter that we can hope to expect a solution to the problem of differential adoption of religious forms—that is, if we agree to view religion mainly as an interpretative and legitimating structure.

The social-structural situation that is typical of the Bituma region, does obtain in parts of the Nzila region as well. But in the latter’s centres of gravity (Mongu, Livingstone-Mulobezi) the structural context is very different. Here people live in an urban or peri-urban environment, where their life to a large extent consists of participation in economic and political formal, bureaucratic organizations: civil service, industrial and commercial enterprises, schools, medical services, churches, political parties, recreational associations, etc. Understanding of, and identification with, complex patterns of formal task definition and bureaucratically legitimated authority based on legal rules, are essential for both psychological comfort and for economic survival in town and in the peri-urban and rural extensions of urban-based formal organizations. Parkin (1966) has emphasized the potential of urban voluntary organisations to become mechanisms of adaptation to the organizational structure of modern urban life. Nzila is likely to play a role in this respect. But what is more important in the context of the present argument, is that Nzila’s being embedded, from the 1940’s on, in a general urban organizational environment, did provide, in the perception of Nzila’s potential patients and adepts, obvious organizational models and patterns of bureaucratic legitimatidn by reference to which this cult could successfully make the transition from direct to routinized charisma.

Whilst no doubt providing a solution for some of the problems engendered in its adepts by their confrontation with modern urban organizational structures, Nzila copied an organizational model derived from this very structure (and therefore recognizable, and acceptable, to its adepts). This appears to have been an important factor in Nzila’s development into a successful regional cult.

The knowledge of such models and patterns was surely available in the Bituma region as well. But there such a formal organizational structure did not correspond with the peasants’ subjective reality: the organizational micro-structure of their rural society. Nzila developed a regional organizational structure borrowed from and commensurate with the urban society by which it was surrounded. Bituma, after attempts towards regional organization, to a large extent lapsed back into a fragmented, factionalized organizational structure which is not only typical of the pre-existing non-regional cults of affliction, but which is also (in ways I cannot elaborate on here) in line with the fragmented economic and political structure of Nkoya village society.

I am not suggesting, though, that the structure of this rural society was entirely incompatible with formal, interlocal religious organization. Christian missions throughout Western Zambia demonstrate that such religious organizations are viable provided they receive ample political and logistic backing from national and metropolitan centres (a condition obviously not met in the case of Bituma). Missions are important for an understanding of the development of the regional cults in several other respects. They were among the few formal, bureaucratic organizations that were visibly present in the rural areas. In this way they introduced the rural population to a pattern of social organization prevailing in urban contexts. In addition the missions contributed to the ideological shift towards greater prominence of the High God, among the general population of Western Zambia.

Both Simbinga and Chana were exposed to local missions. But whereas Nzila subsequently spread in an area that was very heavily missionized, and where Christianity grew rapidly in the 1940’s, Bituma spread in an area where missionary activity was very limited and in fact declining in the 1940’s (map 2).[48] By consequence, Nzila could benefit much more than Bituma from the missions’ instilling, in the minds of potential adepts, such attitudes and organizational models as are conducive to a regional cult organization.

The relation between Nzila and Christianity was and is extremely tolerant.

Muntemba points out (1972: 12f) that Nzila has developed a considerable degree of symbiosis with Christian churches. Many of the adepts and patients are Christians, even some Nzila leaders are church leaders, and Chana has remained a member of the Seventh Day Adventist Church throughout his life. In the early stages of Bituma, on the contrary, Simbinga denounced membership of Christian churches, claiming that “now we have bur own church ’. In later years this antagonism made way for tolerance and at present the cult is indifferent with regard to its patients’ and adepts’ religious affiliations.

A final point of difference between Nzila and Bituma revolves around the role that ethnicity has played in the development of either cult. Belonging to the Luvale ethnic group, Chana was an Angolan immigrant: one of the many thousands that have moved into the rural areas throughout Western Zambia since the 1920’s (Clay 1945; Colson 1970). Chana utilized Luvale ethnic identity to get his cult off the ground in the first instance. Nzila’s spread closely followed the penetration of Angolan immigrants in the area. However,

“by the later 1940’s the movement had spread to other parts of the Western Province and drew adherents not only from the Luvale (...) but from the Lozi as well as other groups in the Province.” (Muntemba 1972:3)

The mono-ethnic element now appears no longer to play a role in Nzila.

By contrast, Simbinga could hardly rely on ethnic support. He belonged to the Mbunda ethnic group and was an Angolan immigrant like Chana. Unable to find recognition and support among his own people, at the time of his first appearance as a prophet, he turned to the Nkoya who have lived in Kaoma district at least since the last century. The majority of Bituma leaders that Simbinga initiated were Nkoya, Kapata being a notable exception. Instead of utilizing the enormous numerical potential of the Angolan immigrants, as Chana did, Simbinga thus concentrated on a relatively small ethnic group tied to a rather small territory that was increasingly encroached upon by Angolan immigrants, and with no substantial portion of stabilized urbanites among their number. Concentration on the Nkoya is likely to have contributed to the lack of expansion of Bituma. On the other hand, this very concentration made the Bituma cult into a distinctive cultural feature, a sign of Nkoya identity especially among Nkoya migrants in town. Before that time, only the mayimbwe non-regional cult had played that role to some extent.

Ethnicity, or rather local rivalry in an ethnic disguise, also played a role in the later decline of Bituma as a regional cult. The central and eastern parts of Kaoma district form two subdivisions (“Nkoya proper” and “Mashasha”) of the Nkoya ethnic group. Despite frequent intermarriage and intra-rural residential mobility between these two subdivisions, they display mutual antagonism based on a different political, administrative and mission history. Each subdivision has its own senior chief. There are close kin relations between the Mashasha chief and Bituma leaders in the eastern subdivision, and the same is true of the chief in the central subdivision and the Bituma leaders there. The latterday cleavage between Simbinga and Kapata, even though both are non-Nkoya, coincided with one between ethnically-articulated power-blocks in the political structure of the district.

The preceding argument suggests systematic reasons why, with the decline of Bituma, Kapata could yet retain around him some remnants of the cult’s regional organization. Density of population is much higher in the central part of Kaoma district (and particularly north-west of Maoma township, where Kapata’s village is situated) than it is in the eastern part. Moreover, the large Mangango mission (created in the early 1950’s some 15 kilometres north-west of the earlier Mukunkiki site), the annexed hospital, and the secular enterprise it has generated in the way of trading firms and cooperatives, along with the proximity of Kaoma and of Luampa mission, create a structural and ideological milieu more favourable for an interlocal regional cult, than exists in the eastern part of the district.

4.5 Conclusion

My analysis of two recent Zambian regional cults is based, to a large extent, on far from complete, oral-historical data. Even though I can analyse these data against the background of acquaintance with the area, with the non-regional cults of affliction from which the regional cults in question emerged, and with main trends in the religious history of Central Africa in general, the whole argument remains too tentative to approach a theory of the dynamics of regional cults. However, my argument does lead to a few comparative generalizations.

Every regional cult seems to face a dual problem. On the one hand, it has to develop a distinct, specific cultic idiom (in terms of both ideology and ritual). On the other hand, it requires a formal organizational structure which shares out ritual authority sufficiently widely among officials to enable them to cover an entire region, whilst retaining these officials within an authority-legitimizing structure of such effectiveness as to prevent them from either breaking away from the cult or, within its organization, pursuing too deviant an idiom. The dialectical process in which solutions for this problem are worked out with varying success, is determined by a number of factors. In addition to personality traits and interpersonal relationships of the personnel involved, crucial factors appear to be of two kinds. First, the specific nature of a cult’s organizational devices to a large extent determines its chances of survival as a regional cult. Secondly, these organizational devices are partly determined by the structural characteristics which obtain in the geographical area that the cult is to transform into a region. My argument suggests the importance, among these structural characteristics of the region, of demographic patterns and of the occurrence, besides the cult in question,, of other religious and non-religious, formal, interlocal organizations. However, only application to more, and better data from other parts of the world can throw light on the validity of these suggestions.

References

© Wim M J Van Binsbergen 1977

Part Three: Regional Instability: Cult Policy and Competition for Resources

7. Continuity and Policy in Southern Africa’s High God Cult[49]

*by Richard P. Werbner

Introduction: The Definition of Regions

The more a cult spreads across major ethnic and political boundaries, the more its organization may have to be variable and its ritual and symbolism transcultural. A cult in the middle range—neither exclusively parochial nor extremely universalistic—poses special problems. It calls for a theoretical approach that grasps variations in organization and ritual from one rather narrow context to another. Yet it also calls for another order of data and a wider perspective in order to make sense of the cult relations which transcend these narrow contexts. There is no question of alternatives: either to focus on the narrow or the more inclusive context. Both must be regarded, and in relation to each other, because the cult is in some senses both parochial and universalistic. That is why in this essay on a vast yet middle range cult, a High God cult in Southern Africa, fine nuances of cult activity and microhistories are brought into focus in a single perspective along with the connected regional relations and macrohistorical change.

Southern Africa’s most widespread High God cult, the cult of Mwali or Mwari, has its congregations divided in spatially distinct regions, each with its own central staff and oracle and its own adepts and communal shrines. Each region extends over thousands of square miles across ethnic,[50] district and even international boundaries. Yet each is somewhat concentrated in one part or another of the cult’s total domain. Not that any region has a continuous territory of its own: every region is scattered irregularly. Moreover, the congregations of different regions come between each other. A region is a set of many, separated pieces—the locally defined congregations —and it is in its pieces that a region is spatially distinct. There is more to a region, however, than the distribution of its congregations. From region to region there are also overlapping vicinities. The area from which supplicants come to consult a regional oracle as individuals rather than on behalf of congregations is a region’s vicinity. Every region has a wider vicinity around it, beyond its congregations, from which it receives individual supplicants. Moreover, supplicants who come from a single congregation consult oracles in various regions, and these thus have vicinities which overlap. The overlap is selective, however, in that supplicants from one region go to certain others only, and they tend to favour oracles’more in one direction than another.

Regions also have other features of distribution, besides distinctness and selective overlap. Other features appear in the relations between regions and their enclaves and borderlands. The enclaves are territorial communities which do not themselves join any region, although they are virtually surrounded by affiliated congregations. Conversely, the borderlands are the territorial communities which block cult expansion and remain at the outer limit beyond any region. The enclaves tend to be less permanent and fluctuate back and forth from non-affiliation to membership in a region, whereas the borderlands are more enduring as barriers against cult expansion.

Finally, to complete an initial view of the regions’ spatial relations, something must be said about centralization. Not only does each region have at least one regional centre for an oracle, but the regions all together are centralized. Their heartland is where virtually all the oracles are sited, and it is a geographically central location, an inner area of highlands that has special importance for the surrounding lowlands in the rest of the cult’s domain. Moreover, the centralization is such that the heartland itself has a gradient, a decreasing order of oracles from centre to periphery: roughly, the closer an oracle is to the heartland’s centre, the higher is its standing in terms of recognised seniority. The most central oracle is paramount, and in relation to it there are at least two junior lines, each with no more than four oracles under a senior or cardinal oracle. This order relates to permanence in location in that the cardinal oracles, including the paramount one, are relatively fixed; they move their sites the least, if at all.[51]

Both in number and distribution the regions are limited. They are virtually confined to a major ecological zone of savanna or mopane bush veld which runs from Botswana’s central and eastern districts across the south of Zimbabwe and into Mozambique and the Transvaal of South Africa (for this zone see Keay 1959; Newman 1971: 119).[52] Their heartland is the high veld or highlands which cross Botswana and Zimbabwe, and their cardinal oracles are centrally sited in the Matopo Hills within a smaller ecological zone of wooded steppe[53] that is surrounded by the major ecological zone. Not that the regions cover the whole of this major zone: with peripheral, and thus revealing exceptions, only areas of dispersed, though sometimes quite dense settlements with relatively mobile and ethnically heterogeneous populations have congregations. This virtually[54] excludes the territorial communities that are based on large nucleated villages and led by Tswana in the zone’s southwest. There is an oracular cult in the southwest in the Tswapong Hills, but it holds sway only in the smaller, less strongly centralized villages. Moreover, this is a cult of the microcosm,[55] rather than a cult of the macrocosm (see Horton 1975), as is the High God cult.

Within the major ecological zone of savanna, regions compete for congregations and wax and wane in relation to each other. They are unstable, but they do not change arbitrarily. Gradually and continuously, congregations which in themselves vary transfer their allegiance back and forth, from region to region. This transfer takes place during the lifetime of the priest who leads the region,[56] often for several decades or more, though it is all the more frequent at his death. Indeed, after a priest’s death, a long period often passes without a successor in full control of the priest’s office. During such an interregnum, perhaps for as much as a decade or more, the region appears to disintegrate and be partially taken over by others. Sometimes the disintegration lasts; somewhat new regions then emerge, without an increase in the total number of regions.

The cult’s beliefs recognize and allow for such change. Even the site of the most senior oracle, now at Njelele (Snake Eagle),[57] is believed to be one site in a series that began elsewhere among Venda, who are now in the Transvaal to the south.[58] This oracle, like all others, is held to have ultimately come from a foreign site, though its origins were within the cult’s total domain. Any priest may resite his oracle and virtually everyone does. He may move it from cave to cave or hollow boabab tree, on one hill then another. He need not keep to one site throughout his lifetime. He may consult the oracle at a series in succession or at a pair of sites or more, during the same period. From time to time, he may even leave his own centre to live with another priest for months or perhaps longer. Similarly, succession to office is also variable, and people recall only a few successors. A priest’s son is his preferred successor, or alternatively a brother (see also Richards 1942: 55; Daneel 1970: 46). However, these are sometimes disqualified or lacking. The successor may then be another close relative or a newcomer who may claim descent from Njelele’s founder, Njenje, within a rather short patriline.[59] Usually, two generations is the most a priest claims in succession from his region’s founder. However, no direct succession need occur within a region, and an oracle may end with its founder.

Just as a region waxes and wanes so too do its many shrine congregations. The congregations change in span, kind and composition: a chiefdom or any of its parts, one territorial community or several, may become one shrine congregation. The variation is from the kind of congregation which is a whole territorial community (i.e. a chiefdom or a large or small division of it, a section or ward) to the kind which is a local cluster, (i.e. a select set of nearby, roughly contiguous communities and not a whole territorial division). Examples are shown on Map 1: a whole chieftancy (Habangana —D), a section (the south of Musojane—C, wards Cl, C2, C6, C8, CIO), a single ward (C9 of Musojane) and a local cluster (wards B3, B5, B7, B8, B9, of Masunga). Rarely does a congregation last as much as a generation with the same composition in territorial communities. Each congregation expands or contracts in accord with highly localized conditions, and these fluctuate from one small valley to the next (see Werbner 1975). The process is sensitive to local shifts in population, in the support a territorial ruler can mobilize, and in his commitment to a particular priest or even the cult itself.

To understand how congregations vary four aspects must be examined in detail. These are the scale of a congregation’s assets such as its donated fund and offerings, the grade of its shrine, the categories of its staff, and its coverage and composition. Later I take up each aspect in that order. But first lest my view be misunderstood, I must make it plain that cult change is not a process in which the cult’s parts, such as regions or congregations, make a closed adjustment to each other, subject merely to constraints of a political, economic, or ecological kind. There are alternative cults, such as faith-healing and separatist churches; and the High God cult changes in relation to the organization and theology of these alternative, and sometimes opposed, cults.[60] Nor is this a passive relation, or a completely one-sided one. Representatives go back and forth, on an ad hoc basis, between the staff of various regions and the shrines and centres of different cults and also ritual associations, such as diviners’ unions, for example, ‘The Botswana Dingaka Association’ (Mpaphadzi 1975a and b; Modongo 1975) and ‘The Herbalist Association of Africa’. The opposition and rivalry between the cults and ritual associations contributes to rivalry and competition within the High God cult itself. Because of this a priest or would-be leader of a region has a changing series of managerial problems, some of which arise outside of his own cult or region. In due course I discuss these problems along with others that affect policy-making at an oracle. However, a further caution must be given in order for the later discussion to be tightly understood.

This caution has to do with the analysis of strategic action. My argument takes account of long and short-term planning by well-informed cult leaders. Sometimes they hold to the same general orientation for a considerable period of time, and this affects the way they understand the very fine and up-to-date intelligence which they may get even about remote local changes in their regions and elsewhere. But this is not to say that there is some grand and deliberate design and that cult leaders act or choose their policies in accord with it.

Change in Congregations

Every congregation goes through a career of graded shrines in accord with its cult assets and investments. The grades are from a local, to an interlocal, to a regional shrine. Chart I shows the defining features of each grade along with its current distribution on Map 1. Very few communities progress to the higher grades. None remain there for as much as two generations without interruption and return to a lower grade or withdrawal from the cult and want of a shrine. This holds true in the western regions and, I infer, elsewhere too. The congregations which invest more than others attain the higher grades and accumulate the more enduring cult assets—both goods and personnel. But even these are, in certain critical respects, fluid rather than completely fixed. They do not, indeed cannot, anchor a congregation permanently to a particular region, though they may delay a congregation’s decline within the cult. Nevertheless, lest I convey the impression of completely localized fluctuation, I must point out, in advance of my later discussion, that the assets of congregations and regions are checked by the cult’s central organization. Changes in such assets—the slaughter of a black cow, the death of a priest, adept, messenger or affiliated chief and their replacement—should all be reported to a cardinal oracle in the Matapo Hills. Moreover, a cult fine for these changes may be levied centrally.

There are two categories of staff that are perhaps most important for change in a congregation. These are: first, the congregation’s messenger (ntumwa), who is accompanied by one carrier or more, (nsengi)[61] and second, the cult adept (hosana or mwana wa Mwali, child of Mwali). The messenger is regarded as a courier, rather than a delegate with independent authority. His errand is on a territorial ruler’s behalf to communicate and carry out transactions with other congregations, shrines and oracles. To the oracle, he carries offerings along with the congregation’s petitions for rain, fertility and seeds, that are blessed and ritually treated against pests. From the oracle he brings back, besides the seeds, a declaration of Mwali’s response in accord with the state of the land and moral order of the congregation, whether it observes Mwali’s rest days and respects His injunctions such as not to fight and spill blood or kill certain insects, birds, and reptiles; not to sell too much grain or produce, like melons, customarily given free; not to play, especially during the rainy season, the incessant records that blare throughout sales at beer drinks; and, positively, to settle a man’s estate, fill his vacant office, and destroy his abandoned hut after his death. Daneel (1970: 73f) reports an instance of a prominent messenger who got the oracle at Matonjeni and its staff to take a side in a community’s succession dispute. However, an oracle does not regularly declare the winner of that kind of political struggle; and it is not usual for a messenger to seek to have the oracle mediate a dispute over succession to territorial office.

Chart 1: Grades of Shrine

	GRADE I— LOCAL	GRADE II— INTERLOCAL	GRADE III— REGIONAL
Distribution on Map I	A18, A7, B3, C2, Tsamaya	C7	D3, A15
Kalanga Terms	Tanga (Nanga), ‘Ring’	Gota (Daka), ‘Hut’	Dombo (Ntolo), ‘Hill’
Form	Clearing in ring around trees in fallows of local descent line	Clearing in ring; huts	Clearing in ring; more huts; oracle within congregation
Public	Mainly local (a ward or adjacent wards)	Local; some of near communities	Regional representation
Adepts’ Band	Own local band lacking or small	Local band; visitors	Local Band; visitors
Circulation of Messengers	Own staff to and from oracles and other shrines	Own staff; occasional visits by nearby congregations’ staff	Own staff; seasonal visits, at least, by staff of distant congregations
Offerings	Beer, rarely goats	Beer, goats, occasionally cattle	Beer, goats, at times cattle almost annually
Traffic in Supplicants	Absent, except rarely en route to regional shrine	Absent, except rarely en route to regional shrine	Continuous flow of afflicted and other supplicants

Note: Trees of shrines are Mpani (Colophosperma Mopane), Nzeze (Peltophorum Africanum), Nthula (Sclerocarya Caffra).

The messenger is the cult official most bound to a particular congregation, its concerns and interests. He owes his appointment to his standing as an elder within that congregation, though he may also claim that his father was a messenger elsewhere. In some instances, a ward headman himself acts as messenger for his own ward and a set of others. Preferably, the one chosen at a meeting of the territorial ruler’s relatives is a close paternal relative, a seminal brother or parallel cousin, or under special circumstances a close affine. It is crucial for the incorporation in the cult of diverse congregations that a man can hold the messenger’s office irrespective of his ethnic origin or his kinship with a past messenger: Kalanga, Karanga, Venda, Khurutse, Ndebele, Ndau, all hold this office on behalf of their own congregations, whereas the offices of priest and adept are more restricted. As the ethnic composition changes within the cult’s total domain so too may the composition of its messengers change, according to the petitions for affiliation by new congregations.

The messenger has a crucial part to play in the maintenance and growth of a shrine congregation. It is his duty to look after the collection of funds and livestock, to organize the brewing of beer for seasonal dances and shrine feasts, and to arrange for the annual repair of the shrine, if it has huts. On his efforts depends much of the local conversion of private goods into the cult’s assets or its staff’s benefits. A man stands little chance of becoming or remaining a prominent messenger, responsible for a prosperous or higher grade shrine, unless he is himself relatively rich and surrounded by numerous relatives who are ready, along with him, to contribute substantially to the cult. Only when they become, like him, great investors in the cult, can his congregation transcend its boundaries, in various respects, by securing and sustaining the grade of shrine that requires the congregation’s own set or band of adepts.

Adepts, who are at once a major channel for and index of local investment in the cult, are part of its more universalistic staff. Like a priest and unlike a messenger, they may perform anywhere in the cult’s total domain. Indeed, wherever invited, they may give their ritual services. They pray and chant laments to Mwali[62] and sing of the oracles and famous adepts of the past and present. They mime things and memorialize events of broad significance to every congregation in the cult’s domain: they dance with a stick or a wildebeest’s tail in a shrine’s clearing, or ring, sometimes like a fatted cow, an eagle, a game animal or horse, an elder bent with age, a marksman or hunter with a gun, a soldier on military drill with a rifle, or an afflicted victim. After a bout of dancing to three drums, they suffer stylized fits of dramatic, epileptoid possession, when they are ‘tied’ (svungwa) by Mwali, rather than possessed by a lesser or ancestral spirit.[63] They writhe and roll about in the dust, like a person punished by Mwali’s love. Some adepts perform feats: they rub hot coals from the fire on their backs and remain unburnt, as far as I could tell. While their suffering is believed to be on behalf of each congregation as a whole, they are also offered gifts for the sake of individuals-who seek a private blessing. Members of the public at a ritual donate beads, safety pins, or a few pence at a time, and make personal petitions silently or in a whisper to the adepts. For example, an old woman whose daughter’s breasts had run dry not long after childbirth, fixed a safety pin on a female adept’s pleated skirt and whispered an appeal for milk to one of Mwali’s three manifestations: Banyanchaba (The Mistress of Tribes)[64] whose one breast suckles all, mankind, animals, and birds.[65]

Such universalistic aspects of an adept’s status can be understood more, once highly particularistic criteria for succession, recruitment, and seniority are grasped. This will also clarify a crucial fact and its implications: that an adept’s status by comparison to a messenger’s is restricted ethnically and more under a priest’s control. A cult belief is that Mwali chooses each adept to succeed a close relative, and makes His choice known through possession after a more or less severe affliction. The symptoms in the roughly sixty cases which I recorded were severe anxiety, persisting headaches, attacks of hysteria, swelling of the elbows and aches all over the body, constant fatigue and weakness, crippling illness, and infertility. Most adepts claim a predecessor who is an ancestor no more than one generation removed from the oldest living generation and at most three generations from Ego. The relationship claimed is usually through the adept’s mother or father’s mother, but the predecessor is not the adept’s mother herself. In the cases I recorded only fourteen out of sixty-two adepts claimed to succeed a patrilineal ancestor. These cases cover the adepts who resided or performed during 1964–65 in the former Tati Reserve, a part of the cult’s domain shown on Map I and all within about 30 miles of a priest’s home and oracle. (For the distribution in detail, see Table I: Succession of Adepts.) Kalanga say that a person must be of the ‘stock’ (ludzi) of adepts to become an hosana of Mwali. This is a restriction which compels all adepts, irrespective of their own ethnic identities, to claim a predecessor, within living memory, with a Kalanga or Venda ethnic identity. All adepts thus claim common ethnic origins. Significantly, whether to recognise this and a person’s other qualifications, such as Mwali’s choice through possession, is a matter for the priest and his oracle to decide; and I will consider this shortly. The point that must be stressed here is this: on the basis of a claim to a common ethnic origin, ethnically different people can be admitted to the adept’s status. The cult is open to expansion and it includes ethnic newcomers as adepts. Indeed, matrilateral succession, the prevailing mode, fits, and, even more, fosters such inclusion through intermarriage among people from the various patrilineal societies in the cult’s domain. Above all, it is the penetration of newcomers’ communities by their predecessors in the cult that the succession and distribution of adepts represents; for example, the penetration of Ndebele or Khurutse by Kalanga and of them, in turn, by Venda.

Table 1: Succession of Adepts

Deceased Predecessor’s Relationship to Ego	Ego’s Sex
	Male	Female
father’s father	5	5
father’s father’s sister	0	3
father’s father’s father	0	1
father’s mother	1	8
father’s mother’s sister	1	0
father’s mother’s brother	1	1
father’s mother’s mother	0	1
mother’s father	4	4
mother’s mother	1	9
mother’s mother’s sister	1	0
mother’s mother’s father	0	1
mother’s father’s mother	1	3
mother’s father’s father	1	0
mother’s father’s father’s sister	0	1
mother’s mother’s mother	1	3
unknown	1	4
Total adepts	18	44

Table 2: Recruitment of Junior Adepts

Immediate Senior’s Relationship to Ego	Ego’s Age and Status at Recruitment
	Unmarried Boy	Unmarried Man	Unmarried Girl	Married Woman
son	0	0	0	0
daughter	0	0	0	1
brother	0	0	0	1
sister	1	0	3	0
father	2	0	1	0
mother	2	0	1	0
father’s sister	1	1	2	0
father’s sister’s daughter	0	0	1	0
father’s brother’s daughter	0	0	1	0
father’s brother’s son	1	0	0	0
father’s mother’s brother’s daughter	0	0	2	0
mother’s sister	0	1	6	0
mother’s sister’s son	0	0	1	0
mother’s sister’s daughter	0	0	1	4
mother’s brother’s daughter	0	0	1	1
mother’s mother’s sister	1	0	0	0
mother’s mother’s brother’s daughter	0	1	0	0
unknown	0	2	1	2
Total	8	6	21	12

Table 3: Recruitment of Senior Adepts

Age and Status at Recruitment
Unmarried Boy	Unmarried Man	Unmarried Girl	Married Woman
3	1	5	6

Two further points must be made about restricted succession and its implications for the penetration of groups and communities. The first and immediate point illuminates how adepts and their kin connect within and across congregations. The second point needs to be considered later, since it concerns the extension of the priest’s own connections, and requires a detailed discussion of how the adepts recruited are a selection who have close kin or origins in the priest’s own home area.

Adepts enter the cult by joining bands or sets of relatives, in a series. Seniority is recognised thus: the earlier in the series, the more senior the adept, irrespective of kinship or other statuses. Usually, a recruit’s closest living relative in the cult, listed as the immediate senior in Table 2, is within the range of the recruit’s first cousins and their parents (i.e. at least 38 out of the 47 junior adepts; see Table 2 for further details). Recruitment follows from one rather close relative to another. The set of links between all of a band’s adepts has a greater range. A band must extend beyond the range of any individual link. Otherwise, little or none of the diversity within a congregation can be represented in the composition of its adepts. Among Kalanga, at least, each congregation—indeed, each component ward— includes sets of relatives, related in a variety of ways which, of course, change over time. Moreover, no single descent group is in the overwhelming majority or is even the largest in a ward, for very long (see Werbner 1971a, 1975). Thus to fit the highly variable and diverse composition of congregations, a band has to extend like a chain of short but cumulative links. The more the diverse sets of relatives meet the entry qualifications and—as will be seen immediately—the entry costs, the more comprehensive a band becomes. Similarly, as adepts change their residence, so too does a band change its local extension and thus the extension of ritual collaboration across congregations. Distance is an important factor: usually the closer the adepts remain, the more regularly do they return to their original congregation’s shrine and rejoin a band during a performance. Genealogy I illustrates, for 1964–65, one band’s series of adepts, shown in Roman numerals, and its current extension in terms of kinship and wards, in Arabic numerals. This band belongs to the interlocal shrine located in ward B7 of Musojane Chiefdom; it began in the early 1950’s, when the daughter of its first messenger became an adept.

It is usually a messenger who establishes his congregation’s band by sponsoring one of his dependants (his mother, sister, wife, son or daughter) to become the band’s founder and most senior adept. His kin and affines then follow his lead, and sponsor their own dependants. The cost is considerable. Not much need be spent at first, while reaching a decision to sponsor a candidate: less than a pound or two in diviners’ fees, paid by a close relative (a father, mother’s brother, husband, or other elder with jural responsibility for an afflicted person) or, rarely, the candidate himself. This or another relative makes the next, larger donations once a candidate gets endorsed and urged to ‘go forward’ at seances. The close relative who becomes the sponsor takes the candidate forward to the priest’s oracle, preferably accompanied by an immediate senior or some other related adept. For the candidate’s treatment and initiation the priest receives an obligatory gift from the sponsor (fupa, to give for ritual services), often as much as ten pounds. The priest does not charge a fee. Payment must also be made for the adept’s costume (which may become an heirloom), a kilt, preferably dark like a rain cloud, leg rattles and strings of many, brightly coloured beads. From the oracle’s supplies, however, the priest allows the adept a black sash, limbo, a badge of recognition which is given at the oracle’s command. Besides his gift, the sponsor may also namata ‘make an offering to supplicate for a favour’ at the oracle and provide beer and perhaps a goat for slaughter. The sponsor then incurs a further obligation, to give a black cow to be dedicated to Mwali for the adept’s sake and, eventually, to be sacrificed and eaten mainly by the priest and adepts, though anyone from messenger to anthropologist may be honoured with a thigh. Such giving is not for once only. The same man may sponsor as many as two or three adepts. Repeated donations of stock for sacrifice are expected over a period of years, especially from messengers and territorial rulers whose generosity sets the pace for the growth or decline of a congregation and its band.

Some of this expenditure circulates or remains within a congregation rather than concentrates in the priest’s hands. Members of the congregation themselves thus have a stake in the material as well as intangible returns from investment in the cult. Within three or four years, depending on an adept’s participation and performances at home and elsewhere, the adept may earn back for himself (or herself) a sponsor’s initial cash expenditure. This income, which continues as profit, is in small sums from the public at performances, plus about a pound a year from the priest himself for an active adept. Similarly, the priest usually returns to a messenger a pound or more from the congregation’s donated fund, for expenses en route to the oracle. To a major donor, an important messenger or ruler, he sometimes gives a larger sum and goods. For example, for about ten years, while the southwestern priest, Ntogwa, kept an oracle in ward D15 of Ramokate Chiefdom, and until the decline of its regional shrine late in the 1960’s, the priest annually gave the ward’s headman about five pounds and a big sack of sugar.

Nevertheless, their stakes in the circulation of cult goods and investments are such that congregations may still shift from one region or priest to another. The stakes are not of a kind to stabilize a congregation’s affiliation, irrespective of the succession of messengers and territorial rulers and irrespective of adepts’ personal preferences. The messenger who escalates the pace of donations and thus the importance of a shrine achieves a personal stake, materially and, in the spread of adepts as a band among his close kin and affines, organisationally also. This increasing personal stake attaches him more and more to a particular priest. However, the greater the stake one messenger achieves, the greater the hurdle for his successor. Primarily, this is due to change in the composition of the congregation, which also alters the set of relatives around the successor in that generation or the next. Despite such change, the successor is expected to begin with payments in accord with the established grade of the shrine, including a fine for his predecessor’s death. He cannot gradually build up investment as his predecessor did. A decline in contributions and offerings readily becomes a matter for recrimination between the priest, messenger, and members of the congregation, until exasperated the priest may announce that he has broken with them and they may return to a former oracle of theirs.

Thus a messenger often begins by continuing with his predecessor’s priest and then shifts to another one or simply rests so that his wards, left without any messenger, may cease to be one congregation in the cult. Similar careers hold for adepts. All are free, as befits the universalistic aspect of their status, to serve the priest of their choice, though sets of adepts usually choose collectively. Some remain attached to their initial priest, despite their congregation’s shifts. Others become dissatisfied, for various reasons, and press their congregation to change priest along with them. Adepts thus hasten or retard a shift. They are not a barrier against instability in regions and their congregations. On the contrary, they are an essential force in the progress of a congregation through the grades of shrine; that is, in the process by which a congregation defines itself through a local shrine, transcends its own boundaries through an interlocal one, and becomes a major focus of ritual collaboration through a regional one.

Here a point that must be stressed is this: ritual collaboration is no deus ex machina, and does not resolve or redress conflict between communities or congregations. Admittedly, ‘to settle down the country’ (gadzikanya shango), to make peace with and on the land, is an avowed aim of adepts’ performances; and though quarrels do sometimes break out, it is forbidden to disturb the peace of a shrine. However, there is no necessity that opposed communities, such as chiefdoms which compete for territory, collaborate in ritual. Such communities have an alternative to ritual collaboration: avoidance. A region is a set of scattered congregations, and opposed communities often resort to membership in alternative regions in order to avoid ritual collaboration with each other. Nevertheless, occasions for ritual collaboration do provide opportunities—sometimes even the necessity—for testing and proving the relations between competitors. This is so particularly in those ambiguous border areas where people from different territorial communities come to be mixed together and interspersed (see Werbner 1975: 115–17). For example, in a valley which the chiefdoms of Habangana and Musojane (D and C on Maps I and 2) disputed, people from each chiefdom were so interspersed that no clear boundary divided them, until Habangana expanded across the valley at Musojane’s expense. Ritual action as a phase in the competition was a test of strength. Irrespective of their affiliation to one chiefdom or the other, people in the valley joined together in one congregation and thus proved that the valley was not divided within the cult. Even more, they anticipated the gain by one chiefdom, Habangana: they affiliated their congregation to the region centred on that chiefdom and where the chiefdom as a whole belonged. Thus a clear anticipatory definition was given through ritual collaboration across the valley, before the territorial and political ambiguity was resolved. Moreover, the congregations of one chiefdom continued to avoid those of the other for several decades until after a further adjustment between the chiefdoms took place, f*

The Development of a Region and the Limits of its Expansion

To see change in congregations within a wider context, it is essential to consider at least one region and its relations in detail. The region I know best is the southwestern, formerly under the late priest Ntogwa Matafeni Dube, and I will therefore concentrate my discussion on this region. The conditions which are perhaps most crucial for the southwestern region relate to emigration and territorial encroachment. During Ntogwa’s long career as priest, high rates of individual mobility (Werbner 1975: 99) have swelled into great tides of emigration towards the west and south. Partly, this has been due to a higher rate of human and animal population increase with resulting greater pressure on land and, in places, severe scarcity of land. Fundamentally, it has been due to the expropriation of the central highlands of Botswana and Zimbabwe for European ranches. These divided or displaced numerous chiefdoms and led to the founding of others, such as Habangana and Musojane. Most importantly, the land available to their people was restricted, both in Zimbabwe and in Botswana (Werbner 1971a and forthcoming).

Expansion from one chiefdom, Habangana, and migration from it to various others as much as a hundred miles away provided a main stem around which the southwestern region developed. As a youth the priest himself, Ntogwa, had been an immigrant from Zimbabwe, though after settling in Habangana Chiefdom he came to be known as ‘a man of Habangana’. Between 1914 and 1940, the Protectorate administration repeatedly had to allow this chiefdom to take over territory along with people from its weaker neighbour, Musojane Chiefdom. The region’s staff and oracle were thus drawn along with the chiefdom into problems of territorial encroachment and expansion. The expansion and emigration enabled Ntogwa to extend his cult connections widely yet retain some control over them through the selection of adepts with close kin or origins in his home chiefdom. However, a halt to Habangana’s expansion, for various reasons, became a threat to the further development of Ntogwa’s region. This he tried to overcome late in his career, somewhat unsuccessfully, through a more direct commitment to other areas of the region. What he did not do, though his children did, was marry anyone from the chiefdoms in the region’s heartland (shown in Map 1) other than Habangana. In his old age he took his other wives from the periphery of the region in the west (from Tonota, Chadibe, and Mathaipngwane in the Central District, where he also established a hamlet) just as earlier he had married eastern wives (from the areas of his youth in Zimbabwe). Throughout the region’s heartland, he spread his points of access strategically, in two ways. First, he distributed his wives and children in hamlets at the frontiers of each of the heartland’s chiefdoms, and his oracle at a site accessible to a railway station. Second, he established an alternative oracle with its own regional shrine in the only chiefdom that had much room for immigration or more livestock, i.e. Habangana’s greatest rival, Ramokate Chiefdom (A on Map 1). However, this strategic placement and wide division of his family brought its own pressing problems, mainly due to local instability and succession disputes (see above page ); and after less than a decade, he retreated to Habangana, with his entire family except for his far western wives and children. Until his death, four or five years later, he continued to devote himself much more to the western areas than any others, and admitted new adepts from there almost exclusively.

This account of Ntogwa’s career and his region’s development raises a tough question, which calls for further research. What sets the limits to a region’s expansion and extension? The most obvious part of the answer relates to competition between regions and to the priest’s own perceptions. Much of Ntogwa’s region was defined relatively in a competition for congregations between his region and at least one other, mainly the north region of Manyangwa II. Significantly, this competition was absent in the far southwest, which may have been a further reason for Ntogwa’s preoccupation with this area late in his life: he largely withdrew from the main areas of established competition. Yet it was Ntogwa himself who perceived quite early in his career that there had to be another region besides his in the whole of the west. Within his region alone he could not manage the total area covered by several regions now. Using a customary formula, Ntogwa reported to the cardinal oracle at Njelele that “All these people are too heavy for me.’ He requested another priest, ‘Give me another to help me.’ Ntogwa then insured that this other was a protege of his own and closely bound to him. He trained his protege, apparently a son of Manyangwa I; brought him for confirmation at Njelele; installed him in the north; took Manyangwa H’s sister and gave him his own daughter in marriage; and continued to visit and help him for long periods. When I asked this daughter (Manyangwa Il’s wife) about the interspersion of the regions’ congregations, she insisted that the two priests never fought over the country. She said,

‘It is the black people themselves who did this. When they complained against father [Ntogwa], they went to Manyangwa and when they complained against him, they came here. We have our home in Botswana; Manyangwa has his in [Zimbabwe] ‘Rhodesia’. But if people want to jump over from one to the other, it is alright; there is no fault.’

The mutual definition of the regions is, thus, in part a product of co-operation and adjustment between the priests in response to their recognized and shared interests in the management of competition.

The evidence from the southwestern region suggests another part of the answer to the question of what sets the limits to a region. Ethnic links or historic ties and past affiliations are inadequate and are not in themselves enough of a basis for extending a cult region far into communities of large, nucleated villages. The inadequacy and limitation persists even when the cult’s staff try to enhance these links and further various interests through their personal contacts and through heavy investments of resources. I must stress that this limitation by large, nucleated villages is not simply a matter of a lack of cultural familiarity with the cult. Kalanga themselves spread far and wide among ethnically different people and beyond the cult s domain into the borderlands’ nucleated villages. Moreover, they have done so for at least one century and perhaps several. However, Kalanga have not managed to establish cult congregations in the borderlands, with perhaps one peripheral exception, although as individual supplicants they come to the oracles[66] even from Serowe, the Central District’s sprawling capital.

My immediate point about the borderlands requires that I discuss briefly the apparent exception, which I mentioned. This is on the periphery of the southwestern region at Tonota, the northernmost of the large nucleated villages. The messenger there, Radipitsi, holds the most senior title among the Khurutse, who are Tswana-speaking. But he does not hold the major political office or chiefship that once went with his senior title. Radipitsi claims direct descent from a Chief Rauwe who sent cult messengers to an oracle, while he lived among Kalanga. Under the colonial Tati Company this chief was paramount over Kalanga in the region’s heartland from about 1898 until he and other Khurutse were compelled to withdraw south to Tonota in 1913 (Werbner 1971a: 33f). However, Radipitsi’s interest in the north is not a matter of past history only. Recently, he approached the head of the Tati Land Board (his Khurutse cousin Chief Ramokate), which is now responsible for much of the Company’s former land in the north. He expressed his desire to return north soon in order to settle in State Lands near Ntogwa’s daughter, Galani, at Thema Shanga. In the meantime, as messenger, Radipitsi continues to assert a political claim, seniority, and a connection with the north and the region’s heartland. He acts as messenger, however, on behalf of villagers from his own immediate locality. It is a distinct, and somewhat independent part, rather than the big village as a whole or its chief (the Senior Sub-Tribal Authority), which participates in the southwestern region. In Radipitsi’s small and exceptional congregation the southwestern region has reached an outer limit of the cult’s domain.

One further point about expansion must be made here. Just as ethnic and historic ties are inadequate as a basis for the extension of a region so too in the cult’s history ethnic differences in themselves have not been a barrier to expansion. At a moment in time, however, certain enclaves may be defined ethnically, but this may not be permanent. An example of such an enclave is in the heartland of the southwestern region; it is the whole chiefdom of Moroka (E on Map 1) which has no adepts or shrine, though the regional shrine in Habangana is virtually on its borders. The founders of Moroka Chiefdom and their descendants along with a main body of Barolong immigrants who arrived after 1915 have largely remained ethnically separate from and, in some respects, opposed to the Kalanga among whom they live. The more the ethnic separation is modified through marriage, the more likely is the incorporation (or re-incorporation) of this area into the cult, a process already well advanced, as shown by the marriage of the priest’s son Vumbu to a Barolong wife.

To complete the answer to the question about the limits to cult or regional expansion and extension calls for a comparative analysis beyond this essay’s scope, indeed beyond the data now available to me. A fuller answer will be reached when more is known about religious organization in relation to basic aspects of competition in big, nucleated villages, such as how their component communities’ boundaries fluctuate and, how their ruling elites vary along with concentrations of power and wealth.

The Priesthood and Its Managerial Problems

How does a priest cope with the continual and sporadic changes in his region’s many congregations? How does he, at the same time, also accumulate assets and control adepts and their connections? The cult funds and offerings in a region generate varying opportunities to accumulate assets. Every household’s adults are expected to contribute a few pence a year, voluntarily, to their congregation’s fund which is brought to the priest to be offered (namata) at his oracle. However, actual contributions are made more selectively.[67] As a congregation waxes and wanes so does its fund and other offerings such as grain for beer or black cows or goats for slaughter at shrine feasts. It may happen that as many as ten or fifteen head of cattle are kept in trust by a priest for a particular community, (Daneel 1970: 41; Richards 1942: 55). Nevertheless, rarely are even a few head kept in trust within the congregation for long; they are sacrificed, die naturally, or are embezzled.[68] Similarly, every region ought to have a least one shrine at which a granary is kept full of grain for passers-by or supplicants at the oracle. However, only one of the western regions, Manyangwa’s, now maintains such a public granary?[69] By far the greatest accumulation of assets in cattle, cash, and dependants is a personal one that belongs to the priest himself.

At the peak of his career, a priest may be rich in capital, the head of a very large family, and in charge of a great turn-over in cult transactions. For example, the southwestern priest, Ntogwa, kept ten wives and their children, for whom he paid heavily in bridewealth, a personal man-servant, and many retainers as herdsmen. At the time of his death in his eighties in 1972, he left an estate of hundreds of head of cattle, numerous two-hundred pound sacks of grain, a substantial bank account and large sums in small change, a donkey cart, a ruined tractor, and a greater assortment of consumer goods than most Kalanga could afford. His gross turn-over in the cult can be estimated, roughly, from the extent of his region during 1964–65 (see Map 4): at least seven communities in Zimbabwe[70] and fourteen in Botswana,[71] plus about half this total where his adepts lived in wards not affiliated to his region. All together this represents a considerable gross turn-over from a catchment area well over a hundred miles in length. Yet to keep many dependants and be very rich, a priest needs a source of income besides messengers’ congregational transactions and sponsors’ gifts for initiation. In these his gross turn-over may be great yet his net income low or inadequate for all his dependants, since he must return a share to the cult’s staff.

A priest has a more lucrative source for his net income from the cult: a great stream of gifts, often as much as £10 a visit, from afflicted supplicants’, who, throughout the year, consult the oracle about individual affairs, rather than seasonally about communal ones. However, the individual supplicants’ traffic, so crucial for a priest’s accumulation of great wealth, is not and cannot be divorced from the congregational traffic: one sustains the other. A priest must manage both together, or risk a decline in both. This is so, in part, because the priest gets funds from the supplicant’s traffic which he can use, as he sees fit, to subsidise transactions with messengers and their congregations. For example, to continue to give five pounds to an important messenger, despite a shortfall or decline in his congregation’s contributions (i.e. the messenger from D15, mentioned on page 196). Yet perhaps even more fundamental is the significance that one traffic has for the other in organisational terms. Each provides personnel and contacts for the other: the seasonal and more fixed for the continual but somewhat more ad hoc; and vice versa. However, the more far reaching traffic, the supplicants; seems to represent a region’s potential and may set limits for its variation. This traffic comes from well beyond a priest’s current region, and brings to it some of its future staff. Thus members of wards about to form a congregation or would-be-messengers may first come to an oracle as individual pilgrims; or a messenger from a congregation that is established in one region may throw out a feeler for an alternative, while on a personal visit to another oracle.

This traffic, although somewhat more ad hoc, is directed, nevertheless, through regular channels. A supplicant, referred to the oracle by one diviner, may be referred back by it to another diviner elsewhere for further help and treatment; first, of course, the oracle may pronounce on the cause of the supplicant’s affliction or misfortune. There is, thus, a two-way traffic between diviners and an oracle, directed through specific referrals and carried on with authoritative advice and counselling. The diviners themselves may visit the oracle. However, the priest also maintains his contact with them and first-hand, up-to-date information on local conditions, through the frequent circuits that he makes round his region’s congregations and beyond them. For example, even in very old age, Ntogwa visited most of his congregations at least once a year, and annually danced, for payment, at places hundreds of miles apart, from Botswana’s capital to Bulawayo in Zimbabwe. In the past, through such circuits, a priest carried out a further role: public witchfinder for a congregation. Kalanga elders recall occasions when a priest such as the northwest’s, Njenjema, while on a visit to a community, spoke to an apparently empty hut from the front outside it and got a response from within in the oracle’s voice. This accused someone by name or clan and ordered the person to cease the practice of sorcery. I observed no such occasion; nor could I get adequate confirmation of claims, which I doubt, that public witchfinding no longer happens in the cult. Consequently, some important managerial problems in a priest’s career must await further study, and can merely be indicated here.

However, my observations of these circuits do enable me to reach certain conclusions. Perhaps even more than the major property from a priest’s estate, which is inherited rather straightforwardly by rule and practice (the bulk for a senior wife’s first-born son, in effect as a trustee for the family as a whole), a priest’s successor needs to take over the cult sphere which the late priest managed through frequent circuits and long journeys. This is a highly personal sphere. It demands that a successor acquire intangibles of reputation, i.e. for authoritative knowledge of the herbs needed for patients or to treat seeds, ritually, against pests. It also demands, of course, organisational expertise—effective first-hand knowledge of and contact with patients, diviners, and the other cult officials spread over great distances. Yet reputation and organisational expertise, unlike inherited property, have to be gained gradually. Admittedly, they may be gained in good measure by a potential successor, before a priest dies, through the potential successor’s service as his closest associate and companion on the circuits. But this gain hardly eliminates delays and disputes about the transmission of a priest’s assets or about succession to his office. Indeed, such a gain may be the crux of delays and disputes, because a recognised heir to a priest’s main property may have to come to terms with another potential successor, the closest associate and companion of the late priest.

Ntogwa’s case illustrates the opposition between inheritance and personal achievement, and its consequences. For various reasons, Ntogwa relied more heavily, for cult purposes, on two of his daughters, Nlahliwe and Galani, than on his senior son and main heir, Vumbu. Nlahliwe was his first wife’s first-born who married and then separated from the northern region’s priest. She went with her father on trips to the central oracles. Galani was his third wife’s first-born, and she accompanied and helped her father on most of his other trips. She regularly took charge, on his behalf, of performances at shrines, and she became, after him, the region’s most famous dancer: indeed, her father often sang, at shrines, ‘Galani is the senior’ (i.e. of the hosana adepts). In contrast Vumbu, an accomplished dancer though not an hosana,[72] almost never accompanied his father. He knew little about his father’s herbs, and he remained away working on the railway in Bulawayo, except for brief visits to his wife and children in his father’s hamlet. After Ntogwa’s death Vumbu was chosen to succeed his father by many of the region’s messengers and their rulers, led by Chief Habangana from Ntogwa’s home chiefdom. However, some important messengers and rulers did not go to the selection meeting, notably Chief Ramokate from the chiefdom which rivals Habangana in size and other respects. For confirmation and installation, Vumbu was sent, along with Habangana’s messenger and porters, to the cardinal oracle at Dula in Zimbabwe. He returned triumphantly with a black sash, given by the cardinal oracle as a badge of recognition; but when he addressed his own region’s oracle at Habangana Chiefdom, in the presence of a great assembly of messengers, he failed to get a response from Mwali’s Voice. On his visit to Dula, neither of his prominent sisters, Nlahliwe and Galani, accompanied him: a quarrel and bitter recriminations between him and Galani had already become public knowledge. Moreover, a member of the family, who had good reason to be certain, assured me that the Voice would not be heard so long as they continued to quarrel and not make peace with each other.

In the meantime, Galani has been increasing her prominence in the region through her continued circuits around it and by establishing a new, personal base of her own. After a year’s pause to mark her father’s death, she resumed charge of performances at some Kalanga and Khurutse shrines[73] to the southwest, in the Central District. To the southeast, she chose a site suitable for a new oracle. She left the former Tati Reserve and her father’s hamlet in Habangana Chiefdom, and moved south to an area of growing settlement—indeed, the region’s area of greatest growth in settlement— where she built her own hamlet at Thema Shanga. This site is next to a rocky kopje (a granite knoll) with Bushman cave paintings and thus, as she told a confidante, especially suited to impress supplicants with its ancient significance. Moreover, its location in the State Lands gives her a contemporary vantage ground that is crucially comparable to Habangana Chiefdom at the time it gained the regional oracle around 1914. Just as Habangana had then, so too the territory around her site has now a very recently enhanced political significance, besides room for immigration and expansion of settlement (presently lacking at Habangana). Then Habangana was, in various respects, foremost among the set of newly independent Kalanga chiefdoms in the Tati Reserve (see Werbner 1971a). Similarly, the new political community on the State Lands, formed under one territorial ruler[74] by the Government, is potentially the most prominent in the region. It covers a large part of the State Lands (including Thema Shanga) that were acquired by the nation in 1969 from a colonial company; and it exceeds Habangana in size and perhaps already in population also. At Thema Shanga, Galani thus has a base for the kind of shrine which only a major political community on the wax can sustain as the centre of a whole region. However, it is essential for such a shrine that the territorial ruler, his messenger, and other local followers make a strong commitment of support along with considerable investment in the cult. Galani, for various reasons, requires more time to secure such support, if she can.

Primarily, the unresolved problem is whether she can turn her personal base in the supplicants’ traffic into a springboard sufficient to reach command of the congregational traffic. Besides her great organisational expertise, she has an important advantage due to her widespread reputation for mastery of her father’s herbs, which she is said to have taken in great sacks to her own hamlet. So far she has been able to attract to her for treatment not only supplicants but would-be-adepts also. Since her father’s death she has continued to initiate some from the region’s periphery in the Central District, though not from its old heartland to the east in the former Tati Reserve. These and some other adepts closely related to her call her their leader, ntungamili, and insist that she must be recognised as regional priest. Along with them a few rulers and messengers from the Central District, most prominently the Khurutse Radipitsi from Tonota, have publically taken her side and invited her to take charge of performances at shrines. Others from the region’s heartland, including the Khurutse Chief Ramokate, have aided and encouraged her without such public recognition. But most have continued to favour her brother Vumbu or begun to throw out feelers for an alternative, affiliation to the northern region.

Such opposition within the region’s heartland, led by most of its messengers and rulers, has forced Galani to develop her personal base more with the help of separatist churches, particularly the one heavily supported by Radipitsi and other Khurutse. Through the churches she has access to local communities and organised bodies and thus an alternative to those represented by established messengers. In return, she has had to make a personal commitment to the separatists that is greater than her father’s was. Even her father sent and received sufferers who went in a two-way traffic to and from separatists. He attended their services, and included Christian prayers in shrine rituals. The further step Galani herself has had to make is the assumption of a separatist church leader’s role along with public conversion and affirmation of church membership. It is a step she told me she was forced to take to protect herself ritually against sorcery from closely related enemies. However, this step, in turn, has put another part of her supplicants’ traffic at risk and provoked further opposition, this time from some of her father’s other collaborators who are the separatists’ own opponents and lead a diviners’ union. In June 1975, they staged an encounter with her, according to their president’s statement in the Botswana Daily News, ‘Because she used a Christian church as a cover up, (Mpaphadzi 1975)’.[75] Their intention was ‘to set aside Galani who self-installed herself as our god’s chief minister’ (ibid.). However, the encounter was indecisive: Galani has retained a personal base in the cult, but neither she nor any one else has succeeded Ntogwa and taken charge of a region as a whole.

Every interregnum has its distinctive problems. Yet this cult phase is one among others that are not only connected with it but must also respond to certain structural constraints which continue from phase to phase and affect the careers of a series of priests.

Hierarchy and the Making of Oracular Policy

So far, with regard to areas of dispersed settlement and small competing political communities, I have examined the regional instability, relative impermanence of oracles, variability in priestly succession, and fluctuations in congregations and their staff. I must now show that the regions do remain constant, in a sense, and further consider the relations between priests and oracles in order to clarify the making of oracular policy.

Regions do not proliferate in the High God cult. My observations of the western regions over a fifteen year period, participants’ memories, and documentary sources from 1896 onwards lead me to this conclusion. In response to population increase, migration, war, the dislocation of individuals and communities under colonial rule, and other major political change, the regions altered vis-a-vis each other. Their relative coverage and definition changed, but not their maximum number. Chart II shows that no more than three regional oracles have been in the west for at least the past eighty years, throughout a succession of priests and oracle sites. I do not have equally good evidence for other regions.[76] The available evidence shows the continuity of three cardinal oracles: Njelele (Cockin 1879; Richards 1942: 55; during an interregnum, Gelfand 1966: 37), Matonjeni (Cockin‘ibid., Daneel 1970: 40) Dula/Umkombo (Cockin 1879; under Maswabi III, Gelfand 1966: 35f; see also Daneel 1970: 47 on the priestess Kombo at Matonjeni).

Chart 2: Western Regions of the High God Cult

REGION	DATES	PLACE	PRIEST’S NAME AND RELATIONSHIP	DOCUMENTARY SOURCE
1. SOUTHWEST	Pre-1896	(a) Near Mangwe; Bango Chiefdom?	Zviposa. Child of Njenje. Zimbabwe.	(On messenger Hobani, Burnham & Armstrong 1896, in Ranger 1967: 187)
	Early 1900’s	INTERREGNUM
	ca.1914	(b) Jackalas 1; southwest of Habangana Chiefdom, Botswana.	Fulele ‘Cloud (Rainless).. Child of Njenje.	Plumtree N.C. 1914: N 3/31/1-5
	Early 1920’s	(c) Nimakwali; southwest of (b). (d) Luswingo; southwest of (c), Habangana Chiefdom, Botswana.	Ntogwa, ‘Taken (by God)’. Daughter’s son of Njenje.
	Late 1950’s	(e) Tamba Palale; West of (d) Ramokate Chiefdom, Botswana
	1972 to present	INTERREGNUM Preliminaries at Thema Shanga Southwest of (d)		Mpaphadzi 1975 a & b; Modongo 1975.
2. NORTHWEST	1876	(a) Maitengwe River; Meng we Chiefdom?	? Unknown	Holub 1881: 64.
	1896	(b) Dombo Dema; South of (a), Nilikau Chiefdom (west of Tegwani River), Zimbabwe.	Njenjema ‘Twinkles (Variegation) Child of Njenje	Reed 1896:LMS 18/9/1896; Ranger 1967: 190.
	??	INTERREGNUM
	1930’s?	(c)— ? south of (b)	Maboni Son of Njenjema	Sebina 1947: 84.
	?? early 1950’s	INTERREGNUM		Modongo 1975.
3. NORTH	ca. 1904	(a) — East of Maitengwe River, Nata Tribal Trust Land, Zimbabwe.	Manyang wa I.	Administrator’s Office, S.R. 1904: Al 1/2/12/11 in Ranger 1967:205.
	ca. 1920	INTERREGNUM
	late 1920’s	(b) — Nkwasenga, Matiwayile Hill, Nata Tribal Trust Land, Zimbabwe.	Mandibenga (Palali), ‘You rejected me’. Son of Manyangwa I.

Such overall constancy in regions is, crucially, an aspect of cult hierarchy. Far more checks the number of regions or priests than merely local conditions. Each cardinal oracle has its juniors. To be installed, with full authority, a would-be priest of a junior oracle requires a cardinal oracle’s recognition, which may be given through the grant of a black sash. Among the western oracles under Njelele and Dula, the line of seniority, and thus of some authority, corresponds to their geographical location: the farther east and the closer to Njelele, the more senior. Traditionally, this line is an historical sequence: Mwali arrived at each oracle according to its seniority (see note 8). There are also kinship and affinal links between some priests, if not all. But these links do not provide a conceptual framework for relations between regions or oracles. By reference to his kinship with Njelele’s founder, a priest may claim that he has more or less personal seniority among other priests. Such personal seniority, however, does not locate his oracle in a line of seniority within the cult hierarchy. To illustrate, Ntogwa recognised that his oracle was junior to Manyangwa’s. Personally, he claimed, he was senior to Manyangwa. Manyangwa I was a sister’s son of Njelele’s founder; that is, junior in kinship to Ntogwa. Moreover, Ntogwa boasted that his relationship to the founder placed him among the most senior priests, ‘We have no “senior fathers” (batatenkulu) only “junior fathers” (batatenini)’; thus he claimed that Tapa, another priest, with whom he had lived at another oracle (not of the western line), was his ‘junior father’.

As for the relations between the cardinal oracles themselves, these must be, for the moment, the subject of guesswork. I have yet to visit a cardinal oracle. I venture to guess, however, that the trinity of cardinal oracles— Njelele, Matonjeni, anu Dula—stand in relation to each other as do the Beings in the cult’s Trinity: Father, Mother and Son. The distribution of oracles in the hills corresponds to the celestial location of the Trinity: in the south, the Father and Njelele; in the east, the Mother and Matonjeni; and in the west, the Son and Dula.

Although it would be ideal to give a detailed account of the lines of oracles along with the hierarchical relations in the cult, less than the ideal must do, at present, for lack of information. The only list available to me that has been recorded at a cardinal oracle is incomplete, (it omits Ntogwa and Njenjema), and it has no stated order. Maswabi III told it to Gelfand (1966: 40) as an account of attendance at Dula, rather than as an account of the order in which oracles were established and thus the order or line of seniority (see above note 10 for the order as it is known among Kalanga in the western regions). Again, it would be ideal to spell out the transactions, rights and duties which make the lines into a hierarchy, and to show how laymen and cult officials use the hierarchy. Unfortunately, too little is known for more than an initial sketch. I have already discussed various hierarchical relations in installation, in the required accounting for cult assets, and in the directed flow of replacement payments (i.e. a cardinal oracle demands an ‘axe’—currently a cash equivalent—for replacement of a dead adept).

Two further aspects of co-ordinated control and authority must be mentioned, briefly. First, a junior priest is expected to take certain actions with his senior’s consent, after consultation. An example significant also for my later analysis of policy-making illustrates this. Ntogwa described three steps when he told me how Njenjema of the northwestern oracle rescued a missionary, Reed, before Ndebele warriors attacked (a rescue confirmed by Reed’s letter of 1/9/1896). Njenjema first took Reed under his protection to Manyangwa, his senior (wola), then reported (biga) Reed, and finally brought him to the safety of a settlers’ town, Bulawayo (see also Reed 4/4/1896 in Ranger 1967: 189). In anticipation of my later analysis, I must stress that these priests, and later their oracles, acted in co-ordination, according to a hierarchy of authority. The second aspect is this: besides such occasional co-operation, priests meet and consult the High God in council, when they collaborate seasonally in ritual at a cardinal oracle. They are said to sit then in a semi-circle, according to seniority, on both sides of the cardinal oracle’s priest, in their midst. Such councils have yet to be studied. They do appear, however, to provide a central and collective check on the affairs of individual priests and their standing, along with an oppor tunity for the most general pronouncements on issues of broad or universal concern.

A critical conclusion follows from my argument about cult hierarchy along with competition. Negatively, it is the rejection of a view, which I call the crisis conception of the cults. More positively, it is an appreciation of continuity despite, and in part in response to, political and economic change, even disruption. As for the crisis conception, this overruns the past eighty years of literature on the cults, since the so-called rebellion in nineteenth century Zimbabwe (Selous 1896: 14f; Campbell 1926; Jackson 1925; Tredgold 1956; Gann 1965: 134f; Ransford 1968: 8). Initially, this conception was fostered by white settlers along with their Ndebele allies or helpers to explain—or rather explain away—the war’s aims and inspiration. Yet, oddly enough, the most modern version comes from the British historian who provides the most closely documented and masterly assessment of the earlier versions: Ranger (1967).[77] Moreover, Daneel (1970, 1971), who insightfully reports clues contrary to this conception, follows Ranger’s lead, nevertheless. This oddity is, of course, a proper subject for writers of historiography to explain. My task is to make plain the inadequacy of the crisis conception, and put forward a better alternative. I must reserve for elsewhere a close examination of the historical data.

In brief, the most modern version runs as follows. A grave crisis, indeed the unprecedented catastrophe of white conquest, drove cult leaders to renounce their religious opposition to violence. Immediately after the onset of colonial rule, they ‘lent both their moral support and their organisational apparatus to the preparations for the rebellion... Curiously enough there is least evidence about the Njelele shrine, which all authorities agreed to be the senior and most influential’, (Ranger 1967: 148–9 my italics).

Fundamentally, this crisis conception of the cults is wrong because it obscures that the diverse interests within the cult’s domain called for a broad consensus in oracular policy. A war policy against white settlers, and thus for a restoration of a conquered Ndebele kingdom, suited the interests of people in no more than a small fraction of this domain. It was a fraction that included some non-Ndebele; they were connected with the Ndebele regiments, for example the people at the area’s eastern fringe in Belingwe among whom a small number of Ndebele also lived (Selous 1896: 238). The people in the rest of the domain had one policy, and they were the ones who co-ordinated their action over the widest area. The policy of protecting the white settlers, even fighting in their defence against the restoration of Ndebele dominance, was, as seen at that time, more in the interests of the people who adopted it throughout a great crescent around the Ndebele kingdom’s stronghold in the highlands of Zimbabwe. This crescent runs from east to west across Ndau, Karanga, Venda and Kalanga country (see Map I in the Introduction).[78] The cardinal and regional oracles of this vast crescent had to have—indeed, did have—a policy in accord with that of their people: there is neither evidence nor reason to think otherwise.

Even before 1893, when the colonial conquest put an end to the Ndebele kingdom, the staff of the cardinal oracles had interests which were, in various senses, unlike and greater than the Ndebele kingdom’s. This staff stood for the High.God’s hierarchy of authority which was beyond, prior to, and much more far-reaching than one king’s. They were thus accused of being presumptuous and of committing acts of lese-majeste. In order to punish them, the Ndebele king let loose against them his personal bodyguard (the Imbizu regiment, and also the Insuka regiment), and some of the cult staff were killed. However, to argue that cult leaders did not take part in the 1896 war party merely because of their past history of conflict with a recently dead Ndebele king would be to clutch at a straw. What must be grasped are the underlying constraints and the predicaments at the very heart of their theology and their authority.

‘The shrines (daka) are many; God is one,’ Kalanga say, Their High God’s oneness transcends, in their theology, the recognized disparities, even hostilities between the communities that seek His blessings. It is the macrocosm, the order beyond that of the congregation or any single community, towards which their theology directs their attention. The right to free movement on religious errands across communal borders is basic; such borders must not be closed against the cult staff or supplicants. Moreover, while any murder offends God, the shedding of an adept’s blood is a grave sin, considered tantamount to killing oneself, since it causes a ‘standstill’ (chamwi), when rain fears to sow the earth, and men may die of hunger. Furthermore, the portion of the earth that individuals and communities have they hold from God, its Creator and ultimate Owner. In fighting over it (see Werbner 1975: 109f), they commit a sin which must be paid for. Not that their conception represents a single wider society: that would be, in theological terms, to reduce the macrocosm to a microcosm with its defined boundaries. Conversely, the microcosm must be sustained vis-d-vis the macrocosm: the autonomy of communities must be respected. Thus a person ought to die or at least be buried in .his or her own community; i.e. preferably the person’s home ward or in a European’s case, a hometown and not the countryside. Otherwise, the corpse becomes a chamwi.

Again, it would be clutching at a straw to argue that such cosmological conceptions or even the whole of their theology dictated the actual responses of the cult staff in a crisis. Yet equally trivial is insistence on another elementary and primary cause such as the one in any materialist view which disregards the cosmological conceptions. These conceptions are a significant factor in behaviour because they are of such a kind that communities can continue to define their broadest consensus through them irrespective of their differences, hostilities and competition. In a crisis, the more inclusive the majority who are rallied through the cults, the more comprehensive and fundamental is the appeal to such cosmological conceptions. Even the minority, in its somewhat unsuccessful call for wider support during the 1896 war (the so-called mass commitment—Ranger 1967: 353), had to recast these conceptions rather than make radical religious innovations. Thus a more fundamentally wrong view could hardly be put, in my opinion, than the following:

‘These hints justify us in finding in the risings of 1896–7 [including the war to the east among Shona], though in different proportions, the same elements which are so clearly detectable in the twentieth century millenarian movements with which they are in other ways so comparable’ (Ranger 1967: 354).

In my view the need to respond to a broad consensus, the cult procedures for the discovery and definition of such consensus (i.e. through councils at the cardinal oracles), the commitment to macrocosmic conceptions, all acted—and continue to act—as constraints on the cult leadership as a whole. Admittedly, the cult hierarchy was and is sustained through a capacity to collect, store, and allocate great resources and riches. But, in turn, this capacity must be sustained through voluntary consent for collaboration between communities. And it is here, in the winning of consent, that the predicament at the heart of the hierarchy’s authority lies.

One last point about oracular policy-making must suffice here. According to the crisis conception, in its modern version, the cults’ alternative to armed resistance has been, throughout most of the colonial period, a ‘sort of despairing passivity’ (Ranger 1967: 378) (see my note 14). Again the contrary is the case, as shown very clearly in the following account, which has numerous parallels elsewhere in the cult’s domain:

‘Nyusa [Messengers] brought the instructions that [black people] were not to sell grain to white people except for salt, and then as little grain … as possible, as Mwari desired them to obtain their salt at Brak [sz’c., brackish] places or by burning salt grasses, as they did before the white people came. Should they barter grain for salt, they were to cover their baskets so that Mwali could not see it. [They] were told through Manyusa [messengers] that Mwari would send a great wind and that all the white people would leave the country in the night. On their departure, Mwari would introduce other white men who would charge I shilling tax only, and sell goods at a quarter of their present price.’ (Franklin 1932: 82).

As seen by some sophisticated Kalanga, such as an ex-headmaster, ‘Mwali is a conservative God. He does not like change. He wants the old ways.’ Indeed, to some Kalanga Mwali appears to have a devotion to customs for their own sake. Yet this appearance, though not a disguise, is nevertheless a cover for the enduring orientation of cult policy. It is conservative, but with economic and political concerns which are as basic in a neo-colonial period as in a colonial one. The oracles’ messages have repeatedly advocated resistance to the inroads of a cash economy. They have urged the people to store their grain, to rely on mutual aid in agriculture, to brew beer for the co-operative labour of work bees, and not to be at the mercy of sales. ‘Why do you sell all your grain, and not keep some to help the children?’ As Ntogwa’s daughter told me, ‘Mwali objects to sales, for the country is not sold. It is given to its owners.’ Similarly, in accord with the tide of popular opposition in Zimbabwe to measures for European control of agriculture (i.e. through the making of ridges and fences), the oracles’ messages have condemned the measures: they spoil the land and cause ‘standstills’ (zwichamwi) of rain. Cult policy has been even more directly active politically (rather than ‘passively despairing’) in other respects also. At the time of the first elections prior to Botswana’s Independence, the messages warned of the dangers of party politics; i.e. that office is given by people according to law, and not fought over. Furthermore, in 1974 the message brought back to Botswana shrines from the northern oracle in Zimbabwe was. ‘The rain is held back because of those who fight over the country, those of Smith, Nkomo and Kaunda.’ Thus the central concern of the oracles’ messages was and still is to conserve order in the world and maintain the welfare of the land, its people and their economy. Continuity is the policy of the cults, and to explain it fully calls for a consideration of changes in the ecology and geopolitics of a vast domain of Southern Africa.

My account is concentrated primarily on the cult’s western regions in Zimbabwe and Botswana. The reason for this is simple: I know them through long-term observation. However, as I have shown, these regions have to be understood in relation to others. There are wider sets of regions, and there is a definite order among the limited number of regions in each set. Moreover, the cult itself is an organized, if somewhat unstable, whole. It is hierarchical and centralized, and its staff are’ ranked and recruited accordingly: the higher the rank, the narrower and more exclusive are the criteria for recruitment, and recruitment to the inner circle of priests is on the most exclusive basis. Furthermore, certain conditions have long prevailed across regions and virtually across the whole of the cult’s domain. Perhaps most crucially, these conditions have favoured, even compelled, the relocation of communities or their boundaries and great waves of individual migration. The populations that have thus come to live side by side in dispersed hamlets and homesteads have become more and more diverse ethnically. That is why the cult has had to have the capacity to override cultural differences and admit newcomers, irrespective of ethnic origins. In directing their attention towards the macrocosm and a High God, the cult appeals to a more universal order: this order transcends their differences. It is essential to appreciate, however, that various constraints continue to operate, and are anticipated by the people themselves, because the cult’s domain continues to be set within a field of relations which is somewhat like a patchwork. This field is made up of the diverse and competing congregations and communities in each region, the variable sets of supplicants in its vicinity, and disparate, even opposed churches and ritual associations. All these parts are hardly constant or alike in importance. Indeed, through a particular priest’s efforts the importance of a part of the patchwork may alter or a new part may be introduced. Moreover, a priest must commit his various resources, and thus gets more involved in one part than another. Yet a part is not the whole, and radical change in a part need not transform the whole. It is the disparity and diversity of competing or opposed parts that continues to constrain responses within the cult and the orientation of its leaders.

My analysis of this High God cult illuminates that it is essential to study changes in regional relations in order to understand a cult which reaches, both in ideology and organization, beyond the boundaries of any single political community, whether this be a tribe or a nation. The supposed wholeness of a single society cannot provide the starting point for an analysis of such a cult. Nor can it be explained by reference to some ‘correspondence’ between religion and the social order of any single community: the cult’s regions are not a replica of the social divisions which exist apart from ritual relations. On the contrary, the cult’s regions, like all its ritual relations, have an autonomy of their own. Their transformation must be studied in its own right, therefore, and not as some dependent variable of other phenomena.

References

© Richard P Werbner 1977

8. Cult Idioms and the Dialectics of a Region

J. M. Schoffeleers

Introduction

In the ritual cycle of the Mbona cult of southern Malawi and adjacent areas of Mozambique the most prominent event is of a sociologically distinct kind. It is the only occasion when its various participants act in unison. Outside the context of the Mbona cult they belong to different countries, to different administrative units within these countries, to different churches and to several other groups and sets which pursue their own, sometimes conflicting interests. Only in the Mbona cult do they have a conceptual basis for concerted action which transcends these boundaries and which brings them together in a kind of overarching region.

The event which requires their ritual collaboration is the rebuilding of the shrine and organisational centre of the cult. This shrine is like a hut, with the usual proportions, and highly perishable. It often has to be rebuilt, on average every five years but there are no rigidly fixed intervals. The decision to perform the ritual of rebuilding depends as much on the condition of the shrine as on the need for a ritual in order to counteract some calamity. In any case, rebuilding requires protracted consultations between the formal leadership of the cult and the rulers of the various chiefdoms of the region, on which occasion old animosities are habitually revived and fresh ones created. Often the ritual has to be postponed, because the parties involved are unable to reach unanimity. However, they may be forced to act by the cult medium who, while in a state of possession, personifies the deity and compels the various factions to make peace so that the act of rebuilding may proceed. One also observes on this occasion a general resurgence of people’s faith in Mbona. Ancient myths are retold and refashioned; rumours about strange omens circulate and, when the shrine is finally rebuilt, a great many contributions are made in the form of food and labour.

Mbona’s shrine is located at a short distance from the township of Nsanje in southern Malawi. The earliest known recorded mention of its existence dates from the time of Livingstone’s travels through the area, but apparently it is much older (see Schoffeleers 1972). Local tradition associates its origin with the occupation of the region by the Mang’anja, a Chewa-speaking people (Tew 1950), who had already developed a centralised state system by the end of the sixteenth century. According to this tradition, Mbona was a rain-maker of great repute who, on account of his influence with people, aroused the wrath of the first paramount, Lundu, who had him killed. After the execution the rain-maker’s head was severed from his body and shown to the people to prove that he was no longer alive. Subsequently, however, a violent storm began to rage. In it all the executioners perished. Mbona’s spirit then took possession of a villager, and through his mouth Mbona made known that his head had to be buried and a shrine erected over its resting place. Mbona also ordered the paramount Lundu to provide a woman to live at the shrine as Mbona’s wife. The paramount complied; then rains fell, and the country prospered. This is the essence of the myth, as it is recounted to this day.

What characterizes this region in the first place is the shared belief that only the cult and its staff has the means to counteract such calamities as droughts, floods, locusts and epidemics, and that it is therefore in the interest of all to support the cult. Yet the authority of the cult and its staff reaches beyond these pragmatic considerations. On the basis of the equally common belief that natural disasters do not happen independently but are caused by an angry deity in punishment of sins, the cult’s staff also pronounces on the nature of these sins and thereby on social behaviour in general. In consequence of its acceptance of this authority the cult region is more than a collection of pragmatically motivated people; it also presents itself as a moral community with its own axioms and values and its own standards of conduct.

To understand how people conceptualize their region in the context of the cult two conceptions and the dialectical relations between them need to be analysed. These contrasting conceptions are respectively hierarchical and egalitarian in orientation. The first is embedded in the activities and mythical representations of the formal leadership of the cult, and it underscores the legitimacy of the traditional (or quasi-traditional) political organisation. The second underlies the activities and myths of the population and the medium, and it emphasizes values and concerns which transcend all forms of political and economic inequality including that between chief and subject. The hierarchical conception asserts itself in the day-to-day operation of the cult. The egalitarian conception, on the other hand, although continuously present as an undercurrent, asserts itself only intermittently, when the medium is possessed and during the processes which generate such possession. During the possession sequence, the authority of the medium takes precedence over that of the formal leaders of the cult, all of whom are also government acknowledged chiefs. Indeed, the medium may request that they act for the cult and their people in ways which they might otherwise have been unwilling to undertake. The cult leaders, on their part, may refuse to attend or accept the medium’s demands: they may declare that his possession is simulated. The possession event thus assumes the character of a contest between a formal and an informal type of leadership within the cult organisation. Furthermore, such an event constitutes a moment in time when the two conceptions of the region are brought into direct confrontation by their respective protagonists.

Clearly, there are critical factors which influence the decision of the cult leadership whether or not to attend the mediumistic seance. The most critical of these factors is the degree of support that the medium receives from the people. It is not accidental that this support finds its expression in the outbursts of rumour which precede possession. The evidence suggests that such outbursts provide the contents of the communication transmitted by the medium. Consequently, the possession sequence has been regarded both as the summation and the climax of a wider process which, in principle, involves all the people.

Three main steps are necessary for my argument about the cult processes and conceptualizations. The first step which begins this paper is to examine the formal leadership in relation to the cult organisation and its political setting. Next the medium’s position has to be considered in relation, on the one hand, to the formal leadership of the cult and, on the other hand, to the entire population. Finally, the actions of the two kinds of cult leader — the formal leadership and the medium — have to be related to the two opposing conceptions of a region.

The Formal Organisation of the Cult

The organisation in its present day form may be envisaged as operating within four spatial categories that form four concentric circles (see Map I). In the innermost circle are the shrine grounds, some ten acres in extent, which comprise a sacred forest, the shrine proper, a place of assembly, and a small settlement where the attendants live. The shrine grounds are considered to be outside the jurisdiction of any secular authority, and they are therefore not subject to government rules and regulations concerning land use; nor do its inhabitants pay taxes. The next circle corresponds to a land area of about twelve square miles which directly surrounds the shrine grounds. This area, known as Khulubvi, is demarcated by topographical features associated with Mbona’s martyrdom. Prominent among these is the sacred pool which is supposed to have originated when blood flowing from Mbona’s body turned into water. Within this area reside the territorial chiefs who are the principals of the cult, the minor officials, and the cult medium. All of these hold their positions partly by virtue of their association with this spatial category. The third circle corresponds to the district of Nsanje which is the administrative division within which the principals act as territorial chiefs, subject to the policies of the central government. The outer circle corresponds to the entire area over which the Mang’anja are dominant in a traditional sense. This area is co-terminous with the cult region and covers part of Malawi as well as Mozambique. Apart from the Mang’anja, who consider themselves autochthonous, its population consists of a number of other ethnic groups that have moved in more recently and mostly from the Zambezi valley. The Mang’anja have remained dominant in traditional politics by dint of occupying all the chieftaincies, whereas the immigrants are only represented in village headmanships. On the other hand, many immigrants have been successful economically as rice and cotton growers, fishing entrepreneurs, cattle raisers and shopkeepers. Moreover, on this basis they have been able to secure prominent positions in the Malawi Congress Party. The cult organisation formally emphasizes the same division between autochthons and immigrants that is characteristic of the system of chieftainship: foreigners are not supposed to hold major positions in the cult.

The Chiefs and the Shrine

All Mang’anja territorial chiefs are agents of the cult by virtue of their office. They are consulted about the performance of major rituals; they receive requests for and organise the collection of contributions to the shrine, and they have the right to attend on major ritual occasions. Moreover, when a particular area suffers from drought or some other calamity, petitions to the shrine organisation for ritual assistance are made through its chief. Theoretically, therefore, the Mang’anja chiefs, whilst divided by different administrative systems, appear as one body in the Mbona cult. They can be said to share a common estate, i.e. the cult organisation; to hold common jural responsibilities in respect of that estate, and to be capable of concerted action for the maintenance of that estate.

As agents of the cult, the chiefs are differentiated according to their location in relation to Khulubvi, the ritual area marked off by events in the myth of Mbona’s death. The principals are the chiefs within Khulubvi, and it is through those outside it that the shrine organisation maintains its contact with the other areas of the region. The principals, like the rest of the chiefs, have independent jurisdiction over cult matters in their own constituencies, but in addition they also exercise authority over the shrine grounds, its personnel, and the performance of rituals at the shrine. They have authority to grant permission for the performance of rituals on behalf of other chiefdoms; to appoint or dismiss personnel; to judge cases of violation of the shrine grounds, and to preside over any deliberation which concerns the conduct of the cult. For most of the colonial and post-colonial period there have been three principals, but in 1968 one of them was divested of his office for reasons to be discussed immediately. The relationships between the principals are characterized by perpetual conflict due to continuous attempts on the part of each of them to achieve ascendancy over his peers, which in turn leads to frequent accusations of abuse of authority and to allegations of sorcery.

This quest for primacy is expressed in the origin myths of each principal’s office. I will consider each in turn and later present a schematic resume of two versions in Chart I. Chief A claims the right to primacy because his apical ancestor was granted exclusive jurisdiction over Khulubvi by one of the Mang’anja paramounts. Chief A’s ancestor had then delegated the right to offer sacrifices to Chief B’s ancestor, but without granting him any secular jurisdiction. This, Chief A maintains, was the situation till the introduction of the colonial government, when B obtained his chiefly title from the government under false pretenses. C’s right to the office of principal. was even less defensible, according to this version, as his ancestors had been no more than ordinary headmen in precolonial times.

Chart I: Resume of Origin Myths of Two Principals

Issues of Conflict:	A’s version	B’s version
Original nature of ego’s authority	Secular and ritual	Ritual and secular
Grantor Onego’s authority	Secular authority (Paramount)	Ritual authority (Mbona)
Original nature of rival’s authority	Ritual only	Secular only
Grantor of rival’s original authority	A’s ancestor	B’s ancestor
Nature of rival’s illegal action	B assumes secular authority B makers himself independent	A makes himself independent and assumes ritual authority

Chief B’s version of the myth is that Mbona himself appointed Chief B’s apical ancestor to be priest and granted him secular jurisdiction over the entire area of Khulubvi. Chief A’s predecessors were then given part of Khulubvi to administer on behalf of B’s lineage. Chief A, in his turn, subdelegated part of his jurisdiction to an ancestor of Chief C, who then became a sub-chief.

Chief C’s account is less clear, which is probably due to the fact that his office is of rather recent origin. One of his predecessors was among the first converts of the Protestant mission which was established in 1900. It was partly due to mission protection that he was officially recognized chief in 1913 and raised above a predecessor of A, who became his sub-chief. This relation was maintained until 1968, when the Malawi government restored A to the chiefship and reduced C to the status of senior headman. In 1972, C was even divested of that position following charges of slander and sorcery against A.

To underStand these conflicting accounts it is convenient, to limit discussion to A and B, and omit C, for the moment. Their differences have been schematized in Chart I.

The two rival versions show a common sequential structure which evolves in three parts: (1) A superior authority (either that of the paramount or the cult organisation) grants someone combined ritual/secular authority over the Khulubvi area; (2) the grantee delegates either the ritual or the secular part to another person; (3) the person thus given one part appropriates the other part which is lacking and makes himself the equal of the person who delegates him.

Reduced to its essentials, the cause of the conflict appears to be the illegal appropriation, by one’s rival, of one type of authority on the basis of the other type. Thus one of the two is accused of extending his ritual authority to the field of secular authority, and the other is accused of having done the opposite. Each in fact assumes that the combined exercise of the two authorities is legitimate in his own case (and, by implication, also in the case of his grantor), but not in the case of his rival. This, I believe, is more than a conflict between two rivals over a particular right; it is also, and more profoundly so, a questioning of the right itself. It reveals in its own way the profound tension between ritual and secular authority which appears to pervade the cult at all levels: its mythology, its formal organisation and its informal operation.

The permanent condition of conflict between the principals reaches its climax on two different occasions and in two different settings which are nevertheless in many ways related. One of these occasions is the rebuilding of the shrine; the other presents itself whenever the district administration contemplates changes in the boundaries of chiefdoms or in the positions of individual chiefs. The Nsanje district administration has for one of its constituent bodies a chiefs’ council whose chairman is the district commissioner, now known as the government agent. The chiefs’ council, in its turn, is responsible to the district council on which it is represented and to which it makes recommendations. A perennial problem of the Nsanje district administration has been the relative smallness of the chiefdoms in the southern part of the district, a state of affairs historically attributable to the presence of the Mbona shrine. Consequently, there have been a number of attempts, partly for the sake of administrative efficiency, partly for the sake of more intensive political control, to arrange them into larger units. Between 1913, when chiefdoms for the first time became officially recognized administrative units, and the present, there have been eight alterations of boundaries, five of which involved the Khulubvi area. In each case this led to severe competitive struggles and to the creation of a powerful undercurrent of hostility in the relationships between individual chiefs.

Competition for chiefships in the district affects the cult in a number of ways. Such competition is inextricably enmeshed with the competition between the principals: their standing with the district administration is a major factor in their quest for primacy. It also affects the relationships between principals and neighbouring chiefs to the extent that the latter frequently refuse to support the shrine organisation. Even a catastrophic drought may not be sufficient to rally the local chiefs around the shrine. A case in point is the drought in 1948–9, this led to a general revival of rain cults in Malawi, but in Nsanje it coincided with one of these periodic power struggles, and a number of chiefs did not rally around the shrine. Finally, the competition emphasizes the ambiguous position of chiefs operating within a system of modern government. A chief’s position under such a ’system is more directly dependent on the support of the government in power than on the support of the population. Capable individuals may be able to steer a cautious course between the demands of one party and the other, but the conflict is never far from the surface. In the Nsanje district, where the most sensitive issues have traditionally been those relating to taxation, cotton prices and agricultural regulations, this situation has at times led to occasions when the chiefs, including the cult principals, were expected to implement government policy whereas the cult medium encouraged disobedience and resistance.

Against this background of conflict and ambiguity it may be asked how the principals are able to exercise effective authority. Part of the answer is that their position is similar to that of other chiefs in that their authority largely depends on the support which the central government is willing and able to provide. This constraint appears at its clearest in the history of principal C who was ultimately deposed but who, while his favour with the government lasted, was able to maintain his authority in the cult. In addition to this, the principals, like all Mag’anja chiefs, are credited with the power to withhold rains or cause some other calamity. However, in their case, this power along with their cult jurisdiction is believed to operate also beyond the boundaries of their chiefdom. To illustrate briefly, in December 1967 principal B had requested his people to repair the grass roof of his house, but without success. At the same time there was the threat of a drought, the rains having already been a few weeks late. It was decided to offer a sacrifice, but B refused to be present. No rain fell after the sacrifice, and since it was the general opinion that this was due to the sorcery of the principal, apologies were offered and the roof repaired.

The religious and secular authority of the principals, although protected by the same sanctions, are in part challenged by different agencies. In their role of secular rulers they are subject to challenges by the central government represented in the person of the district commissioner; as religious authorities they are particularly open to challenges by the people represented in the person of the medium.

The Spirit Medium and the Informal Operation of the Cult

Within the cultural idiom of the Mang’anja a medium is a person who is able and authorized to receive and transmit messages thought to emanate directly from supernatural beings. The word ‘directly’ serves to distinguish mediums from ordinary diviners whose contact with the supernatural, insolar as it is thought to exist, is mediated through some external object. I’he distinction between medium and common diviners is socially important, among other reasons, because divination is sometimes used deliberately by chiefs who wish to avoid the services of spirit mediums. Contact with the supernatural is commonly established through trance during possession. The medium is said to be seized (-gwidwa) by a spirit (mzimu) and, while in a condition of dissociation (-gwa), he babbles and raves (-bwebweta). His utterings are often unintelligible to outsiders unless ‘translated’ by the medium’s acolyte or interpreter.

In the course of their history the Mang’anja have developed or adopted a variety of mediumistic institutions. Broadly, Mang’anja distinguish between mediums who operate on behalf of private interests and those whose concern is with public interests. The former specialise in the diagnosis and treatment of diseases, whilst the latter — whom I designate territorial mediums — predict and explain disturbances in the natural environment. Territorial mediumships show significant variations according to differences in sex, means of contact with the divine, relationships with chiefly lineages, and the spatial extent of their jurisdiction (Schoffeleers 1974). Thus some mediumships are exclusively held by women; others by men and women. A few receive their revelations in dreams, while the majority receive and transmit them through possession. Some are members of the local chief’s lineage; others operate independently. Finally, some have local significance, whereas others such as the Mbona mediums have regional significance. Politically, the most crucial variable is that of the relationship between medium and chief. For it is considered that, where a medium is a member of the ruling lineage, he will be close to the interests of that lineage, and that, where he is not so related, there is a definite potential for opposition to the ruling lineage’s interests. Indeed, it would appear that the history of territorial mediumship among the Mang’anja can be validly interpreted only in terms of these two conflicting tendencies.

Chart 2: Mbona Mediums in the Twentieth Century

Name of Medium	Sex	Period of activity (approx.)	Biographical Notes
I Mbote	Male	c.1900–1914	Immigrant lineage; independent; divested
2 Mbandera	Female	c.1920–1930	Full sister of 1; independent; died in office
3 Onsewa	Female	c.1930–1933	Relative of principal B; died m office
4 Edward Harrison	Male	c. 1933–1943	Relative of principal B; died in office
5 Jailos Thom	Male	1946–1951	Younger brother s son of I and 2; independent; died in office
6 Thapuleni Mybeck	Male	1952-	Claimed to be relative of principal B; rejected as medium
7 Joseph Thom	Male	1954-to date	Younger brother of 5; independent

Much the same can be maintained of the post-1900 history of the Mbona mediums (see Chart 2). The leading mediums during this period have been members of an immigrant lineage whose ancestress had been granted asylum at the shrine. The present medium of that line, like his predecessors, is quite definitely independent of the principals with whom, moreover, he is in almost continuous conflict. Other mediums of this period, however, have belonged to the lineage of one of the principals. One of them, Edward Harrison, enjoyed considerable renown between the mid-thirties and 1943. His was a case in which attachment to a chiefly lineage could be successfully combined with the confidence of the population. His success was due to the convergence of several favourable factors, including Edward’s own personality, the popularity of the principal whose kinsman he was, and the political circumstances of the time.After Edward’s death, however, the credibility of the chiefs came increasingly to be questioned during opposition to unwanted agricultural reforms and the imposition in September 1953 of the Federation of Rhodesia and Nyasaland. It would have been difficult in such a climate for any medium closely related to a chief to maintain himself. Although at least one aspirant did try, he was unsuccessful, and the Mbona mediumship has from 1946 onwards remained in the hands of independents.

The Medium’s Entourage

One particular facet which needs to be made explicit is that the Mbona mediumship, at least in its present form, is not being used as an oracle in the sense that the medium can be called upon to provide divinely inspired solutions to such problems as succession disputes. Possession is conceived of as an event which is entirely dependent on the initiative of the divinity, and solutions to specific problems, insofar as they are provided, are not given at the request of one or other party but only in response to a general feeling of crisis. This has the important implication that the medium need not be in any way better informed than the average inhabitant of the region. All he needs is to keep in tune with the general sentiment of the population. In accordance with this principle, the medium does not travel widely; nor does he maintain a network of informants. What information he needs he can, in principle, pick up from village or market talk. There is one exception to this, which is that he must be better informed than the average person about what goes on at the shrine. However, since he lives nearby, and since little remains secret on account of the conflicts between the principals, this constitutes no particular problem.

Still, the medium has a small entourage of kinsmen and sympathizers who in various ways assist and support him. Within this entourage the important roles are those of the medium’s guardian (kholo) and of his acolyte/interpreter (womasula). The former is the head of the lineage section (mbumba) to which the medium belongs, and he is always male. The latter may be any adult member .of the lineage, but preference is given to a woman of mature age. The guardian as the head of the lineage is co-responsible for the actions of its members, including the medium, unless their relationship is severed on account of some conflict. He is further entitled to act on behalf of a lineage member when he or she is permanently or temporarily incapable of acting on his or her own behalf. This is considered to be the case when the medium is possessed. It is the guardian who then takes command by sending messages to the principals and by formally receiving them, if they decide to come.

The task of the interpreter, usually a woman, is to render the ravings and mumblings of the medium intelligible and understandable to those to whom they are directed, and to convey the replies of the latter back to the medium. Her role is a critical one in the sense that the outcome of the seance partly depends on her ability to formulate. She has therefore to be a person of proven ability. The choice of an interpreter is the exclusive affair of the lineage to which the medium belongs. Should one medium die and another emerge from the same lineage, as has been the case with the present medium, then the interpreter will also serve the next medium. If, on the other hand, the next medium comes from a different lineage, then she will cease to function as an interpreter. The point really is that medium, guardian and interpreter co-operate in the interest of their own lineage which stands to benefit from its association with the mediumship. In the past, it seems that successful mediumship could, given the right conditions, be converted into one or other form of hereditary secular authority. This no longer is the case, but the present medium and his kinsmen seem to think that he or his successor may possibly become the sole head of the cult. This at least is the interpretation which the principals give of the fact that the staff have built a separate possession hut which at the same time is used as a shrine. It appears, therefore, that the hostility between the medium and the principals, which is so much a feature of the present situation, is not merely incidental, rather, it is to be viewed as a logical consequence of the fact that the medium is a potential rival of the principals, both in their role of secular rulers and in their role of heads of the cult organisation.

The Mbona Medium and the Chiefs

The Mbona medium becomes possessed rather infrequently, at intervals of several years. Such possession maybe provoked by three different types of events, viz. environmental calamities, offences against the social order, and offences against the cult. In actual fact, it may be difficult to establish in any particular case which of the three is to be regarded as the activating factor, since nature, society and cult appear to form an indivisible triad. Environmental and climatic disturbances are routinely explained in terms of social disturbances and vice versa, while both again are connected with the cult as the censor of public mores and as possessing the means of expiation. Whatever the immediate cause of possession, it invariably involves a pronouncement on the social and moral condition of the community. Moreover, there is a further general aspect of the message transmitted by the medium: it is always centrally concerned with the behaviour of the chiefs and with the relationships between the chiefs and the population. Characteristically, the chiefs have to bear the brunt of the medium’s criticism, either because they are accused of siding with the central government against the population, or because they are accused of failing to put an end to social abuses within their territories.

The antagonism between medium and chiefs, which manifests itself in possession, is equally apparent outside it. The present medium continuously criticises the actions of chiefs, particularly the Khulubvi chiefs. In their turn, they criticise the medium for trying to enhance his status by unlawful means. The chiefs are particularly critical of the fact that, with the help of kinsmen and sympathizers, the present medium has built for himself a large shrine with an altar to which he retires when possessed. They regard this as a departure from tradition and as a sign that he wants to make himself independent of the cult organisation. It is a striking fact, however, that despite these allegedly unlawful acts, they have left the medium in peace. One of the principals whom I questioned on this point replied that, notwithstanding such obvious irregularities, one still ought to respect the medium, since he spoke with the voice of Mbona. Nevertheless, in view of the fact that mediums in the past have sometimes been discredited by the chiefs, one may ask why they have not done so to the present incumbent. The most likely answer is that he enjoys considerable reputation among the people, both because of the capable manner in which he performs his role and because of the courage which he has shown in crises. He was for instance one of the..persons arrested as one of the suspected leaders during the political riots of March 1959.

The principals may also express their disagreement with the medium by refusing to attend a possession seance. They themselves state that this depends on whether possession is real or feigned, and they claim that they are able to judge this by the medium’s behaviour. Thus principal B once said of a candidate medium who presented himself in the early 1950s (number 6 in Chart 2), “I went to listen to his ravings, but when I heard what he had to say, I turned away in disgust. The man was a drunkard.” Nevertheless, the fact that they may decide to stay away from the seance before they have actually observed the medium suggests that their judgment may be based on an entirely different criterion. My suggestion would be that the criterion actually applied is their view of the degree of political pressure which generates a seance in a particular case. If this pressure is considered insufficient, chances are that possession will be declared false, and that none of the principals will attend.

Lest this suggestion be taken to imply that those involved do not actually believe in the reality of spirit possession, I must stress that all my evidence indicates, that they do. My suggestion does point to the fact that they may be using conscious and unconscious criteria at one and the same time. Much the same can be said of the medium. Although, from the political viewpoint, the art of mediumship consists in choosing the right moment for possession, the medium himself may be under the impression that it occurs spontaneously.

The Encounter between Medium and Principals

Formal encounters between medium and principals take place in the context of the possession sequence which is always precipitated by a combination of social, environmental, and cultic factors. A closer look at the rebuilding of the shrine will clarify some of the factors in the cult itself which trigger such a possession sequence.

The shrine is made of perishable materials — grass, reeds, wood and bamboo — all particularly subject to rot and destruction by white ants. Permanent materials must not be used and are not allowed; nor may repairs be carried out. The shrine remains untouched, apart from the occasional sweeping of its interior, till the next rebuilding. The act of rebuilding is accompanied by two further rituals, namely the induction of the spirit wife and an animal sacrifice on behalf of the entire population. In the past the spirit wife would live permanently at the shrine, but nowadays a woman is chosen who stays there only for the duration of the ceremony. These actions require the co-operation of the paramount chief, Lundu, and of a number of senior chiefs, all of whom are supposed to perform certain parts of the ritual. The lesser chiefs are mainly involved through the collection of grain and other victuals for the sacrifice and the feeding of those who participate. As mentioned earlier, a good many always refuse to co-operate because of their conflicts with the principals. The principals may make attempts at reconciliation or they may try to bring pressure to bear on the unwilling through the offices of friendly chiefs, but the outcome is always the same: the rebuilding is performed with the help of substitutes and in what is considered to be a simplified and adapted form. The traditional ritual, as preserved in oral accounts, has in its entirety never been performed since the middle of the last century, and it is doubtful whether it ever has been. At any rate, the way the rebuilding is to be carried out nowadays is the source of endless altercation between those who may be conveniently labelled the ‘purists’ and the more ‘practically minded’. One party demands that tradition be followed to the letter; the other allows for adaptation and simplification. This dispute is in reality a political one. No chief will ever admit that he refuses co-operation because he has been outdone politically. Rather, he will maintain that his reason is a conservative one: he is opposed to the principals and senior chiefs who no longer care for the cult or perform its ritual according to the way of the past. The political losers thus tend to become the ‘ritual purists’. The usual result of these altercations, however, is that negotiations drag on without a decision being taken.

The negotiations about the rebuilding of the shrine are a major source of rumour among the people. While on the whole uninformed about the details, people, even when they live at a great distance from the shrine, are aware of the essentials which they interpret in their own fashion. The medium, on the other hand, living close by the shrine, is fully informed about what goes on, although he does not directly participate in the negotiations himself. In the case of the present medium possession announces itself days beforehand through severe aches in the back and the head. The decisive sign, according to his own information, comes when he Sees a small cloud moving at great speed through an otherwise bright sky. If he is away visiting, or working his gardens, he returns and tells his wife. Throughout the rest of the day he shivers and mumbles, and towards evening he is secluded in his special possession hut where he wraps himself in a black cloth. Messages are then sent by his guardian and relatives to the principals and their assistants.

This summoning is a critical point in the possession sequence, for those called may or may not come depending on their reading of the situation. If possession occurs in the midst of a drought or some other calamity, the decision will be easier to take. Failure to come might be interpreted as a sign of sorcery, and the principal would then pit himself against the medium and the population, which involves considerable risk. The less pronounced the element of environmental and climatic crisis, therefore, the greater will be the chance of a clash of wills between medium and principals. If indeed the majority do not show up, the medium’s career is in jeopardy and it may be irreparably damaged. If, on the other hand, a majority has assembled, the medium begins to stir and his acolyte, who is his interpreter, holds a roll call from behind the reed door of the hut. This is followed by the message proper, which is in three parts, First, the nature of the crisis is stated; then an explanation is given in terms of some delict committed; finally, the means to avert disaster are indicated and a promise solicited.

The next critical moment arrives when it is time for the spokesman of the principals to reply. If everything goes well, he makes a formal act of submission by addressing the medium as his senior; by confessing to neglect of the shrine and such other offences as may have been mentioned by the medium; and by promising that things will be done as requested. In that case, the medium gives a parting admonition in a conciliatory tone and all disperse. If, however, the principals find it impossible to accept the terms posed by the medium, or if they feel offended by the medium’s denunciations, the session may break up and the medium’s future is once again in jeopardy.

The specific and immediate effect of intervention by the mediumist is the rebuilding of the shrine which ideally signals the end of a period of contestation and the beginning of a new era of unity and co-operation. The traditional accounts of the ritual lay particular stress on this aspect. Thus, no household was to give more than one cob of maize or head of millet “so that the gifts of all would form part of the libational beer”. Nobody was allowed to carry more than one shaft of reed or a bundle of grass thicker than one’s own arm “so that the labours of all would be required to rebuild the shrine”. This injunction is no longer followed, if it ever was, but the contributions in the form of food and labour are still impressive as is the participation by the people in the ceremony itself.

Rumour and Possession

The three types of events which may provoke spirit possession may also provoke spates of fantastic rumour. The following are instances of rumours which circulated in the late 1960s, at a time when particularly vigorous efforts were made to rebuild the shrine and to revive certain features of ancient organisation.

According to one of these stories Mbona had sent a monkey from the sacred forest to Lundu with the message that he, Mbona, wanted a wife from the paramount as had been customary in the past. Lundu, however, not only declined to grant the request, but he also had the monkey put to death. Because of this atrocious behaviour all monkeys are said to have left the sacred forest, which in turn was interpreted as a sign that Mbona himself had withdrawn from the shrine and that some serious calamity would follow. In fact, and not only in legend, Lundu had indeed received several requests from the shrine organisation to re-institute the old custom of providing a spirit wife; he had, however, refused to comply on the ground that such a thing was no longer possible in modern times. Perhaps the killing of the monkey also had some substance: someone may have seen such an animal killed near Lundu’s village. However this may be, the rumour itself was clearly being structured like the myth of Mbona’s death; only the sequence was reversed. Instead of the killing preceding the request for a woman it now came after the request. This again expresses another and more fundamental difference, viz. that the original Lundu had been penitent, whereas the present incumbent had not.

Another rumour was that fishermen on the Shire River had dragged up a small iron box from which emerged a little man with a pronounced limp. When asked how he got his limp, he said he had broken his leg walking through gardens which had been ridged, and that he was now on his way to the District Commissioner to tell him that this practice must be discontinued. In another version of the same rumour Mbona was supposed to have declared (presumably through the medium) that he would no longer bless the gardens, because he might break his leg due to the ridges.

According to a third rumour, the waters of the sacred pool had turned red. This was taken as an omen of worse things to follow. Finally, there circulated rumours to the effect that the medium had already become possessed, and that he had spoken out on several issues concerning the cult and the region.

The body of rumour of which these particular instances formed part had at least four distinctive effects. They undoubtedly contributed to a revival of the people’s awareness of the cult on a regional level. Moreover, some of the rumours at least seemed to represent a collective effort by the people to establish unanimity on issues of general concern. Furthermore, they appeared to provide much of the pressure which finally generated the possession seance. Finally, they also determined the contents of the message which the medium was to communicate.

As to the spatial aspect of these rumours, only a general estimate can be given. I made no effort to establish with any degree of accuracy how widely they had spread, but the fact that they were repeated to me in locations some ninety miles apart may serve as a general indication of their appeal. Nor was it difficult to establish that these rumours were part of, and in their own way contributed to, a revival of people’s interest in the cult. Indeed, for several weeks there was talk of little else. As for the causal connection between the outbursts of rumour and the possession event, particular attention needs to be paid to the rumours concerning the medium, since these provide the most explicit indication of such a connection. The very fact that it was widely rumoured that possession had already taken place meant that it was part of people’s expectations in this situation. In addition, those rumours purported to broadcast what the medium had said and thus expressed what people expected him to say.

I was not able at the time to establish whether information about his own alleged prophecies did in fact filter back to the medium. If it did, as I suppose was the case, he never mentioned it. On the other hand, it became clear in the course of several interviews with him that he very much shared the concerns which they expressed. His complaint, however, was the quite significant one that the people remained too divided and too indifferent for him to force the principals and their fellow chiefs into action. He said this primarily in connection with the reinstatement of Mbona’s wife, but he obviously meant it to apply to other issues as well. The significance of this complaint is that it suggested, at the very least, that the medium himself was conscious of the fact that he could not operate without the support of the people in general.

Peter Lienhardt (1975: 130) has recently argued that “rumours of the more fantastic sort represent or may represent complexities of public feeling that cannot be made articulate at a more thoughtful level”, and furthermore, that “they join people’s sympathies in a consensus of an unthinking or at least uncritical kind”. Whilst agreeing that, in the’case described here, rumour very much seemed to have the effect of establishing and expressing widespread consensus on certain issues, I do not think that this consensus was of an unthinking or uncritical kind. Nor would I agree that these issues could not be made articulate at a more thoughtful level. Indeed, it is not difficult in our case to show that at least some of the issues had been the subject of articulate discussion. This was not only so with regard to the reinstatement of Mbona’s wife, which had been endlessly discussed by the shrine officials, but also with regard to the ridging of the gardens, the pros and cons of which were continuously being assessed by villagers and agricultural instructors. What the rumours did though, was to establish an alternative level of discussion, where quite different, non-rational arguments were brought into play, and where the emphasis shifted from technical to social considerations. It is possible and indeed most likely that a number of people whose attitude to modern agricultural methods was not altogether unfavourable, were thereby swayed in the other direction, but even so, one hesitates to apply such epithets as ‘uncritical’ and ‘unthinking*. It would be more realistic to view rumour in this case as part of an ongoing process of self-assessment in which the people of the region tried to come to terms with factors of change and division within its boundaries.

The Conceptualization of a Regional Community

In the foregoing description an explicit distinction has been made between the formal and informal operation of the Mbona cult. Formally, it has been suggested, the cult operates through a leadership which, although divided by participation in different states and administrations, is regarded in the cult as one elite group. From this viewpoint, then, the cult may be said to support a system of social differentiation in which a relatively small aristocratic group sees itself as possessing an unquestionable right to political rule. Our discussion of the informal operation of the cult has, in a sense, reversed this picture by showing that in certain situations and under certain conditions the interests of the people could be made to prevail over those of the principals and the chiefs. The key figure in this process is the medium who has recourse to possession and who could use it, given sufficient popular support, to effectively confront the traditional political establishment.

In this final section further aspects of the contrast between modes of operation can now be considered. First, it has a pronounced ideological content which has to be made explicit here. Second, the social antagonism enacted through it has to be seen more broadly. This antagonism is not confined to the chief-subject relationship: it extends to antagonism of every kind which affects the population as a whole. To pursue the first point, it is useful to start with the striking ideological differences in the bodies of cultic myth which are characteristic of the group of principals and chiefs on the one hand, and of the medium and the people on the other.

The myths of the principals and chiefs are fundamentally concerned with history, the legitimation of their positions and, in a more general way, the legitimacy of social inequality by discrimination between aristocrats and commoners and between autochthons and immigrants. The myths of the medium and the population, by comparison, have a pronounced ahistorical character. This is pointedly illustrated by the fact that the present medium, for all his influence and importance, has almost no knowledge of or interest in the history of the cult. His interest is entirely in the present, notably in the deeds of principals and chiefs and the manifestations of public discontent. It would be fairly easy for the medium to acquaint himself with the details of history, as easy as it would be for any inhabitant of the district. The fact that he does not avail himself of this opportunity and that he feels no need to do so reflects a deep-seated aversion. This is readily understandable, given his identity: he himself belongs to an immigrant family, and gains no legitimacy from cult history.

What has been said of the medium equally applies to the people at large. Their myths reveal a similar lack of concern with history and a similar emphasis on Mbona’s relevance to the present condition of society. They differ further from the myths of the chiefs in that they project an image of a region which is unified on the basis of common concerns and a common morality. Distinctions between chiefs and commoners or between autochthons and immigrants are left unmentioned and are implicitly denied. Whereas the myths of principals and chiefs are cast in a purely traditional idiom, those of the medium and the people at large have increasingly come to be expressed in biblical language. Mbona is now said to be the ‘Son of God’ (Mwana wa Mulungu) and the ‘black Jesus’ (Yesu wakuda), who is the guardian spirit of all ethnic groups, and not just of the Mang’anja. The fact that the older traditions do not mention Mbona’s father, a logical correlate of a matrilineal system of succession and inheritance, has been converted into a myth of virgin birth which once again de-emphasizes the ethnic factor. The Christian bible has thus provided an external idiom to which neither Mang’anja nor immigrant can lay exclusive claim, and on the basis of which both may unite.

The concept of a non-differehtiated society is put forward as the ideology of a region through the formal organisation of the cult. For it is the formal organisation which defines that region as a spatial entity and makes it visible to itself through its ritual. Paradoxically, therefore, the very organisation which asserts social difference and privilege becomes the vehicle through which the denial of difference and privilege is affirmed. There occurs here & functional transposition which is activated, as we have seen, by a concurrent awareness of social and environmental threats. In the light of this observation, climatic and environmental calamities, or the fear of such calamities, are not only to be viewed as causes of cultic revival but also as activators of an alternative mode of conceptualizing its constituency.

A particular quality of this process of conceptualization is that, in principle, it concerns itself with all instances of social inequality which affect the population as a whole, but that it subsumes them all under the chief-subject relationship. The same suggestion has been made earlier in different terms: i.e., that the communication transmitted by the medium in the possession sequence is always centrally concerned with the behaviour of the chiefs and with the relations between chiefs and people. The chiefs are held accountable for the existence or persistence of almost every kind of social grievance, whether this springs from central government policy, from the actions of the chiefs themselves or one or other section of the people. That this should be so is partly due to the extensiveness of the chiefs’ jurisdiction and partly to the ambiguousness of their authority, which makes them both the representatives of the central administration vis-a-vis the people and the representatives of the people vis-a-vis the central administration. In situations of social discontent they thus become, from the viewpoint of the population, either the ones who have been instrumental in causing the discontent, or the ones who are in a position to eliminate the cause of discontent, or both. It is this viewpoint which finds its dramatic expression in the encounter between the medium and the principals.

It would be gratifying at this point to conclude the analysis, were it not for one unresolved question: What explains the situation in which the medium is close to the interests of the principals, as apparently was the case between 1930 and 1946? In a wider context this raises questions about the structural position of the medium. Given the chief-subject relationship as a constant factor in the region’s organisation, how is it that at one time the mediumship appears to support the social differentiation which this relationship implies, whereas at another time it manifests itself as decidely antagonistic to the same differentiation? My explanation would be that this depends primarily on the degree to which the functioning of chieftainship is problematic in the perception of the people. Such was the case, for instance, after 1946, when the credibility of the chiefs had increasingly come to be questioned in connection with the federation issue and the introduction of agricultural reform. This, as has been noted, made it difficult, if not impossible, for any medium to maintain close ties with the principals and, through them, with the chiefs. The structural position of the medium, then, is changeable, and not fixed. It would, however, be incorrect to view it as oscillating between the interests of the traditional political establishment and those of the people, for it has to remain within the orbit of the latter. Only insofar as public sentiment gives assent to the chiefs’ performance of their office can the medium support them and the chieftainship. Whenever this is the case, one would expect also that the contrast between the formal and informal operation of the cult and between the two conceptions of a region would become submerged. The structural position of the medium would thus be an indicator of the degree to which tensions within the region have become acute.

Essentially, what has been maintained in the present essay is that the Mbona cult functions as a mechanism by means of which the inhabitants of a particular geographic area perceive themselves as constituting a region. This perception springs from two different sources, each of which endows it with an entirely different character. One of these sources is the formal organisation of the cult which projects the image of a hierarchically structured region, and which thereby underscores the legitimacy of social differentiation. The other sources are widespread feelings of anxiety and discontent with issues that cannot be effectively dealt with by isolated groups. These feelings are characteristically focussed on the chiefs, and they are cast in a symbolic idiom which in various ways denies the legitimacy of social differentiation. The situation which ultimately emerges is that of a co-existence of two opposing conceptions of a region and of an alternating movement by which now one then the other conception is given emphasis.[79]

References

©J M Schoffeleers 1977

Notes on Contributors

van Binsbergen, Wim M. J.

Born 1947, the Netherlands; educated at University of Amsterdam (Municipal), “Kandidaatsexamen” (1968), “Doctoraalexamen” (1971).

Assistant Lecturer, later Lecturer, in Sociology, University of Zambia, 1971–73; Research Affiliate, Institute for African Studies, University of Zambia, 1973–74; WOTRO Research Fellow, University of Amsterdam (Municipal), 1974–75; Acting Professor of African Sociology and Cultural Anthropology, University of Leiden, 1975-

Author of ‘Durkheim’s Begrippenpaar “Sacre Profane”’, in Kula, 1968; ‘Saints of the Atlas: Ernest Gellner’, in Cahiers des Arts et Traditions Populaires, 1971; ‘Ritual, Class and Urban-Rural Relations: Elements for a Zambian Case-study’, in Cultures et Developpemement, 1976; ‘The Dynamics of Religious Change in Western Zambia’, in Ufahamu, 1976; ‘Law in the Context of Nkoya Society, Central Western Zambia’, in Law and the Family in Modern Africa, ed. S.A. Roberts, 1976; ‘Social Change and Intergenerational Conflict among the Nkoya of Western Zambia,’ in Cultures et Developpement, 1976; Explorations into the History and Sociology of Territorial Cults in Zambia’, in Guardians of the Land: Essays on Central-African Territorial Cults, ed. J.M. Schoffeleers, 1977; ‘Religious Innovation and the Deprivation Thesis, with Special Reference to Zambia’, in African Dossiers, 1977; Shrines and Ecstasy in the Social Structure of North-West Tunisia (forthcoming), and Explorations in Zambian Religion (forthcoming); editor (with R. Buijtenhuijs), Religious Innovation in Modern African Society, 1977.

Colson, Elizabeth

Born 1917, U.S.A.; educated at University of Minnesota, A.B. (1937); A.M. (1940); Radcliffe College, A.M. (1941), Ph.D. (1945).

Assistant Social Science Research Analyst, War Relocation Authority (1942–43); Research Assistant, Harvard University (1944–45); Senior Research Officer, Rhodes-Livingstone Institute (1946–47, 1956–57, 1962–63); Director, Rhodes-Livingstone Institute (1948–51); Senior Lecturer, Manchester University (1951–53); Associate Professor, Goucher College (1954–55); Associate Professor and Research Associate, African Studies Program, Boston University (1955–59); Professor, Brandeis University (1959–63); Visiting Professor, Northwestern University (1963–64); professor, University of California, Berkeley (1964-); Fellow, Center for Advanced Study in the Behavioral Sciences, Stanford (1967–68); Fairchild Fellow, California Institute of Technology (1975–76).

Author of The Makah, 1953; Marriage and the Family among Plateau Tonga, 1958; Social Organization of Gwembe Tonga, 1960; The Plateau Tonga, 1962; The Social Consequences of Resettlement, 1971; Tradition and Contract: the Problem of Order, 1974; Autobiographies of Three Porno Women, 1974; co-editor with Max Gluckman, Seven Tribes of Central Africa, 1951; “Spirit Possession among the Tonga of Zambia” in Spirit Mediumship and Society in Africa, eds. John Beattie and John Middleton, 1969; and various other articles.

Eickelman, Dale F.

Born 1942, USA; educated at Dartmouth College, A.B. (1964); Institute of Islamic Studies, McGill University, M.A. (1967); University of Chicago, M.A. (1968), Ph.D. (1971).

Assistant Professor of Anthropology, New York University, 1971-;

Fellow, Center for Middle Eastern Studies and the Committee for the Comparative Study of New Nations, University of Chicago, 1973–1974; Member, Institute for Advanced Study, Princeton, 1967–77.

Author of ‘Musaylima: An Approach to the Social Anthropology of Seventh Century Arabia’, Journal of the Economic and Social History of the Orient, 1967; Is There an Islamic City? The Making of a Quarter in a Moroccan Town’, International Journal of Middle East Studies, 1974; ‘Islam and the Impact of the French Colonial System in Morocco’, Humaniora Islamica, 1974; Moroccan Islam: Tradition and Society in a Pilgrimage Center, 1976; ‘Time in a Complex Society: A Moroccan Example’, Ethnology, 1977, (In press); and other papers.

Garbett, G. Kinglsey

Born 1935, England; educated at London University, B.Sc(Soc) 1957, University College of Rhodesia, 1958–1959; Manchester University, Ph.D (1963).

Research Assistant in Social Anthropology, Manchester University, 1961–1962; Lecturer in Social Anthropology, University College of Rhodesia, 1963–1965; Lecturer in Social Anthropology, Manchester University, 1966–1974; Senior Lecturer in Sociology, Aberdeen University, 1974–1976; Reader in Anthropology, Adelaide University, 1976-

Author of Growth and Change in a Shona Ward, 1960; ‘The Land Husbandry Act of S. Rhodesia’ in African Agrarian Systems, ed. D. Biebuyck, 1963; ‘Religious Aspects of Political Succession among the Valley Korekore’ in The Zambezian Past, eds. R. Brown and S. Stokes, 1966; ‘The Rhodesian Chief’s Dilemma’, Race, 1966; ‘Prestige, Status and Power in a Modern Valley Korekore Chiefdom, Rhodesia’, Africa, 1967; ‘Spirit Mediums as Mediators in Korekore Society’ in Spirit Mediumship in Africa, eds. J. Beattie and J. Middleton, 1969; ‘Circulatory Labour Migration in Rhodesia: Towards a Decision Model’ in Town and Country in Central and Eastern Africa, ed. D. Parkin, 1975.

Marx, Emanuel

Born 1927, Munich; educated at The Hebrew University M.A. (1958); University of Manchester Ph.D. (1963).

Research Associate, University of Manchester 1960–63; Senior Research Fellow, 1963–67; University of Manchester; Simon Fellow, University of Manchester 1969–70; Lecturer and Head of Department of Sociology, Tel-Aviv University 1968–69; Associate Professor, Tel-Aviv University 1971- Head of Research on Bedouin Settlement, Desert Research Institute, Ben Gurion University 1976-

Author of Bedouin of the Negev, 1967. Some Sociological and Economic Aspects of Refugee Camps on the West Bank, 1971 (with Y. Ben-Porath); The Social Context of Violent Behaviour, 1976, and various papers.

Schoffeleers, Matthew

Born 1928, Netherlands; educated at Montfort Theological College, Netherlands; St. Catherine’s College, Oxford (Nuffield Scholar), Diploma in Social Anthropology, B.Litt., D.Phil.

Pastoral and educational work in Malawi, 1955–1970; Senior Lecturer in Social Anthropology, University of Malawi, 1971–75; Reader (Lector) in Religious Anthropology, Free University, Amsterdam. 1976-

Author of “The History and Political Role of the Mbona Cult among the Mang’anja” and co-author of “The Resistance of the Nyau Cult to the Catholic Missions in Malawi” in The Historical Study of African Religion, eds. T.O. Ranger & I. Kimambo, 1972; “Livingstone and the Mang’anja Chiefs” in Livingstone Man of Africa, ed. B. Pachai, 1973; “From Socialisation to Private Enterprise: History of the Nomi Labour Associations in the Nsanje District of Malawi, 1892–1963”, Rural Africana, 1973; “The Interaction between the Mbona Cult and Christianity” in Themes of the Christian History of Central Africa, eds. T.O. Ranger & J. Weller.

Thoden van Velzen, Bonno

Born 1933, The Netherlands; educated at the University of Amsterdam (M.A. and Ph.D;).

Research assistant of WOSUNA (Netherlands Foundation for the Advancement of Research in Surinam and the Netherlands Antilles), 1960–1963; Research Fellow of the Afrika-Studiecentrum, Leyden, 1966–1971; Professor of Cultural Anthropology, the National University of Utrecht (1971-).

Author of ‘Residence, Power Groups and intra-societal Aggression’ (with W. van Wetering), International Archives of Ethnography, 1960; Politieke Beheersing in de Djuka Maatschappij (Political Control in Djuka Society),

1966; Land Scarcity and Rural Inequality in Tanzania (with van Hekken) 1972; ‘Robinson Crusoe and Friday: Strength and Weakness of the Big Man Paradigm’, Man 1973; ‘Staff, Kulaks and Peasants: A Study of a Political Field’ in Socialism in Tanzania, vol. II, eds. Lionel Cliffe and John Saul, 1973

Werbner, Richard P

Born 1937, U.S.A.; educated at Brandeis University, B.A. (1959); University of Manchester, Ph.D. (1968).

Research Assistant, University of Manchester 1961–62; Assistant Lecturer, University of Manchester 1962–66; Lecturer, University of Manchester 1966–1974; Senior Lecturer, University of Manchester 1974-

Author of ‘Atonement Ritual and Guardian Spirit Possession among Kalanga’, Africa, 1964: ‘Federal Administration, Rank and Civil Strife among Bemba,’ Africa, 1967; ‘Consitutional Ambiguities and the British Administration of Royal Careers among Bemba,’ in Law in Culture and Society, ed. Laura Nader, 1969; ‘Local Adaptation and the Transformation of an Imperial Concession in North-Eastern Botswana’, Africa 1971; ‘Symbolic Dialogue and Personal Transactions among Kalanga and Ethnology, 1971; ‘Sin, Blame and Ritual Mediation’, in The Allocation of Responsibility, ed. Max Gluckman, 1972; ‘The Superabundance of Understanding: Kalanga Rhetoric and Domestic Divination’, American Anthropologist, 1973; ‘Land, Movement and Status among Kalanga of Botswana’, in Essays in African Social Anthropology, ed. Meyer Fortes, 1975; and other papers.

Author Index

[Page numbers don’t correspond to this version]

Subject Index

[Page numbers don’t correspond to this version]

Note: References are listed at the end of each article, and these lists are not indexed separately.

[Back Cover]

ASA MONOGRAPHS

Reently published

13. Social Anthropology and Medicine

edited by

J.B. Loudon

University College of Swansea, Wales

November 1976, xxvi + 600pp.

0.12.456350.3

14. Social Anthropology and Law

edited by

Ian Hamnett

University of Bristol, England

April 1977, x + 234pp.

0.12.322350.4

Volume in preparation

15. The Anthropology of the Body

edited by John Blacking

Department of Social Anthropology

Queen’s University. Belfast, Northern Ireland

Academic Press

London — New York — San Francisco

A Subsidiary of Harcourt Bruce Jovanovich, Publishers

24–28 Oval Road, London NW1 7DX, England

111 Fifth Avenue, New York, NY 10003, USA

Australian office: PO Box 300, North Ryde, NSW 2113, Australia

[1] This paper is based on fieldwork in Boujad, Morocco, and its environs from October 1968 until June 1970. It was sponsored by the Foreign Area Fellowship Program and the National Institute of Mental Health, to whom I am deeply grateful. Additonal fieldwork ensued in the summers of 1972 and 1973 and was sponsored, respectively, by the Foreign Currency Program of the Smithsonian Institution and the Social Science Research Council. The transliteration of Arabic terms in this text is based upon colloquial usage. The plurals of most Arabic words are indicated here by the addition of to the singular forms, except for those few which frequently occur in Western literature. However, to avoid awkward circumlocutions, the maraboutic descent group discussed in this paper are Sherqawi in the singular and adjectival forms, Sherqawa in the plural, as they are in Arabic. Because of publishing economies, I have been unable to indicate emphatic consonants by the use of special transliteration marks. For comments on this paper I wish to thank Karen I. Blue, T.O. Beidelman, Nicholas S. Hopkins, and especially Richard P. Werbner.

[2] Although rarely used as such marabout is an English word. I deliberately use this term rather than saint to avoid facile analogies to the Christian context. Although both North African marabouts and the saints of Mediterranean Europe serve as intermediaries to the supernatural and frequently are associated with specific places and clientele, they occur in significantly different cultural contexts which cannot usefully be equated, as Turner has persuasively argued (1974: 56–71).

[3] For similar analogies concerning saints in the European Mediterranean, see Boissevain (1966: 30–31) and Christian (1972: 44).

[4] The acquisition of clientele independent from immediate ecological considerations continues to be the case. In recent years a Boujad Sherqawi active in electoral politics (when they are possible) has successfully sought maraboutic status in certain villages in the Sus area and has collected substantial donations. These are provided in part by wealthy Casablanca merchants who originate from these villages and seek to facilitate their business enterprises in return. Other villagers sought to become his clients on the basis of less practical “religious” ends.

[5] This baraka, as these attributes are popularly considered, can be acquired equally from kinsmen and nonkinsmen. The “closeness” through which these qualities are transmitted is frequently envisaged in highly concrete terms in maraboutic myths. In one case, bread is prepared by a marabout and eaten by his disciple. In another instance, a Sherqawi child sucks the fingers and toes of a sleeping marabout.

[6] I specify currently because a number of other sons of the marabout are recognized, some of whom have shrines in Boujad. Both in the past and today, Moroccans recognize the possibility of other descent groups being “discovered” (or sociologically speaking, of descent groups crystallizing out of descent categories).

[7] For a further discussion of these brotherhoods, see Eickelman (1976: 224–227).

[8] The same descriptive term designates pilgrims who visit maraboutic shrines.

[9] A similar process takes place in the formation of urban quarters (Eickelman 1974b). By rough calculation, there is a maraboutic shrine for every six square kilometres, or one for every 150 persons, on the western plains of Morocco. Comparable figures prevail for the High Atlas mountains (Berque 1957: 7). It is futile to place much faith in such a census, but it does indicate the prevalence of maraboutism. Of course, the vast majority of such shrines draw only miniscule offerings and clientele.

[10] Because Jacques Berque’s interpretation of maraboutic cults is the most prominent one in French literature on Morocco, it is useful briefly to analyze it here. While Gellner relies upon the logically elegant imagery of segmentation to describe “tribal” social structure, Berque elaborates a botanical metaphor to conceptualize Moroccan social forms, especially those of High Atlas Berbers. For Berque, these forms resemble “a hopelessly tangled undergrowth, formed of miniscule twigs, with the roots of each extending to all points of the horizon” (1953: 263). Having extracted full value from the botanical metaphor, Berque proceeds to an incendiary one to describe historical shifts in the distribution of power: “tribal hegemonies, personal power, spiritual or dynastic expansion flame up and die out like torches. Their radiation lacks depth. It is spread out over a milieu diverse and diffuse at the same time, fragmented but universal” (1953: 269). Berque sees maraboutic cults as an integral part of Moroccan social identity from at least the eighteenth century to the present (1953: 270). The title of the article, “Qu’est-ce qu’une ‘tribu’ nord-africaine” is probably meant to evoke Renan’s “Qu’est-ce-qu’une nation?” (1887) and thus suggests the significance which Berque attributes to the study of “tribal” social identities. Unfortunately, however, his general argument, as opposed to his Ethnography of the Seksawa (1955), is limited to a presentation of his key metaphors and an analysis of prior scholarly assumptions toward local and tribal identities. Yet in various other writings, Berque acknowledges the persistence of maraboutism in diverse historical and social circumstances (e.g. 1949: 82–84; 1957).

[11] Field work in South Sinai was carried out between 1971 and 1976. I spent a total of eight months in the field. In 1973–4 Miss Shuli Hartman participated in the research and carried out field work. The Ford Foundation, through the Israel Foundation Trustees, supported the project and this help is gratefully acknowledged. The South Sinai Administration, and especially Moshe Sela, greatly facilitated my work. The comments of the participants in the 1976 ASA Conference, particularly Ernest Gellner, Raymond Firth, Shuli Hartman and Richard Werbner, were of great help in the revision of this essay. Map I is at the end of the article on page 49.

[12] Fieldwork was carried out among Zezuru of Musami, Mrewa for three months in 1958 and among Korekore of the Sipolilo and Darwin Districts over a period of 25 months during 1960–61 and 1963–64. I make grateful acknowledgements to the Leverhulme Trustees, to the International African Institute and to the Research Council of the University College of Rhodesia, for generous support.

I particularly wish to express my debt to the late George Kupara, the ‘Lion’ of Mutota, who welcomed me to his seances and patiently instructed me in the ways of the Mutota cult.

In writing, I have used the ethnographic present, occasionally, for convenience. However, it must be borne in mind that throughout the seventies the whole of the Zambezi Valley has been the scene of intense military activity. The Korekore are now very much in the front line of the struggle for an independent Zimbabwe. Sadly, many have been concentrated in ‘fortified villages’ and a number of mediums of the cult of Mutota have come under the close scrutiny of the colonial military. The medium of the Korekore Nehanda fled to Mozambique to escape their attentions and one, the medium of Chihwahwa, was imprisoned for 20 years, allegedly, for aiding guerillas. The senior medium of Mutota ended his days in a village surrounded by security officers who vetted every supplicant who came to him.

[13] I was not able to enter Mozambique to establish how far the cult extended. Korekore say that the cult’s region extends to the northern escarpment of the Zambezi valley. I was able to establish that the spirit provinces of Nebedza and Nehanda did extend into Mozambique.

[14] Often spoken of as ‘Lions of the Heavens’ (Mhondoro dze Matenga) and called ‘supra-tribal’ in much of the literature (see Gelfand 1959; Abraham 1966).

[15] All names, unless otherwise stated are of rivers or streams. With the exception of the Bwase (‘Gwase’) River, all streams and rivers which appear on official maps have been named as they appear there. Many are wrongly recorded. For example, in Chikorekore, Kadziruve is Kadzuruve, Ambi is Hambe, Hunyani is Manyami, Mahuwi is Mauhwi, Katsawa is Katsahwe, etc.

[16] Boundaries were established by working from one land-shrine community to the next and asking to whom the “beer of the earth” (hwahwa hwe pasi) was offered. Forest boundaries were established by joining hunting parties and noting the different spirits to which hunters made symbolic offerings of leaves and twigs. The eastern part of Map 1, including the chiefdoms of Mzarabani and Kasekete and the west bank of the Msengezi, was mapped in 1960–61. The remaining areas were mapped in 1963–64.

[17] References under A refer to Genealogy 2 and under B, to Map I. Under A, T refers to Tavara, Ch. to Chthonic, S to Stranger. The affiliation of Mukarazi was not fully determined. Almost certainly, this spirit forms part of the Chikunda cult of Kanyemba.

[18] Mutota dies because his sons conspired together to prevent him knowing that he had reached the Zambezi Valley. He was thus placed in mystical danger. As he died, he stated that unless one of his sons had intercourse with his daughter, Nehanda, his body could not be buried. Eventually, Nebedza slept with his sister. The ground then opened miraculously to receive Mutota’s body. From this incestuous union all present day Korekore are held to derive.

[19] I have distinguished chthonic beings from ordinary Tavara chiefs in terms of the powers accorded to them in the myth by Korekore. Tavara chiefs are described as being killed by ordinary, physical means; the chthonic beings are always overcome by mystical means. Some Tavara do not make this distinction and claim the chthonic beings as their ancestors. In some accounts Musuma and Nyamukokoko are distinguished as two beings. Korekore in the Dande hold that Musuma Nyamukokoko was a son of Dzivaguru who was raised in Chimombe’s village. Later, Chimombe is said to have moved from the Dande, leaving Musuma behind, to land beyond the Hunyani and Angwa rivers.

[20] In the past, mediums were forbidden to cultivate and were maintained by tribute in grain given by members of their spirit provinces. Even now a medium obtains a dispensation from his spirit before he begins to cultivate.

[21] An ngozi is an afflicting ancestral spirit. In most areas of the Valley, one will find ngozi possession cults. Their organisation tends to cut across the land shrine communities. A candidate for ngozi possession is seated on a mat and senior adepts of the cult will rub his head vigorously, from time to time, to aid possession.

[22] An unpossessed medium is referred to as Novana. Often, a medium speaks of himself as a container or pocket (homwe) for the spirit.

[23] Abraham (1959) has presented a composite of several genealogies collected throughout the Zambezi Valley. In this, segmentation occurs earlier. Nevertheless, only Nebedza, of the 16 children of Mutota given, is credited with a line of descendants (see also Beach 1972).

[24] Mediums do not drink immediately before or during seances. The spirit is held to enter the stomach. Drinking water would both cool and displace it.

[25] I attended 40 seances at which 74 separate issues were raised by 63 supplicants. Supplicants together with their companions and retinue numbered in all, 153.

[26] Field work in the Tapanahoni area of Surinam took place from June 1961 till November 1962, during August 1965, in September and October 1970 and again for a few weeks in December 1973 and January 1974. The first and second periods were financed by The Netherlands Foundation for the Advancement of Tropical Research (WOTRO), the third period by the Afrika-Sudiecentrum at Leyden. My stay in 1973–1974 was supported by the University of Utrecht. I feel indebted to Chris de Beet and Diane Vernon for valuable criticism of an earlier draft and for drawin attention to documents in French and Dutch archives.

[27] The name Gaan Tata has fallen into disuse. Nowadays, the Djuka usually refer to Gaan Gadu or Bigi Gadu (Thoden van Velzen 1966a). Kersten, a missionary who visited the Tapanahoni in 1895 (MBB 1896: 185), noted that the name Gaan Gadu was also used by high priest Labi Agumasakka. Other names mentioned in the earliest accounts of the movement are Masa Jehovah and Bakaa (MITT 1895: 52), respectively ‘Sir Jehovah’ and ‘European’. I have preferred to use the name Gaan Tata as this distinguished the movement from the other High God or Great Deity cults that periodically erupted in Bush Negro territory; adherents of these cults often called the central divinity Gaan Gadu.

[28] In 1885, gold-diggers were willing to pay 32 florins ($12.40) per barrel, the unit of freight. Ten years later, freight prices on the Albina-Lawa route were down to 10 florins ($4), to skyrocket again to levels of 32,50 and even 50 florins during the second gold rush (1901–1904) in the Lawa basin (De Beet and Thoden van Velzen 1977). The colonial administration repeatedly attempted to curb price rises by putting pressure on tribal chiefs. They had to wait for the decrease in the demand for transport, before these efforts were successful.

[29] The attitude of government officials tdwards the cult varied considerably. Usually, at the turn of the century, the higher echelons showed themselves opposed to any aggressive policy towards the Gaan Tata priests. However, officials at the lower levels at times supported the missionaries in attempts to suppress the cult. Such divergencies enabled the Gaan Tata priests to operate an oracle at Santigron, fifteen kilometres from the capital.

[30] But of course some clashes did take place: at Moarimbomofo in 1892 (MBB 1893: 179–180), 1894 (MITT 1896: 70–71); at Dritabiki in 1900 (MITT 1901: 598), at Koffiekamp in 1892 (MBB 1892: 367–368) and in 1893 (MITT 1895: 317–318); and again at a Cottica village (MBB 1907: 62–66) where Djuka men and women dragged their kinsman Kwassi Hendriks from the hands of a missionary who was about to baptize him.

[31] Modern ethnography has been deplorably silent on the history of the cult, its spread throughout the interior and its present-day functioning.

[32] Personal communication, Chris and Miriam de Beet.

[33] ‘Outsiders’ is my translation for Bakaa. All those who are neither Bush Negroes nor Amerindians are called Bakaa by the Djuka.

[34] Quite healthy Gaan Tata cults continued to exist in other areas as well. In 1962,-Agiti-Ondro on the Cottica River was thriving with patients and supplicants. To a lesser extent this, also seemed to have been the case at Santigron and Toti-Kampu on the Saramacca River. Gaan Tata worship also continued at several places on the Suriname River.

[35] Apparently, only the Tapanahoni’s parochial cult was destroyed. Other parochial Gaan Tata cults are thriving.

[36] This paper is based on my research into Central African religious change, in which I have been engaged since 1971. Fieldwork was undertaken alternatingly in Lusaka and Kaoma district, Western Province, Zambia, during 27 months, from February 1972 to April 1974. Of this period, rural research was carried out in February 1973, May-June 1973 and August-April 1974. Research both in Lusaka and Kaoma district concentrated on the Nkoya ethnic group but in addition involved religious and political organizations of a much wider scope. A research grant from the University of Zambia covered initial research expenses in the period February-April 1972. In 1973–74 I was a Research Affiliate of the University of Zambia’s Institute for African Studies, in which capacity I benefited from its various research facilities. The Netherlands Foundation for the Advancement of Tropical Research (WOTRO) provided a writing-up grant for the period 1974–75. The University of Leiden paid towards my participation in the 1976 ASA conference, which led to the present paper. While registering my indebtedness and gratitude vis-a-vis these various instutitions, the Zambian authorities and my informants, I wish to thank in particular the following people: Denes Shiyowe for excellent research assistance; Henny van Rijn, my wife, for sharing much of the fieldwork and analysis; the late Max Gluckman, Andre Kobben, Maud Muntemba, Bob Papstein, Terry Ranger, Mathew Schoffeleers, Jaap Van Velsen, and Richard Werbner who over the years through comments and discussions, have contributed to my analysis; and finally Richard Werbner for his most stimulating editorial efforts towards the production of the present paper.

[37] Christian churches are cults in terms of my definition of cult. However, they would be distinguished from cults in the dominant sociological approach to types of religious organization, which has been developed on the basis of the writings of Troeltsch, Von Wiese, Becker, and Yinger (cf. Kolb 1964; O’Dea 1968).

[38] The following associations between non-regional cults of affliction and ethnic groups are recognized in Western Zambia: the songo cult is associated with the Songo people of Angola (McCulloch 1951), and introduced in Western Zambia by the Luvale (Ikacana 1971), chimbandu with the Angolan Ovimbundu (McCulloch 1951); kayongo with the Luvale (Kuntz 1932) or in general with Angolan immigrants (my fieldnotes); kalendamawulu with the Lozi (my fieldnotes); blyaya or slyaya with the Luvale (Ikacana 1971, my fieldnotes); liyala with the Lozi (Kuntz 1932; Gluckman 1951); macoba with the Kaonde (Kuntz 1932); mayimbwe with the Nkoya (Kuntz 1932; Gluckman 1951; Ikacana 1971; my fieldnotes); kasheba with the Kaonde (my fieldnotes); muba with the Totale (Gluckman 1951; Ikacana 1971, who also mentions the neighbouring Toka); Lenje (my fieldnotes), or even groups further to the east, like the Lamba (Doke 1931) and the Ambo (Stefaniszyn 1964).

[39] “Anamnesis” is used here in the technical, medical sense of a patient’s own account of the history of his complaint.

[40] In this context one should bear in mind that most cult leaders are women, who in extra-religious public settings are not supposed to sit on chairs.

[41] This might be read to constitute a moral aspect, in contrast with what I have said above about the amoral nature of these cults. However, this concern with fair play points in a different direction—just as the ethics of fair play do not turn soccer or the retail trade into a predominantly moral institution.

[42] Mongu district files; the report could not be consulted.

[43] Reynolds (1963: 133) mentions the date of 1944, Muntemba (1972: 1) 1940.

[44] Initially a jar of chalk, and church bells (Reynolds 1963: 135, 137); later also a sacred cup for the administering of medicine (Muntemba 1972: 5, 11).

[45] The paraphernalia which Simbinga brought from his journey, and which to this day are in possession of his widow (one of my chief informants), were mainly the following. In three differently-coloured bottles (allegedly brought from Johannesburg by somebody else, and for secular purposes) Simbinga had collected water from three rivers: the Zambezi,.Jhe Kabompo and the Lungwebungu. These bottles he used as divining device: only a patient who managed to indicate which bottle contained water from which river could be considered to suffer from Bituma, and thus be eligible for treatment. In addition Simbinga brought a round, grey stone (about 10 cm diameter); a mpande shell partly covered with copper-wire; a flyswitch made of eland-tail; a number of cowrie shells; a genet skin (mbomba) a circular ornament, made of parts of white water-lily sown into strings of white beads; a small copper bell of European manufacture; and finally a leather-bound copy of an Afrikaans hymnbook (Di Berijmde Psalms, 1936). For the healing ritual, Simbinga would dress in fully white clothes. Although I cannot elaborate on this point here, to one familiar with the material culture and symbolism of Western Zambian cultures it will be clear that Simbinga’s choice of paraphernalia was largely based on widespread conventional symbolism in this area, and reflected personal idiosyncrasies only to a limited degree—except in their combination.

[46] In the 1969 census Mongu had 10,700 inhabitants, Livingstone 43,000 inhabitants, and the other centres mentioned less than 4,000 inhabitants (Davies 1971: 126).

[47] Despite increase in population, the relative demographic difference between the Bituma and Nzila regions was of essentially the same nature as in the 1930’s, when Mongu and Livingstone had already been provincial and national capitals respectively for several decades, and (cf. Merle Davis 1933, appendix map) population density in Western Zambia showed variations similar to the present-day distribution.

[48] Map 2 gives, inter alia, an overview of the size, growth and distribution of Christianity in Western Zambia between the years 1929 and 1948, the crucial period for the emergence and early development of Nzila and Bituma. Basic figures were derived from: Northern Rhodesia, 1929, 1937, 1948.

[49] I began my study of this cult in the course of fifteen months fieldwork in Zimbabwe during *1960–61, and continued the fieldwork in Botswana for the following periods: 1964–65, fifteen months; 1969, three months; 1974, one month. I wish to thank the Fulbright Commission, the University of Manchester, and the Social Science Research Council (U.K.) for grants and leave for the research. I am grateful also to the many Kalanga and Khurutse, cult officials and laymen alike, who helped me to understand the cult, especially the southwestern priest, Mr. Ntogwa Matafeni Dube, Reverend Mongwa Tjuma, Mr. Mpubuli Matenge Ngulube and their families. Public personalities and places referred to here are historically significant and too readily recognisable to be disguised. I have therefore not used pseudonyms, as I usually do to protect personal identities.

[50] Included in the cult are such ethnically different people as Kalanga (Holub 1881; O’Neill 1920; Richards 1942; Sebina 1947; Gelfand 1966), Karanga (Knothe 1888; von Sicard 1952; Daneel 1970, 1971), Venda (Schwellnus 1888; Stayt 1931; van Warmelo 1940) Ndau (Daneel 1970: 57), Khurutse (Shapera 1971), Ndebele (Campbell 1926; Ranger 1967), and others. I am not sure whether Namzwa to the north, near Wankie, have congregations.

[51] Cockin, a missionary at Hope Fountain in 1879, mentions three oracles in the Matopo Hills (Cockin 1879, cited in Ranger 1967: 145), as follows, relative to his mission station, Ematjetjeni (Matonjeni) to the east, Entjeleli (Njelele) to the south, and Umkombo to the south-west. (I am grateful to Mr. Richard Brown for supplying me with Cockin’s full text.) Another oracle, contemporary with them, is mentioned by Holub for the northwest in 1876 at the Maitengwe River (Holub 1881: 64). Note that all these oracles continue into the twentieth century. Ranger misreads Cockin’s report floc. cit.). He eliminates one of the central oracles; he mistakes a lesser oracle outside the central Matopo Hills near Mangwe for a cardinal oracle. This misreading leads Ranger into an ad hoc view of the cult in its response to change; he obscures continuity in the cult.

[52] The cult’s domain also coincides, roughly, with the distribution of Khami-type Iron Age buildings (with the exception of Mtoko, in the east). (For a map see Summers 1961: 9.) The somewhat wider distribution of Zimbabwe-type buildings extends beyond the present cult domain into the Korekore-Zezuru areas to the north, where cults of local and territorial spirits have their domains (see chapter 3 above).

[53] This smaller zone of wooded steppe with abundant Acacia and Comiphora has somewhat higher rainfall.

[54] I discuss an apparent exception below, page 200.

[55] My assessment is on the basis of a preliminary study of the cult of komana, which I carried out in 1972–3 during a year’s fieldwork among Tswapong in central Botswana.

[56] At a cardinal oracle of Matonjeni, two offices are distinguished: that of shrine keeper, customarily a Venda, and that of priest, customarily from the Shoko (Monkey) clan among Mbire. This distinction is not maintained in the western regions. Lumbi, minstrel, is the term commonly used for the priest Ntogwa, who also spoke of himself as ‘Leader’, ntungamili. Similarly, here unlike at Matonjeni, the term hosana is used for both male and female adepts.

[57] Newman (1971) records, ‘In the Matopos Hills in Rhodesia there is a remarkably dense population [of eagles], possibly t(ie most concentrated eagle population known anywhere in the world’ (59). Note that snake eagle is a friend of man in that it preys on snakes. I am tempted to suggest that the two figures of the famous Zimbabwe figurine may be a snake eagle or a martial eagle and its prey, a monitor lizard.

[58] According to Kalanga tradition, the oracle of Mwali came from: 1) Lutombo Lutema to 2) Bambudzi, 3) Zhomba, 4) Chizeze, 5) Mavula Majena, 6) Njelele, 7) Dula, 8) Manyangwa, 9) Njenjema, 10) Ntogwa.

[59] In the region that I know best, the priest Ntogwa succeeded his mother’s sister’s son, Fulele, and also claimed that Njenje was his maternal grandfather. He claimed Venda origin for his patriline in a Zebra clan (Dube or Ntembo). Njenjema, a priest of the northwest, was of Njenje’s clan, a Monkey (Shoko or Ncube, praise name Luvimbi). Manyangwa, the northern priest, is a Leya.

[60] Daneel (1974: 109) stresses hostility and opposition between Zionist churches and the cult staff at the cardinal oracle of Matonjeni, though he also mentions more co-operative relations between the cult and Ethiopian-type churches. In Botswana the southwestern priest, Ntogwa, co-operated in various ways with leaders of the Bethannia Mission Apostolic Church and J. E. Mtembo’s Church of Christ.

[61] A messenger explained the distinction between him and the porter thus, “I carry the truth; he carries the blankets.” In fact, this messenger was slightly deaf, and at a public meeting, the porter gave most of the report about their pilgrimage to their regional oracle. It is common for a porter to become a messenger, in turn.

[62] Ranger (1967: 378) reads his own interpretation, somewhat mistakenly, into a cult lament which Mr. Timon Mongwa Tjuma transcribed for me, and I translated with his help. The text is not ‘a despairing lament for the ruin created in the Shona world by the white man and the powerlessness of the defeated to do anything about it’. Nor does it express ‘a mood of despondency’ (Daneel 1970: 35). It expresses moral outrage and complaint—indeed, a refusal to be defeated—as shown by the chorus and remarks at the time. The prevailing attitude of protest and complaint in cult songs is explained thus by Kalanga themselves, ‘A child must cry so that his father will heed his suffering.’

[63] At a different time, and on a domestic occasion, an adept may act as a host of Mazenge and be possessed by a bush spirit. (See Werbner 1964, 1971, 1972 and also Daneel 1970: 51 on Jukwa possession in the cult of Mwari.)

[64] Her Husband, Shologulu, (The Great Head), dwells in the sky to the south, Her Son, Lunji (The Awl), to the west; and She Herself, to the east. She and Her Son go back and forth and communicate with each other and Shologulu via a shooting star. Note that Kalanga bury a corpse with its orientation in accord with the trinity thus: its face towards the south, its head to the east, and feet to the west, along the sun’s path.

[65] When an elder, after reciting Her praise poetry, professed to me the conventional piety that She has one breast because that suffices for all, another elder interjected, ‘Tell him the truth—it’s because she is too stingy that she has only one.’ Her parts are half a person’s: one eye, ear, nose, arm and leg. (See also Cockin 1879 for an early missionary’s report.)

[66] Note that Kalanga, both from the cult’s domain and outside it, also approach the Tswapong oracle as supplicants; some Kalanga claim Tswapong ethnic origins.

[67] The fund is conceived of as one whole from the congregation. Neighbours know, at the time, who contributes, since the collections are made or announced at public meetings. But it is taboo to make a permanent record of this public knowledge, by writing down a list of donors and contributions.

[68] A close kinsman of a deceased cult messenger once boasted to me about the debt his kinsman incurred through such embezzlement.*

[69] Another is at a cardinal oracle in the Matopo Hills, which is called Dula (Granary), apparently after the shape of its rock. In the past, during the youth of the oldest living elders, valuables were stored at the senior cardinal oracle, Njelele, including durables such as hoes, and consumer goods, i.e. tobacco, hides, cloth. A ruler or locally important elder was, on occasion, given a hoe from the store.

[70] i.e.: Chichi, Mapokani, Vaka, Mpini, Makwela, Jeshen, Madlambudzi.

[71] i.e.: besides the six shown in detail on Map 4 (including Letsholathebe outside the Tati Reserve), Mpapo, Nkange, Dagwe, Goshwe, Sinete, Mzonga, Matobo, Tonota.

[72] Vumbu was, like his father, a dancer of hosho (the gourd rattle) and a singer of its somewhat satirical ballads. This dance is performed for recreation and ridicule at rituals and also beer parties. It is the dance Mwali prefers to the Gumba-Gumba and pop music on records. See above page 187. Note also that Galani, too, never became possessed or wore a black sash, though she otherwise dressed and danced as a hosana.

[73] From north to south in 1974: Mathamngwane, Tonota, Mmadinare.

[74] This was formed about 1972 under Letsholathebe, Senior Headman in the official grade and a chief in my terms.

[75] I am grateful to Mr. Harry Finnigan for the Botswana Daily News references. In reply to my enquiries and at Mr. Finnigan’s suggestion, the News reporter, Mr. Mishingo Mpaphadzi, kindly wrote for me a further, unpublished account.

[76] Other oracles apparently under Mtonjeni are these two: 1) the oracle sometimes known as Magubu (perhaps a hereditary title) which has been southeast of Fort Usher and under Mtabani then Magubu (Richards 1942: 55; ‘Majuba’, Gelfand 1966: 35f; Maguhu, Daneel 1970: 41) and 2) the oracle at Shangonyima/Jahunda in turn under Tapa, Mike, Kamba (Knothe: 1888; Schwellnus: 1888, Posselt 1927: 530f; Richards, ibid.). The first oracle is the one that figures so largely, indeed too largely, in Ranger’s account (1967: 149); the second has been especially important for Venda in the Transvaal and in the Gwanda District south of the central highlands.

[77] Ranger (1967: 187) misreports that I was ‘told that [a] high priest had been shot at Mangwe in 1896 by a fellow American’. Despite his assertion, it is not so that ‘There can be little doubt that the man killed and called Jobani or Tshobani was the High Priest of Mwari for the southwest, remembered by the Kalanga today under the name Habangana (my italics).’ Kalanga told me that during the war, Armstrong murdered Chief Bango’s messenger, his parallel cousin. Hobani (no connection with Botswana’s Chief Habangana of the same totem, Zebra, but not of the Bango clariy, Zviposa, the priest, continued to perform near Mangwe, clandestinely, even after the war. His oracle was a junior southwestern one. In 1960/61, I observed that Chief Bango sent adepts and a messenger to the cardinal oracle of the priest Maswabi at Dula in Kubutu cave.

[78] Ranger (1967) misses the general accord in the crescent from Chilimanzi south and eastwards and including Gutu, Chibi, Ndanga along with others, because he isolates each instance and treats as exceptional any congregation such as Ndanga’s which opposed the war (see Selous 1896: 238).

[79] The fieldwork on which this paper is based was carried out in 1967–8 and in 1973–4. The earlier part was financed by a Nuffield sociological scholarship; the latter by a grant-in-aid from the University of Malawi. I wish to record my indebtedness to both organisations and to my research assistant Mr. E. C. Mandala.