Shrines, cults and society in North and Central Africa
	Wim van Binsbergen's webpage on Khumiriyya (N.W. Tunisia), late 18th - mid-20th century

return to the index page of Historic Berber culture

(c) 2006 Wim van Binsbergen

Wim M.J. van Binsbergen

Shrines, cults and society in North and Central Africa

A comparative analysis

paper read at the Conference on Regional Cults and Oracles

(Annual Conference, Association of Social Anthropologists)

Manchester, March - April 1976.

Contents:

1. Introduction.

2. The Khumiri of Tunisia. 2.1, Segmentation; 2.2. Shrines; 2.3. The ecstatic cult; 2.4. Historical and political aspects.

3. The Nkoya of Zambia. 3.1. The Nkoya village; 3.2. The village shrine; 3.3. The valley and its shrine; 3.4. Non-ancestral cults of affliction; 3.5. Historical and political aspects.

4. Comparison of the Khumiri and Nkoya data.

5. Conclusion: towards a theory of communal and non-communal cults in their social context.

Appendix 1: Systematic comparison of Khumiri and Nkoya society, shrines and cults.

Notes.

References cited.

1. Introduction[1]

In recent years a considerable amount of research has been carried out on cults and oracles, in various parts of the world and within various theoretical approaches. Much of this work is represented at this conference. We are confronted with a growing need to compare notes, particularly with the aim of arriving at such theoretical explicitness and accumulation as are still lacking in this field. For a century or more, anthropologists, historians and students of comparative religion have been writing on cults, oracles and on shrines that tend to constitute the material focus of both. Yet, so far, we have (to my knowledge) little in the way of useful and general definitions, and less of a systematic theory.[2] This is unfortunate especially since many of us believe the phenomena studied at this conference to play a pivotal role in society.

                        The purpose of this paper is not in the least to go ahead and propound the general, systematic theory I hope we are all waiting for. I merely intend to search for a descriptive and comparative framework out of which such a theory may one day evolve. I shall use that simplest form of comparative analysis, in which an anthropologist compares two societies in which he has himself collected extensive data on the subject under study. Comparative studies move in and out of fashion, and anyway one could imagine a more sophisticated comparative approach than the one I adopt here (Köbben 1966, 1970). Simple comparison is used here mainly as a heuristic device, to highlight some of the theoretical problems, and the potential, in this field. For this purpose, it is immaterial that the two societies (the Khumiri of the highlands of N.W. Tunisia, and the Nkoya of the woodland plateau of central W. Zambia) have been selected on rather trivial ground: because I happen to know them.

                        To these limitations in theoretical pretensions and ethnographic scope, must be added those concerning subject matter. I propose to concentrate on the social-organisational aspects of shrines and cults, and largely ignore moral and symbolic aspects. The latter were intently studied in my researches. However, social-structural aspects lend themselves more easily to generalised, summary description, and cross-cultural comparison. Behind this lurks however an enormous problem which I can only briefly indicate here.

                        Religion, including the types concerning us here, takes shape within the social process that goes on between signifying, interacting, moralising and manipulating individuals. It is one of the greatest promises in the religious anthropology of the last few decades, that it has begun to formulate approaches - e.g. in Victor Turner’s work (1957, 1967, 1968, 1969) - which seek to interpret ritual in terms of the shifting and often inchoate social process, involving a highly specific setting of individuals whose interrelations and biographies are traced in great detail and time depth (extended case method)             This signals the emergence of alternatives to the hitherto dominant, Durkheimian heritage of interpreting ritual by reference to an abstract and rather immobile, total social structural order (Durkheim 1912; cf. van Binsbergen 1976a). When it comes to cross-cultural comparison, the very specificity of an extended-case approach to religious phenomena presents complex methodological problems, for which no solution has yet been worked out. For this reason, I shall presently remain within the safe grounds of the dominant tradition, and study religious and non-religious phenomena primarily as clearly definable, persistent and comparable institutions, elements of social order. This illusion of monolithic units of analysis conceals how the actual social process (as well as history on a larger scale) in various parts of the world works at cultural and environmental material so as to produce more variation, manipulation and innovation of both meaning and action, than the comparativist can ever take into account. If it is the specific, local social process, instead of the generalised institutions, that really matters in religion, then an institutional approach, comparative or not, is likely to yield insights in social structure more than in religion. Some of us have accepted this limitation gladly (e.g. Gellner 1969: 138 and passim; cf. van Binsbergen 1971b: 208f). Others have indicated that here, precisely, lies the main shortcoming of current religious anthropology (Geertz 1966: 42). Anyway, the problem is there.

                        I shall now proceed to describe the social-structural aspects of shrines and cults among the Khumiri and the Nkoya, in preparation for the comparative and theoretical parts of the argument. The unavailability, as yet, of comprehensive accounts of my researches in English, necessitates rather extensive description here. The reader pressed for time may wish to skip sections 2 and 3, glance at appendix 1, and move on to section 4.

2. The Khumiri of Tunisia

2.1. Segmentation

The Khumiri highlands (Van Binsbergen 1970, 1971a, 1971c; Jongmans 1968! 1973; Demeerseman 1964; Souyris-Rolland 1949; Creyghton 1969; Hartong 1968) form a system of narrow valleys, separated by wooded maintain ranges. Each valley contains a small number of villages (50 - 250 inhabitants each). Villages are surrounded by fields, and separated from each other by stretches of forest. Within each village, smaller units define themselves by: spatial clustering; by particular amenities (house threshing-floor, spring, men’s assembly ground (raquba) exclusively used by one unit; by collective economic action (horticulture, animal husbandry, food preparation); by common recreation; and by common identity and a name. The smallest of these residential units is the individual house. A few houses together constitute a compound, which has its own threshing-floor. A few adjacent compounds constitute a hamlet, which has its own spring. A few neighbouring hamlets constitute a village, which has its own raquba. This amounts to a system of territorial segmentation. On various hierarchical levels, well-defined and non-overlapping units exist which are in segmentary opposition vis-à-vis similar units at the same level; on a higher level these opposing segments are united as belonging to one higher-order segment. Spatial lay-out and visible, exclusive attributes from house to raquba wholly define the system of territorial segmentation from the smallest unit up to the village level. Above the village level, two more local levels are operative in the contemporary Khumiri society: the valley (segmented into a number of villages) and the chiefdom, a colonial creation segmented into a number of valleys. On still higher levels lies the government administration, with its districts, provinces, and the national state.

                        The spatial arrangement of dwellings and the exclusive use of the characteristic attributes by their segments, demonstrate that this model of territorial segmentation corresponds to an implicit structuring in the Khumiri conceptual system. Yet the Khumiri’ explicit participants’ model of social organisation is a different one. All territorial segments bear names, derived from an ancestor: Mayziya (from Bu-Maza), Zrayqiya (from Zarruq) etc. Participants view their social organisation in terms of a segmentary system of patrilineal descent groups. Mythical ancestors associated with the highest segmentary levels (chiefdom, valley, village) are, through a putative genealogical chain of an unspecified number of generations, linked to the historical ancestors (who died within four generations of the now living). The participants’ ideology provides a skeleton model in which they rarely press for specific genealogical and residential details, and which is therefore allowed to be utterly non-consensual, contradictory, and historically inaccurate. The present-day function of the Khumiri unilineal segmentation model is mainly to comment on actual social relations. Claims with regard to rights of residence, land and assistance are peripheral to the unilineal model’s application: these claims primarily spring from actual, close relationships and current transactions, featuring the total bilateral kindred. Crucial relationships are only secondarily underpinned by reference to a genealogical charter, which is sufficiently flexible to be manipulated and altered along with the rapid shifts in the pattern of actual relationships. Such shifts are brought about by the competition for honour and informal leadership, factionalism, individual geographical mobility, marriages, people engaging in or dropping dyadic[3] contracts, etc. Genealogical manipulation, aided by the ideal of patrilateral parallel cousin marriage,[4] tends to conceal the actual historical diversity of people’s origins under the convenient fiction of common patrilineal descent. Why should this fiction exist at all? It provides an idiom of kinship obligations, to support the actual need for close co-operation between members of territorial segments at various levels. At the lower levels (up to the hamlet), members of segments have common interests in the fulfilment of vital economic needs: domestic budgeting and food preparation (house), horticulture (compound), water (hamlet). Beyond this, there are common interests in terms of social relationships. Politics and conflict settlement mainly take place at village level and centre on the raquba.

                        In addition, local men, as bride-givers and bride-takers, share an interest in local women, most of whom marry within the valley and even within the village (50%). Marriage is mostly (95%) virilocal. Adult men frequently move to join more profitable dyadic partners as neighbours, but these removals are mostly intra-village. Nowadays, most men, and almost half of the women, live and die in their village of birth, amongst people most of whom they have known all their lives.

                        Whether viewed as territorial or as unilineal descent segmentation, the segmentary alignment of groups, and the distribution of characteristic attributes at each segmentary level, is therefore pertinent to the total economic and social dynamics of Khumiri society.

                        As elsewhere in the world, segmentation is process as much as it is structure. Environmental, demographic, economic and local political factors determine at which segmentary level a given, named unit will function. As these factors vary over time, units acquire or lose personnel and functions, and this demands a redefinition of their segmentary relations vis-à-vis neighbouring segments. Segments are known to develop within a few decades from e.g. compounds to villages, thus dissociating themselves from the village of which they originally formed part. Likewise hamlets and villages can dwindle to compound and even house level. The outwards signs of these segmentary dynamics are not only changes in the spatial lay-out of houses within a valley, but also a redistribution of the other characteristic attributes: threshing-floors, springs, raqubas will be abandoned, or new ones created, as beacons in the altered segmentary structure.

2.2. Shrines

Shrines[5] come into this overall segmentary structure in a variety of ways.

                        In the Khumiri highlands, shrines vary from a venerated tree or source (marked by rows of stones and pieces of textile tied in knots), via miniature huts built of large stones and arboreal material, to the white-washed, domed qua well-known from throughout the Islamic world. For the Khumiri, shrines are associated with deceased human beings: saints. For almost any local shrine people can quote the name of the saint associated with it. A simple myth links saint and shrine: allegedly he lived there, was buried there, stopped there in the course of a journey, or the shrine was created upon relics brought from other shrines having such mythical connotations. Myths may also link various saints associated with the shrines in a valley: a saintly pantheon of close kinsmen and servants. But such myths are used inconsistently and without consensus, and they comprise only a selection of the many shrines in a valley. Therefore these myths only roughly represent the hierarchical, tree-like structure of segments by an analogous model of mythical relationships between the segments’ saints. The myths quote as contemporaries of the saints, other saints and mythical ancestors.

                        Saints associated with local shrines lived in the mythical past, they rarely feature in the skeleton genealogies, and are not supposed to have descendants amongst present-day people (even if analytical historical reconstructions suggest the contrary). Yet people refer to the saints by kinship terms: jaddi (grandparent), uboi (father), sidi (elder brother), lalla (elder sister) - the standard terms of address also used for unrelated, senior living people. Dealings with the local saints are usually Characterised by an excited fondness rarely seer. in the dealings between living kinsmen. Shrines feature as additional characteristic attributes of territorial segments from the lowest level upwards. A minority of houses have a tiny shrine as their characteristic segmentary attribute. Most compounds, and all higher-level segments up to the valley level, have a neighbouring shrine attached to them. No shrines are attached to the higher administrative levels from the chiefdom upwards.

                        The association between shrine and segment takes a number of expressions. Individual households in the segment dedicate meals to the saint. Individuals enter into contractual relationships with the saint: they promise him an offering (candles, incense, a flag, an animal sacrifice) in exchange for a major saintly service (curing from illness and barrenness, protection of domestic animals, a job). Members of the segment may visit the shrine individually, particularly on occasion of childbirth, marriage, departure for military service, illness. Twice a year the women of the segment pay a collective visit to the shrine, directed by the senior women, who also co-ordinate the day-to-day economic activities by which the segment is bound together.

                        Village and valley shrines have a twice-yearly festival, when the shrine structure is repaired, and sheep and cattle are sacrificed and distributed. On this occasion all male members of the segment assemble for some days around the shrine to ‘recreate’ themselves with tea drinking, card playing, music and ecstatic dancing (see 2.3.). Around the same time, three sets of women pay compulsory visits to the shrine: all present female members of the segment; all women born into this segment but since married or otherwise migrated to a different segment at the same level; and women who have belonged to a different segment since birth but who have entered into a personal relationship with the shrine’s saint at one stage in their lives (e.g. a visit to cure barrenness) - or who have inherited such a relationship from their mothers or mothers-in-law.

                        In years of extreme drought another collective shrine ritual can be enacted for the shrines at the highest segmentary levels: an animal sacrifice to request rain.[6]

                        Only valley shrines have a shrine keeper. He performs a ritual at the shrine at Thursdays, co-ordinates the festival and receives a share of the offerings. Ideally a close agnate of a former incumbent to this office, the shrine keeper is usually recruited from among a Somewhat larger set, including close matrilateral and affinal kinsmen of a predecessor. Raqubas are always located within view of the major village and valley shrines, and the latter’s saints are frequently invoked to strengthen oaths; in very grave cases parties will visit the shrine to swear upon the sacred objects therein.

                        Finally, many shrines attached at levels from the hamlet upwards are surrounded by cemeteries to bury, again, three sets of people: present members of the segment; people born into the segment but emigrated; ant people born elsewhere but tied to the shrine and the segment by a family tradition, based on emigration in a near preceding generation. Usually, the cemetery and the festival function of the shrine are spatially apart. One higher--level segment has then at least two shrines attached to it; these may be associated with the same saint and bear the same name.

Khumiri shrines appear to have the following social-structural functions.

                        They reinforce the system of territorial segmentation. This is a dialectical function, integrative and divisive at the same time. Within one segment, the shrine ritual clearly makes for integration. It supports the demarcation of the segment’s social group vis-à-vis the outside world, reinforces the collective identity, corroborates (though not: legitimates)leadership, makes for a religious focus and religious collective action in addition to the collective action focused on amenities (from house to raquba) with direct economic and social connotations. But true to the segmentary principle, all this implies divisiveness with regard to the relation between two neighbouring segments at the same level. Myths emphasise rivalry between saints whose shrines are of equal segmentary importance. One cannot visit or celebrate their shrines on the same day. Members of rival segments compete over the power and splendour of their respective saints, shrines and festivals; and negative stereotypes between rival segments are phrased in reference to their saints and shrines. On a higher level, however, this divisiveness tends to be balanced by the fact that rival segments together constitute a higher segment, at the village and valley level, where they are united again by a common shrine and common ritual. This is all the more important since on the segmentary level of the valley, no non-religious characteristic attribute exists as focus of collective activities and identification (cemeteries providing a borderline case). However, beyond the valley level even shrine cults fail to create much integration of highest-level segments; thus valleys and chiefdoms oppose each other as segments without effective means of identification and collective action at any higher segmentary level.

                        In so far as the Khumiri shrine cult stipulates not only ritual for members of segments collectively, but also ritual behaviour of individuals vis-à-vis shrines of other segments than these individuals belong to at any level, the shrine cult also cuts across the system of territorial segmentation. Festivals, home-pilgrimages of out-married women, optional visits of individuals, and incidental burials at distant shrines, bring together individuals of non-adjacent segments who would otherwise have very little opportunity to come into contact and develop more permanent relationships.

                        These cross-cutting contacts are very important. Although the segmentary arrangement of relationships (in which geographical distance is a crucial factor) regulates most aspects of social life, there remains a need for more distant, dyadic contacts L22 which may offer resources locally not available: friendship untroubled by day-to-day close scrutiny of partners and third parties; marriage partners; information on and access to economic opportunities and specialist services elsewhere. Cross-cutting shrine ritual helps to open up and sustain this interlocal field of social relations. Moreover, cross-cutting shrine ritual forms the main opportunity to remain informed on kinsmen in distant places, and to assert one’s claims on them. The pilgrim from an other valley not just visits the shrine but also his (more typically her) kinsmen. He imposes himself as someone who has latent claims locally, who belongs there too. This is particularly the case for an out-marrying woman. Khumiri marriages rarely end in divorce, and women rarely remarry. In theory a woman has a right on her parents’ land equal to her resident brother’s right.[7] These rights are allowed to become dormant when she marries into a different segment (even if within the same village). But by coming back, annually, to her original segment’s shrine, and bringing her children, she asserts these rights, and introduces her children to their maternal uncle who in future may not be able to turn them away if need forces them to ask his support. A similar argument applies to those who keep visiting a distant shrine in view of a personal relationship with the saint, or a memory of migrating from his segment in a previous generation. At the same time, the distant pilgrimage implies a challenge to the local segment to which these pilgrims now belong. It brings out that they are not wholly dependent upon their present segment, that they have still claims and obligations (both ritual and otherwise) elsewhere. Thus the distant pilgrimage assumes the character of a declaration of independence which is resented by the fellow-members of the pilgrims’ present segment. Hon -pilgrimages (reinforced by the idea of heavy supernatural penalties in case of neglect) form almost the only occasion, the only pretext if you like, on which husband and in-laws allow an out-married woman to go and visit her kinsmen in her original segment. Other visits, lacking such supernatural sanctions, are taken as indications of her marriage breaking down. Similarly, men are bound to their segments on sanction of arousing the saint’s anger in case of emigration. Cases of affliction in men and women are often diagnosed[8] as caused by a saint who feels slighted by people moving away from him or failing to celebrate his festival.

                        Clearly this set-up has great potential for the symbolic enacting of cleavages in the local group and amongst kinsmen, in the idiom of the saints of present and original segments. This falls outside our present scope. I should however emphasise that no local saint (nor any other supernatural agent except the High God, Allah), is believed ever to exert direct moral censure on the interaction between the living.

                        The saint is considered a powerful ally, who can be mobilised through ritual and promises of ritual. He can be insulted by perjury, attacks on pilgrims, failure to honour promises to him - and then he will strike. Not visiting his shrine invokes punishment, but neglecting obligations vis-à-vis people living near his shrine does not. Symbolically, the saintly cult does provide an idiom to express and mend breaches in human relations - but indirectly so.

2.3. The ecstatic cult

Every member of Khumiri society engages in the cult of local saints, although women much more frequently and deeply than men. In addition to this cult complex, a sizeable proportion of adult men, and much fewer women, have been initiated into a religious specialism (faqir, druish) whose main manifestation is their entering into trance or ecstasy during public, ritual musical sessions. Upon his initiation, every faqir has selected for himself one cult song (triq) out of about twenty available in the area. Only with this song and the proper musical accompaniment will he be able to enter into trance. The texts feature local saints, more universal Islamic saints, and other supernatural entities venerated throughout the Islamic world (Sidi ‘Abd el-Qadir, Sidi Hamad, Umm’ Lehana, the Prophet Mohammed, Allah). This cult complex is the Khumiri variant of Sufism (Islamic mysticism). Most fuqra are considered members of one of several Islamic brotherhoods (foremost the well-known Qadiriyya) that have regional centres or lodges (zawiyya) in the Khumiri highlands and in the town of El Kef,[9] and which have ramifications throughout the Islamic world. Nowadays, these cults’ organization is extremely weak locally. Few fuqra have in fact visited the lodge and have been initiated there. Most have acquired their status and skills through personal intuition of members in the course of sessions held in the villages; only a minority of such mentors maintain themselves close relations with the lodge.

                        Village faqir sessions are held on two occasions: to add splendour to saintly ritual, and to diagnose and treat affliction. On all occasions the more experienced fuqra, while in trance, may divine on such issues as saintly intervention, imminent death, and disrupted relationships within the segment organising the session. When affliction is the occasion, the session may start off a process by which one afflicted is gradually initiated into faqirship. Underneath such affliction (marked by psychosomatic, neurotic or psychotic symptoms) severe psycho-social tensions can be identified, bearing on the patients’ immediate consanguineal and affinal relationships. However, such crises are neither a necessary nor a sufficient condition for faqirship.[10]

                        Outside the lodge fuqra mostly perform alone or in very small numbers. Fuqra have no common identity, never act as a group, and display no preference for intermarriage or for dyadic contracts amongst themselves rather than with non-fuqra. But the ecstatic ritual does Catalyse other group processes. Faqir session are freely attended by all members of the segment that has invited the faqir; the session has a rallying function. The lodge is a building located in the village where the descendants live of the founder of a local branch of a brotherhood. Though most fuqra are now outside lodge control, there is still a rudimentary Organisation operating in the area. In each valley near the lodge there is one representative who frequents the lodge and the establishment at Le Kef. this representative regularly visits the villages in his area to collect small donations and occasionally stage faqir sessions. The lodge leader does the same; in addition he has a large practice as a diviner, diagnosing affliction as caused by saints or land spirits (jenun). He also stages the actual treatment for the more severe affliction cases. This treatment takes place twice a year at the lodge. Patients from all over the area, who have previously visited the lodge for diagnosis, come and bring animals for sacrifice; they will spend the night being led into ecstasy, in a bid to allay their possessing spirit. In the lodge ritual, local Khumiri saints are ignored and all attention focuses on universal Islamic saints, spirits, the Prophet, and Allah .

2.4. Historical and political aspects

Khumiri shrines and the ecstatic cult can only be understood against the background of universal Islam - of which Khumiri religion is a localised, popular version not unlike the ones encountered amongst peasants in other Islamised parts of the world.

                        The proliferation of faqirship independent from the lodge; the substitution, outside the lodge, of universal Islamic supernatural agents by local Khumiri ones; the weak integration of local faqirship into the international brotherhood Organisations: these are all manifestations of a dominant feature of Khumiri society - the strong tendency towards local particularise and autonomy. Distant parent units (be they distant segments from which people emigrate; distant lodges; or distant parental shrines) are only for a short time acknowledged in a hierarchical relation vis-à-vis their social offspring.

                        Not unlike the High Atlas as described by Gellner (1969), Khumiri society is the scene of a continuous battle between ideology and historical truths: the actual history of diverse origins, migrations, dependence, offspring, - upon which ideology imposes a system of segmentary opposition and integration, claiming common origins, mutual equality between segments performing similar functions, competition for relative autonomy and equal segmentary status irrespective of historical, genealogical ties. In such a segmentary context, religious power and religious foci apparently tend to be atomistic and unrelated.

                        Local saints and independent fuqra are more in line with the local social structure than universal Islamic supernatural agents, and an international formal brotherhood Organisation. The same pattern applies to the saints themselves. In so far as they were historical local residents,[11] they seem to have been immigrant, adherents to a more strict and formal version of Islam than prevailing in the Khumiri highlands: representatives, mainly, of the various Islamic brotherhoods that have time and again attempted to replace, in the rural Maghrib (N.W. Africa), the local popular version of Islam by their own, more formal variant (Bel 1938; Brunel 1926; Dupont and Copolani  [ check Coppolani ] 1898; Montet 1909; André 1956; Anawati and Gardet 1961). Caught in the dynamics of local Khumiri society, these pious strangers after their deaths ended up as cornerstones of the very popular religion they came to transform.

                        Once the saints, through shrines, had become attached to segments, the localising principle again explains much of the specific local history of the distribution of shrines over contemporary segments. Detailed historically study of a few adjacent valleys since the early 19th century reveals how this distribution was brought about by fission, fusion and migration mainly in the last century. Small family groups would, for political and ecological reasons, leave their localised clan segment and settle in a different valley. They would be in need of a shrine, as a focus of identity and ritual. In case they immigrated as clients, dependent upon their hosts, they would have to orientate themselves towards their hosts’ local shrines. But when they arrived as an autonomous immigrant group, securing land rights through purchase, exchange or conquest, then they would create a new shrine, often as branch of the shrine of their parent segment. In the latter case the shrine would be erected upon relics brought from the parent shrine, and be given the same name. Initially, immigrants of either type would keep visiting the festivals of their parent segment but as the immigrants became more securely settled locally, this contact would be allowed to lapse, and one would concentrate on local festivals as a sign of the immigrants’ incorporation into an other valley. As a group’s size, power and wealth increased, its festival and shrine would be made more important and take precedence over other shrines in the valley. Likewise, decline of the local group would lead to its shrine losing prominence. In the past the segmentary, localising tendencies created local saints out of Islamic "missionary" strangers, and demolished, after initial success, such lodge-type cult Organisations as these strangers may have come to found locally. This process is likely to have been going on for many centuries. No doubt ecstatic ritual has a very ancient history locally,[12] but the main present-day Khumiri lodge was founded in the 1880s only. In its early decades, the lodge had formal representatives in many surrounding valleys. These officials organised impressive annual collections of tribute, to be sent on to El Kef via the twice-annual convention at the lodge. Moreover, some of them were in charge of the major local shrines.

                        With the limited effectiveness of non-religious conflict-regulation agencies (peace councils, jama’a, whose members were often themselves parties to conflicts), these pacifist officials intervened in battles and provided a venue for peace negotiations between feuding groups. /This has been not In less than a century, the economic and political functions of the lodge have virtually disappeared This has been not merely as an effect of localising segmentary tendencies, which are inimical to an interlocal Organisation. For in the same period, the segmentary system was confronted with a unifying, bureaucratic power complex of such military and organisational superiority, that the very dynamics of feuding, conquest and ecological competition, the backbones of the segmentary system, were cut out. Up to the French Colonisation (1881) the Khumiri had succeeded in keeping the Tunis central government out - although occasionally tribute was exacted by armed tax-collectors, who often were associated with Islamic brotherhoods (e.g. the Shabbiya). In an until then acephalous polity, the first chiefs ("sheikhs") were appointed in 1883 from amongst locals outside the politico-religious power complex of lodge and major shrines. Though backed by a colonial garrison, it took the chiefs a quarter of a century to effectively impose the central power they represented upon the segmentary system, banning warfare, and usurping the conflict-regulatory functions of the shrine keepers and lodge. As a means towards both power and prestige, the colonial and post-colonial chiefs gained control over the shrines by pushing poor kinsmen, their own clients, to be appointed as shrine keepers; and by opposing (aided by governmental regulations against the brotherhoods,[13] the lodge representatives and leaders whilst allowing their own client kinsmen to embark upon careers as village fuqra unattached to the lodge. The lodge Organisation collapsed partly in response to segmentary localisation, but mainly as a result of the segmentary system being overcome by a more powerful central government.

                        Under colonialism, the banning of warfare and the increase of population froze the segmentary dynamics to a point where migration, and assimilation of immigrants, could only take place at the lowest segmentary level: the individual household. Instead of movable tents, dwellings became immobile stone constructions, and the distribution of groups over land stuck to the situation at the turn of the century. After 1930 no new shrines were created, merely some hut shrines were embellished to become qubbas. If the modern Khumiri highlands displays segmentation, and distribution of shrines, mainly along territorial instead of descent lines, this may be largely due to the Colonisation of the area. The colonial impact upon the area’s politico-social structure is particularly manifest in the preposterous fact that Khumiri segmentation crucially affects all aspects of life except politics.[14]

                        One wonders why the lodge, deprived of most of its political and economic powers has retained and most likely further developed the affliction aspect. What are the links between Colonisation, peripheral stagnation, individual suffering and an idiom of ecstasy and mystical liberation? Whilst statistical analysis shows that Khumiri fuqra are not deprived in wealth and prestige, as compared to their non-faqir neighbours, is it possible that their performances enact a collective deprivation syndrome of Khumiri society: colonial experience internalised ? or is it rather that the affliction aspect was harmless enough from the viewpoint of the colonial and post-colonial central power, to be endured - in contrast with the lodge’s political aspects (cf. 3.5.)? Or both?

3. The Nkoya of Zambia

3.1. The Nkoya village

Nkoya is an ethnic and linguistic label applying to about 50t000 people inhabiting the wooded plateau of Central Western Zambia (Van Binsbergen 1972a, 1975a, 1975b, 1976a, 1976b, and forthcoming a, b; Clay 1945; McCulloch 1951; Mainga 1969; Anonymous n.d.). At the village level, their society shows much of the familiar Central-African pattern. In addition to migrant labour, the economy hinges on shifting horticulture with a fringe of animal husbandry wherever the fly-infested environment allows this. Villages are small: up to a hundred inhabitants, but usually a score or less. Villages continually emerge, mature and decline in response to ecological, demographic and social vicissitudes. The rural male career model stipulates the competition for village leadership and glorious titles. This causes senior men belonging to different villages to compete for patrilateral, matrilateral and affinal junior kin (no matter how remotely related) as co-residing followers. In most other Central-African societies bilateral tendencies are concealed under a formal matrilineal ideology; the Nkoya however are explicitly bilateral, which amounts t) children having equal claims concerning support, residence and land on a rather large pool of geographically scattered bilateral kindred, including remote and putative kinsmen (bathukulu, "grandfathers"). Marriage is unstable, many marriages end in divorce, and successive marriages are common practice. Therefore both women and junior men (up to their forties) tend to live in a number of villages successively, staying on for a number of years as long as they are effectively attached to senior men, and moving on (after divorce, death, disruption of good relationships with their patrons)when local support is failing or when better opportunities arise elsewhere. Adults rarely live and even more rarely die in the place where they were born. The cultural and ecological similarity in the wider area and a clan system providing means of identification beyond traceable or putative genealogical ties, allow individual geographical mobility (in the pursuit of marriage and clientship) to cut across ethnic and linguistic boundaries, over several hundred kilometres.

                        With only moderate exaggeration the Nkoya village could be characterised as a small, ephemeral conglomerate of strangers who have not grown up together, are genealogically heterogeneous, and are ready to leave as soon as misfortune befalls them locally and/or they can get better opportunities elsewhere.[15] At the same time, people thrown together in a village share vital interests: in land, the production and distribution of food, a measure of harmony in day-to-day interaction, assistance in individual life crises, conflict regulation to mitigate internal strife and to prevent sorcery, and finally the maintenance (through food exchanges, mutually visited ceremonies, and martial ties) of good relationships with other villages in order to create a pool out of which material support and personnel can be drawn in case village survival is in danger.

                        How are these vital interests served? First, the senior membership (headman, elders, elder women) spend, in exchange for economic and prestige benefits, most of their time in organising and checking the social process in the desired direction. Secondly> an ideological construction counteracts the heterogeneity and opportunism of individual village membership. In terms of this ideology, all members of the village are close kinsmen. Precise genealogical details, and other historical facts such as historical slave status of part of the membership, are suppressed. The ideology has elaborate religious aspects. Nkoya belief that in addition to the actual living membership, a village’s affairs are the concern of all deceased former members both of this village and of all villages from which ever members were drawn. Whatever their names and wherever buried, these ancestors (mipashi) are held to keep the affairs of the living in constant scrutiny and to dish out success and health, respectively failure and illness, commensurate to the living people’s performance.

3.2. The village shrine

Various actions and material substances allow for communication between the living and the death: beer and meat stock; application of white meal, clay, cloth, or water; divination; clapping of hands; prayer. The main material focus for rituals involving these elements is the village shrine (chihanda, mushuwa), an inconspicuous shrub or forked pole situated in the centre of the village, near the men’s shelter which is the organisational headquarters of the village.

                        Any restoration or enlargement of the village’s strength and unity forms occasion for a small ritual at the shrine: Childbirth, return after long absence, the tracing of a distant kinsman and potential co-resident of whose existence one was not aware, the settlement of internal conflict, recovery from illness that was diagnosed as caused by the ancestors, success in hunting, etc. Ancestors are prayed to at the shrine in cases of illness. Very rarely, named ancestors are addressed; this is only the case if divination has pointed at a recently deceased as causing the illness.

                        In all other cases, the supernatural entities associated with the shrine are addressed as a nameless collectivity, or as "Thou, our ancestor". Name-inheriting ritual represents the most elaborate collective rituals in which the Nkoya village shrine plays a part. On all other occasions only current members of the village (and migrants on visit) take part in the ritual. No outsider is under obligation to visit the village shrine and make an offering. But in the name inheriting rituals, for which a beer party and nocturnal dance are staged, members of surrounding villages in the same and adjacent valleys participate in great number (up to several hundred, if the title of a chief or senior headman is to be inherited).[16] In addition to local visitors, geographically distant members of the extended bilateral kindred of the deceased come and participate in the ritual; many of these have never belonged to this village but have latent claim of membership there, whilst others may once have belonged to the village but moved away - often because they were in conflict with the former, now deceased headman whose title they might now inherit. After long deliberations a heir is appointed by the elders so assembled, and inaugurated at the shrine.

                        The village shrine forms the main focus of village identity and autonomy, legitimates village authority, and is the material focus of a device to enforce conformity and loyalty by reference to supernatural sanctions. No village can do without a shrine. The ritual of planting a shrine by the headman makes the selection of a site for a new village definitive. New village sites are occupied on two occasions: when after the death of a headman the whole village moves away; and as a result of fission, usually followed by the attraction of geographically distant followers from other villages than the one that has split. On all occasions village shrines are created directly from forest material, and no relic from the parent village’s shrine is brought to the new site. Nor is there any other way in which Nkoya village shrines become associated with each other or incorporated into a wider structure. The nameless shrine comes into being and falls into decay along with the material structure of the village.

3.3. The valley and its shrine

The cult complex of the village shrine mainly refers to communal life. Its ecological connotations[17] are limited to hunting; horticulture, the land, rain are outside the scope of this ritual complex.

                        Ignoring such supralocal organisational devises as have been superimposed by the Lozi (Barotse) and British Colonisation of the Nkoya area (cf. 3.5.), Nkoya society has few effective organisational levels above the village level. The main named territorial unit besides the village is the valley, which takes its name from the river flowing through it. Rivers yield great quantities of fish in the wet season. Near the river, the wet gardens of the valley’s villages lie close to each other; they are rich in yield but cannot be expanded, and therefore are heavily contested. The less fertile, dry gardens are situated around the villages and further uphill, as clearings in the forest which separates valleys from each other. Villages are located halfway the gentle slopes. They are connected by paths along and across the river. Most day-to-day interaction of an economic ritual or recreational nature, and most marital ties, of a village are confined to the other villages in the same valley. While a chief is appointed over a number of adjacent valleys, each valley has a sub-chief, whose main function is the regulation of such conflicts (including occasional referral to the distant Local Court) as could not be settled within the village. Thus the valley is the highest effective social and ecological unit in Nkoya society; its inhabitants have a certain density of interaction, social and economic common interests, common identity and a name.

                        A valley shrine cult exists which has primarily reference to the valley’s ecology. The cult focuses on the grave of the most recently deceased, most senior headman or chief of the valley. The place (in the midst of the previous, deserted chiefly village) is marked by a pole.

                        In years when the first rains are delayed or stop too soon, the sub-chief and two or three other senior headmen of the valley visit the shrine, clap hands, and ask the deceased and the High God (Nyambi) for rain. To make their need clear, they pour water on the grave and display seeds or faded seedlings. No other people share in the ritual or visit the shrine at other times. No other ritual is performed for this deceased headman. Not the valley shrine, but the shrine of the last-living chief’s village features in chiefly succession ritual. No ritual (except a greeting prayer) is performed either for the other deceased headmen whose burial shrines one may come across if their deaths were yet recent enough to prevent these traces from being completely swallowed by forest regeneration.

3.4. Non-ancestral cults of affliction

Besides village and valley shrines a third type of shrines exist: shrines erected and owned by individual specialists. These specialists fall into two main categories: ecological specialists (hunters, fishermen, honey-collectors, ironworkers), and those who, as leaders or adepts, belong to one of the many non-ancestral cults of affliction existing in the area. It is on the latter type that I shall concentrate now.

                        It is relatively rare for illness and other misfortune to be attributed to troubling ancestor who can be approached through village shrine ritual. Besides such recognised causes as sorcery and forest beings, for which specialist treatments exist, a large and increasing number of affliction cases is diagnosed as being caused by vague and rather impersonal "disease principles". In the Nkoya area, at least twenty disease principles are distinguished; some major ones are: biyaya, kasheba, mayimbwe, songo, muwa, bituma. Each principle is the subject of a cult. A local senior adept of the cult diagnoses whether people’s illness has been caused by the cult’s disease principle; if so, they may provide minor treatment to alleviate the affliction, or, more typically, they stage a healing session (ngoma) during which the patient (through dancing, singing, music, sweating baths, private therapeutic conversations, manipulation of the leader’s paraphernalia, and ecstasy) is initiated to become an adept of the cult. Often the ecstatic ritual serves as both diagnosis and treatment. Through the ritual the patient acknowledges his special relation vis-à-vis the disease principle, honouring and placating the latter so as to be troubled no longer. Once initiated, the adept may continue to visit sessions and may begin to diagnose and treat others.

                        Cult sessions are held in the novice’s or the leader’s village. In the former case a newly erected cult shrine is the focus of the session. Musicians and a chorus are recruited in the same or neighbouring villages. Beer and specialist fee are furnished by the sponsor, usually a senior kinsman or the spouse of the novice, who in most cases is female. Neighbours and kinsmen may attend in smallish numbers, seldom more than a score. The whole setting suggests a social drama in which tensions and cleavages in crucial social relationships in the village and the kindred are brought to the fore and ritually resolved. This is in fact the case, but a discussion falls outside the scope of this paper.

                        The various cults follow this general pattern whilst differing in details: paraphernalia, texts and music of the songs, vegetal medicines used, form and adornment of the leader’s shrine, specific ideas on the nature and symptomatic manifestations of their particular illness principle, etc. Almost all women and many men have been initiated in one cult, many in several. All cults of this type now occurring among the Nkoya are claimed to be recent innovations, which only in the twentieth century, have been either introduced from surrounding areas or locally developed. They have spread like fashions or epidemics, quickly gaining momentum in a wide area, losing vigour soon, to be overtaken by a similar cult only a few years later.

                        A combination of characteristics enables these cults to spread rapidly, even across linguistic and ethnic boundaries, as they  [ check ]  have been doing in Western Zambia.[18] Affliction is attributed to abstract, impersonal supernatural entities, which are free from all reference to local social-structural or ritual peculiarities. Contrary to other locally prevailing interpretations of illness, the cults have an emphatically a-moral nature: misfortune is explained not in terms of malice, guilt or neglect (as in the case of sorcery and ancestral affliction), but in terms of accidental possession by an unknown entity with whom one had no relation previously. The cults therefore offer a venue for healing and collective ritual without disrupting local relations by the search for a culprit. The cult’s main symbolic idiom is easily understood, lying in line of the ancestral and chiefly ritual throughout Central Africa. Esoteric knowledge, or complex mental or bodily techniques which it takes years to acquire, do not come in. These are some of the factors explaining the great receptivity of the modern Nkoya and their neighbours to these cults of affliction. An extra impetus, on the side of the leader, form the high prestige and financial benefits; but these can only exist, of course, if for other reasons there is a need for this type of cult among the population in general.

                        Most of the non-ancestral cults of affliction described here, have only the most rudimentary organisation. This fact may largely explain why each separate cult subsides quickly. Diffusion through individual adepts who after mastering the cult’s idiom establish themselves as leaders makes it possible for these cults to expand over a large area without any \ organization above the local level. There are however minimal organisational requirements. Sufficient adepts must be available locally to form a chorus; these adepts must be mobilised to attend the session and must be prepared to accept the leader’s authority during the session. Virtually every session is devoted to recruitment of new adepts, of whom the leader keeps thinking as his own patients- although, once the fees have been paid, they will only meet at sessions if they live in each other’s proximity, and even then rarely. Whilst the sessions have a rallying function for non-adept attendants, there are no indications that adepts, as such, have a tendency to develop, outside ritual, closer relationships among themselves, or with the leader, than with non-adepts. Yet for each leader the local adepts represent an indispensable ritual following, which is jealously guarded against the encroachment of other leaders within the same or similar cults. The general pattern however allows for variation in two directions. Given the decisive role of individual adepts in the diffusion `of the cult, and the lack of a supra-local organisation, there is much room for personal innovation. Many leaders have added a personal touch to the paraphernalia and the song texts of the cults they were transmitting. In a few cases, visionary individuals reach further than that and, in the course of a personal crisis, devise what amounts to a prophetic cult of affliction. Such an innovation retains the material of the general pattern, but adds to this elements from the personal vision of the prophet: a more elaborate conception of the illness principle, particular medicines, food taboos, and myths linking all these. These prophetic versions are highly syncretistic, and combine the general affliction-cult pattern with elements springing from Christianity and local ancestral cults. Two examples from the area are Sambaing (founder of the bituma cult), and a prophet who later in life was known as Moray, after the moya illness principle featuring in his cult (cf. Ikacana 1952). Both started in the 1930s.

                        In both cults a second direction for variation can be noted: a supra-local Organisation. Once the prophetic vision has reshaped the general pattern into a very specific and elaborate cult, there is a basis for orthodoxy. The prophet and the story of his vision can act as a charter and focus for the adepts, thus lending the cult much more identity than the non-prophetic cults of affliction can derive from their abstract, impersonal illness principle.

                        Leaders of the latter cults are mainly regarded as skilful doctors who get good payment for their indispensable services; the awe-inspiring nature of the illness principle they deal with, does not emanate onto them. In the prophetic cults the situation is different. In the eyes of those adepts who got cured by the prophet or his representatives, the prophet acquires a personal charisma. This seems to form the main basis on which a supralocal Organisation is developed. Adepts in distant places are appointed representatives for the cult; they keep in touch and have to forward part of the income they derive from the cult sessions they stage. Both Sambaing and Moya set up such a supra-local Organisation, with a small number of representatives dispersed over an area of thousands of km2.[19] Nowadays the bituma and Maya organisations still exist but the former has declined enormously. An early schism within the cult between the rival factions of the founder and his cousin Kapata, has estranged most original representatives from the present leader. These dissidents no longer forward money or visit the annual convention. Neither have they retained any organisation among themselves Simbinga’s original vision means little to them now, and instead they have allowed bituma to routinise into just another, non-prophetic cult of affliction whose local leaders they are. It is thus that bituma is viewed nowadays by most adepts and other Nkoya.

3.5. Historical and political aspects.

The present-day religious situation of the Nkoya can be interpreted in the light of the area’s history over the last centuries. From the 18th century onwards, the penetration of militant immigrants and new political concepts mainly from the North, and the related expansion of regional and long-distance trade, caused a marked increase of political scale. Chiefs began to control larger areas and assumed a more exalted status. A royal cult complex focusing on chiefly paraphernalia, medicines and graves appears to have greatly expanded at that time - if it had already existed before. Among neighbouring groups (Ila, Kaonde, Kwanga) similar processes took place. And particularly among the Lozi, where the development was stimulated by the eminently favourable ecology of the Zambezi flood plain and by the Kololo invasion from Southern Africa. Much of the Nkoya area came under Lozi political influence during the last century: tribute was paid, chief’s drums (symbols of autonomy) and captives were taken to the Lozi capital, Lozi representatives were stationed at the Nkoya chiefs’ courts. The Lozi royal cult was boosted along with Lozi political expansion, and became the central focus of identity and chiefly legitimacy in the Lozi kingdom (Mainga 1972). But Nkoya chieftainship and chiefly ritual declined, due to the expansion of the Lozi and to a lesser extent, through Kaonde and Ila raids. This decline seems the main reason why today the Nkoya chiefly cult (at valley level) is a very modest affair, as compared to that of the Lozi. To this should be added that Nkoya from outside the Zambezi flood plain never participated in the Lozi royal cult -except as court musicians or occasional human sacrifices; coming under Lozi political influence did not imply adopting the Lozi royal cult. Viewed as a neighbourhood ecological cult, the Nkoya valley cult is also much less elaborate than that of neighbouring peoples (Ila, Tonga) who, lacking a pre-colonial state organization, have made neighbourhood shrines the key element in their supra-local order (Smith and Dale 1921; Jaspan 1953; Colson 1962, 1969; Fielder 1965).

                        In the second half of the last century the expansion of the slave trade and raiding (both from neighbouring groups and from distant Ndebele and Yeke) threw together in the fly-infested, less hospitable parts of Western Zambia, concentrations of refugees from a variety of ethnic and linguistic groups (Gielgud, n.d.). They boosted the villages of petty local chiefs, and gave these a considerable heterogeneity. Present-day Nkoya are largely descendants of these people. The present heterogeneity of the Nkoya village has therefore a precedent at least in those days. And so has, probably, the crucial function of the village shrine as a device to create unity, conformity, identity and authority, amongst a heterogeneous set of co-residents. This however suggests that, prior to the political expansion and general upheaval during the last century, village shrines may have played a lesser role. It is not unlikely that, before the 19th century, more elaborate supra-local shrine cults existed among the Nkoya; of these the contemporary valley shrines may then form a dim reminiscence, drawn into the realm of the locally more recent, chiefly cult (Van Binsbergen, forthcoming a).

                        It is equally difficult to fix the time dimension of the general pattern of the non-ancestral cults of affliction, although it is fairly certain that all the specific cults now occurring among the Nkoya are recent innovations. The great similarity in details between these cults and the village shrine ritual suggests that these cults are partly modern mutants of older village shrine cults; this point was first made by C.M.N. White (1949). on the other hand, these cults of affliction contain elements (mediumship) reminiscent of cults of forest beings and of deceased chiefs such as have been recorded in other parts of Central and Southern Africa (Van Binsbergen, 1972b), and fear which a very recent origin is hardly likely. Elsewhere (forthcoming a) I have argued in detail why, amongst the religious variation in Western Zambia over the last few centuries, the cults of affliction in their present form represent a truly modern innovation: geared to the type of society which emerged in Central Africa as a result of thorough changes in the 19th and early 20th century. I shall not repeat the argument here. I do however wish to draw attention to the fact that, with their extremely weak organisational structure, these cults represent a form of religion which cannot accumulate and exert social power - beyond the face-to-face relationship between leader and local adepts, and then mainly during the ritual. It is almost as if these cults were devised so as to preserve the peace of mind of colonial administrators who ever since the early colonial uprisings in Southern Rhodesia (Ranger, 1966, 1961) and Tanganyika (Gwassa 1972) had been afraid of the anti-colonial political potential of African religion. For this reason the administration had very strongly acted against the prophet Mupumani, African Watchtower, and witchcraft eradication movements (mcape) in the period 1910 - 1940, most of which had only rather weak supralocal Organisations too (Rotberg 1967; Hooker 1965; Shepperson and Price 1958; Ranger 1972; Van Binsbergen forthcoming a). Significantly, when Simbinga founded the bituma cult, the Nkoya area had just seen three crises rapidly succeeding each other within a few years: the ceasing of migrant labour recruitment due to the international Depression; famine due to locust plagues; and massive spread of the Watchtower movement, which because of the anti-government and anti-Lozi statements of the preachers had been forcibly repressed by both the administration and the Lozi paramount chief. Had the religious idiom of Simbinga’s movement pursued any of these lines, his cult would not only have been better documented in tile archives than it is now - it would also have been prohibited to develop and maintain even the rudimentary Organisation Sambaing worked out for it. This does not explain, of course, the positive reasons why the people of Western Zambia adopted bituma and similar cults so eagerly; but it does suggest that the colonial situation imposed, as conditions for survival, certain types of content and Organisation upon religious movements. Cults of affliction could flourish only if they did not present a political threat. They did lack an Organisation that might be mobilised for African political action; and even if they had, like the prophetic cults, the rudiments of such an Organisation, the ideological dimension of the cults wholly concentrated on suffering and treatment - an extreme individual-centredness which sharply contrasts with the other cults from Mupumani to Watchtower, and in which an eschatological blueprint of a radically transformed social order formed the main source of mass inspiration. If the cults of affliction were to function in a rural society where a central government and its recognised African associates (chiefs) held a rigid monopoly of power, they could hardly have developed an effective supra-local Organisation. On the other hand, what extra-religious functions could a more effective Organisation have had? I cannot see any conditions in Nkoya colonial societal which could call for such supra-local yet non-political Organisation. Conflict regulation above the village level was already included in the central power monopoly, and such relationships (mainly with kin) as existed above the valley level were sufficiently reinforced by occasional visits, name-inheriting ritual, and frequent individual removals over wide distances.[20] I realise however that in social science, the explanation of the absence of a feature on the basis of the absence of a need for it,  is hardly less dangerous than explaining the presence of a feature on the basis of its being needed!

4. Comparison of Khumiri and Nkoya data.

The preceding sections provide sufficient data for a detailed and systematic comparison between Khumiri and Nkoya society and cults (App. 1). Data of this kind must be treated with reservation. I have scored both societies on over 40 variables, most of which dichotomous. These variables were suggested to me by my analysis of each society, by preliminary intuitive comparison between both societies, and by my general theoretical views on society and ritual. Other researchers in this field might have selected a rather different set of variables as crucial and then differences and similarities between Nkoya and Khumiri society and ritual might have shown themselves in an entirely different way. Although it is in most respects an advantage that I myself collected the data in both societies and did the scoring, an element of personal bias may have crept in. Among other things, discussion of this paper may consider the merits of my selection of variables and scoring. Meanwhile the comparison yielded the following results. In terms of macro structure and history, Khumiri and Nkoya society have some striking points in common. Each is similar in language, culture and religion to neighbouring groups over a large area. Each has existed for some time in the periphery of distant states, until recent European Colonisation imposed upon them a Centralised polity, which after the nation-state’s attainment of political independence underwent few changes at the local level. Local social structure is similar in that valleys constitute the main effective local communities; in that they have a predominantly horticultural subsistence economy; are not stratified; are dominated by multiplex, inclusive relationships; and have a bilateral underlying kinship structure. They differ however greatly in the density of population, the social organisational density within the community, the stability of community membership, local leadership pattern, the marriage pattern (endogamy, stability and seriality of marriage, effective incorporation of women in their husband’s group), and the economic opportunities of women.

                        As regards communal shrine cults[21] the two societies seem to display a common basic pattern. Shrines are attached to residential groups, for which they are a focus of identity and collective ritual; by virtue of this, communal shrines have a major rallying function for community members and outside contacts; shrines at the maximum community level have keepers; if affliction is attributed to beings associated with communal shrines, this symbolically reflects the in-group social process. But upon that basic pattern, Khumiri and Nkoya cults show marked differences, in the following respects: permanency of shrines; their being linked to other communal shrines through myths and ritual; the extent to which the cult at the maximum community level involves all community members in collective ritual; the cult’s providing a device (through compulsory, supernaturally sanctioned pilgrimages) for the maintaining of outside contacts; the preponderance of women in cults;    and the extent to which supernatural beings associated with the shrine are believed to take a moral interest in the interaction between community members. In addition there are minor aspects in which the Khumiri shrine cult now resembles the Nkoya village shrine cult, then the Nkoya valley shrine cult. These aspects include: the material form of the shrine; is the shrine the burial place of an important person; are shrines named and associated with individual supernatural beings; and has the cult strong ecological connotations?

                        As regards the non-communal cults which in both societies occur along with the communal shrine cult, here again a striking similarity between both societies suggests an underlying basic pattern, including: involvement of a sizeable minority of the male population; membership through initiation; ecstasy; cult readers are associated with shrines; cult sessions have a moderate rallying function, which does not encompass the -total maximum community; a strong political aspect is absent; u d the cults’ extension beyond the ethnic and linguistic confines of the societies under study. To this may be added the historical point that current versions of the cults all appear to be of recent origin. Variations of this basic pattern, making for differences within either society as well as between societies, include such aspects as: the cults’ concentration on individual affliction; the ritual’s requiring more than one officiant, in a role structure of leader and adepts; the extent to which cult membership is exclusive; the extent to which the cult’s ritual concentrates on recruitment as the major treatment of affliction; and the extent to which the non-communal cult is linked to the communal shrine cult in the same society. Only three major differences stand out between Khumiri and Nkoya non-communal cults: the percentage of the female population involved; the cult leaders’ making, or not making, public moral pronouncements concerning the sponsoring community; and the occurrence of major cash transactions between cult leader and sponsors.

5. Conclusion: Towards a theory of shrines and cults in their social context

My comparison of Khumiri and Nkoya cults in their social context suggests a number of theoretical and historical questions. How general and basic are the common patterns I tentatively formulated for both the communal and the non-communal cults? To answer this question, we shall have to refer to additional comparative data from many other societies - much of which will become available during this conference. Comprehensive comparative analysis may be somewhat facilitated by my series of variables (appendix 1), which forms a heuristic device to ask apparently relevant questions about societies where cults of this type occur. But, no doubt, some variables could be omitted and others will have to be added.

                        From a wider comparison, I expect two outcomes.

                        First, in addition to the basic cult patterns now formulated, other irreducible varieties may be brought to light, peculiar to types of societies greatly different from Khumiri and Nkoya society - which despite their being 6,000 km apart have much in common. Secondly, the two basic patterns formulated now may turn out to be far from basic, but instead to include some secondary features which these two societies just happen to have in common, and to lack other essential features present in both societies but overlooked in my analysis. Thus the political aloofness of both Khumiri and Nkoya non-communal shrine cults may well turn out to be a secondary feature some of the cults to be discussed at this conference having great political significance, even though otherwise they are similar to Khumiri and Nkoya cults and operate within a colonial and post-colonial framework similar to that affecting Khumiri and Nkoya [ check ]. The next question is then: how can we make Sense, social-scientifically, of the combination of features in the two basic patterns now formulated? With the uncertain status of these patterns, I think we should not try to propound a speculative theory at this stage. We should wait for the outcome of further comparison. But it is not just a matter of more data To link the various features in the basic patterns, we need nothing less than a general theory linking: social structure; the on-going social, economic and political process; material religious objects (shrines); religious symbolism; ritual; and altered states of consciousness. In short, an integrated theory of shrines and cults in society. Prolegomena for such a theory are available throughout the anthropology of religion, from Durkheim (whose ideas on the relation between group and symbol are clearly relevant here) to modern work, including my own. But I shrink from pursuing this line of argument at the end of an already too long paper.

                        Once we shall have succeeded in identifying one, or very few, universal basic patterns of communal and non-communal cults, the perennial question of cross-cultural comparison will have to be faced: are similarities due to a similar functional set-up in synchronic conditions, or to historical diffusion which at best reflect functional similarities at the time of transmission? In other words, is there a type of society which will, sui generis, produce the basic cult patterns as encountered among the Khumiri and the Nkoya; or is the distribution of this basic pattern due to diffusion? A combination of answers can be envisaged. Cults are not the only aspects of societies to have a history; and by attempting a time series of synchronic functional analysis interesting hypotheses can be generated - and perhaps verified.[22] On the other hand, data now available on spirit mediumship in Africa and its apparent spread throughout the continent in the last few centuries (Beattie and Middleton 1969; Carter 1972) suggest that at least in the ecstatic variety of non-communal cults diffusion is important; not, I think, out of any impetus of its own, but largely because such diffusion is sustained by fundamental processes of change which constitute, outside religion, major related themes in the continent’s modern history: expansion of long-distance contacts and trade; state formation; and Colonisation. Strictly speaking, any ad hoc generalisations suggested by the analysis of one or two societies, should be carefully considered in the light of a larger comparative sample before they can be accepted as valid even for the case study for which they were advanced.[23] But while waiting for a theory and for enlargement of our comparative mini-sample, I am yet inclined to attach some explanatory value to the social-structural similarities concomitance with the similarities in basic cult patterns between Nkoya and Khumiri. Non-communal cults might not have crossed (either way) the boundaries of Khumiri respectively Nkoya society, if these local societies had not been embedded into a general linguistic, cultural and social-structural complex of considerable homogeneity, comprising neighbouring societies over a very large area.[24] Moreover, the difficulty these non-communal cults have in both societies to develop or maintain a supra-local organisation with major political functions, seems definitely related to two social-structural factors: the tendency to localisation (springing, I am inclined to say, from the limited size, the sharp delimitation, and relative isolation of the maximum effective community, the valley), and the confrontation with a Centralised political system. Likewise there is an intuitive link between the communal shrine cults providing, in both societies, a focus of identity, collective ritual and a rallying point both within and between communities - and such social-structural aspects as subsistence economy, absence of stratification, predominance of multiplex, inclusive relationships, and bilateral kinship. Without too much effort, these hints could lead to fertile hypotheses which, once subjected to wider cross-cultural comparison and built into a coherent theory, may enhance our understanding of communal and non-communal cults in any human society. Those aspects in which Nkoya-Khumiri differences in communal, and non-communal cults occur in conjunction with social-structural differences, offer equally fascinating fields for exploration. The rather different parts women play in the cults in both societies appears to be closely related to the marked differences in the marriage pattern, women’s membership of their groups of orientation and procreation, and their access to economic opportunities independent from men. Similarly, if shrines are so intimately linked to residential groups, the extent to which these groups’ membership (both male and female) is stable will directly bear on the form and function of communal shrine cults, on the beliefs concerning communal shrines, and even on the material form of the shrines, their permanency etc. In the same vein, if shrine-owning residential communities are hardly integrated in specific ways with other such communities at a higher level, we hardly expect the associated shrines to be linked to each other through myths and ritual; if there is high social-organisational density, then we expect the reverse to be true. If the local leadership is articulate and operative, communal shrine cults are more likely to support such leadership than when local leadership is diffuse and eclipsed by the Central power of the national state. Whilst in both societies both communal and non - communal cults symbolically reflect on the in-group social process, a penetrating analysis of social control mechanisms, symbolism and sex roles is required before the differential distribution of moral aspects over communal respectively non-communal cults can be understood. Finally, further analysis may also provide insights into the interrelation between the two major types of cults encountered in both Khumiri and Nkoya society. Is their co-occurrence mere coincidence? Are there societies where only the communal shrine cult, or the non-communal cult, exists? And, if both types co-exist in a given society, are moral aspects, and personnel of either sex, allocated to either cult type by a blind game of functional and historical free variation? Or is there (as I suspect) a more intimate and systematic relation between communal and non-communal cults?

This paper does little more than proposing a research programme, and hinting at some of the possible results It leaves the reader with a limited and inconclusive argument, and two too sketchy case studies. However, the fascination, if not the progress, of social science lies not in the occasionally stumbling upon a correct answer, but in the sustained pursuit of meaningful questions. Such value as my argument may have, I hope lies in this direction.

Appendix: a systematic comparison between Khumiri and Nkoya shrine cults (click to open)

References cited,

Anawati, G.C.., and Gardet, L., 1961, Mystique musulmane, Paris.

André, J.G., 1956, Contribution _l’étude des confréries religieuses musulmanes, Alger.

Anonymous, n.d., (Rev. .J.M. Shimunika), Muhumpu wa byambo bya mwaka Nkoya, n.p. (Kaoma). SAGM,

Beattie, J., and Middleton, J.,11969, Spirit mediumship and society in Africa, London .

Bel, A., 1938, La religion musulmane en Berbérie: Esquisse d’histoire et de sociologie religieuses, Tome I: Etablissement et développement de l’Islam en Berbérie du VIIe au XXe siècle, Paris: Geuthner.

Bertholon, L., & Chantre, E., 1913, Recherches anthropologiques dans la Berbérie orientale, Tripolitaine, Tunisie, Algérie, 2 vols., Lyon: Rey.

Brunel, R., 1926, Essai sur la confrérie religieuse des Aissaoua au Maroc, Paris: Geuthner.

Carter, M., 1972, ‘Origin and Diffusion of Central African Cults of Affliction’, paper read at Conference on the History of Central African Religious Systems, University of Zambia/University of California Los Angeles, Lusaka.

Clay, G.C .R., History of the Mankoya district, Lusaka 1945.

Colson, E., The plateau Tonga, Manchester 1962.

-, Spirit possession among the Tonga of Zambia, in: Beattie and Middleton 1969 .

Creyghton, M.-L., 1969, Folk illness in een Tunesisch dorp, thesis, University of Amsterdam.

Cunnison, I., 1956, ‘Headmanship and the ritual of Luapula villages’, Africa, 26 (1956), 2–16.

Demeerseman, A., 1964, ‘Le Culte des saints en Kroumirie’, Institut des Belles Lettres Arabes, 27: 119-63.

Depont, O., & X. Coppolani, 1898, Carte de l’Algérie, l’Afrique, l’Asie et la Turquie d’Europe: domaine géographique des confréries musulmanes, Alger:..

Durkheim, E., 1912, Les formes élémentaires de la vie religieuse, Paris

Fielder, R.J., 1965, ‘Social change among the Ila-speaking peoples of Northern Rhodesia, with particular reference to their relations with the Primitive Methodist Mission’, MA thesis, University of Manchester.

Fortes, M., 1953, The structure of unilineal descent groups, in:  American Anthropologist 55:17-41.

Foster, G. M. 1961. The dyadic contract in Tzintzuntzan: a model for the social structure of a Mexican peasant village. Am. Anthrop. 63, 1173-92. 

Foster, G. M. (1963). The dyadic contract in Tzintzuntzan, II: Patron-client relationships. American. Anthropologist, 65, 1280–1394.

Geertz, C., 1966, Religion as a cultural system, in: Banton, MM., ed ., Anthropological approaches to the study of religion, London.

Gellner, E., 1969, Saints of the Atlas, London.

Gielgud, Reports on the Gielgud-.Anderson expedition to the Kafue Hook, Zambia National Archives, BS1/93 and KTJ1/1.

Gwassa, G.C.K., 1972, ‘Kinjikitile and the ideology of Maji Maji’, in: Ranger, T.O. & I. Kimambo, eds, 1972, The Historical Study of African Religion, London: Heinemann, pp. 202-17.

Hartong, A.M., 1968, ‘De Geschiedenis van het Sjeikaat Atatfa op Basis van de Orale Traditie, Drs. Soc. Sc. thesis, Catholic University of Nijmegen

Hooker, J.R., 1965, ‘Witnesses and Watchtower in the Rhodesias and Nyasaland’, Journal of African History, 6: 91-106.

Ikacana, N.S., 1952, Litaba za Makzwanga, Lusaka.

Jaspan, M.A., 1953, The Ila-Tonga Peoples of North-Western Rhodesia, Oxford University Press for International Africa Institute, Ethnographic Atlas of Africa.

Jongmans, D.G., 1968, ‘Meziaa en horma. Samenhangen tussen dienstbetoon, eer en welstand in een veranderde samenleving’, Kroniek van Afrika, 3:1-34.

-, 1970, Jongmans, D.G., 1973, ‘Politics on the Village level’, in: Mitchell, J.C., & Boissevain, J.F., eds., Network Analysis: Studies in Human Interaction, The Hague/Paris: Mouton, pp. 167-217.

Köbben, A.J.F.., 1966, Why exceptions?, in: Current Anthropology.

-, Köbben, A.J.F., 1970, ‘Comparativists and non-comparativists in anthropology’, in: Naroll, R. & R. Cohen, eds, 1970, A Handbook of Method in Cultural Anthropology, Garden City (N.Y.): Natural History Press, pp. 581-96.

Ling, D.L., 1967, Tunisia: from Protectorate to Republic, Bloomington & London: [ add publisher ]

McCulloch, M., 1951, The southern Lunda and related peoples, London: International African Institute, Ethnographic Atlas of Africa.

Mainga, M., 1969, Bulozi under the Luyana kings, London.

-, 1972  A history of Lozi religion to the end of tlhe nineteenth century, in: Ranger and Kimambo 1972.

Marçais, W. ., and Guiga, Q., 1925, Textes arabes de Takrouna, Paris.

Molé, M., 1963, ‘La Danse extatique en Islam’, in: Les Dances Sacrées, Anthologie, sources Orientales 6, Paris: [ publisher ] 1963, p. 145-280.

Montet, E., 1909, Le culte des saints musulmans dans l’Afrique du Nord, Geneve.

Muntemba, M.S., 1972, ‘Zambia Nzila Sect and Christian Churches in the Livingstone Area’, paper read at Conference on the History of Central African Religious Systems, University of Zambia/ University of California Los Angeles, Lusaka.

Ranger, T.0., 1966, ‘The Role of Ndebele and Shona religious authorities in rebellions of 1896 and 1897’, in: Stokes, E. & R. Brown, eds, 1966, The Zambesian Past, Manchester University Press for Institute for Social Research, pp. 94-136.

-, 1968  Revolt in S.Rhodesia 1896-7, London.

--., 1972, Michape and the study of witchcraft eradication, paper read at tlle Conference on the History of  Central African religious systems, Lusaka.

Ranger, T.0., and Kimambo, I., eds., The historical study of Afrikan religion, London, 1972.

Rotberg, .h.I., The rise of  nationalism in Central  Africa, Cambridge, 1967.

Shepperson, G., and Price, T., 1958, Independent African, Edinburgh.

Smith, E.W., & A.M. Dale, 1920, The Ila-speaking peoples of Northern Rhodesia, 2 vols, London: Macmillan

Souyris-Rolland, ML., 1949, Histoire traditione1le de la Kroumirie, in: IBLA 19:127-65.

Speight, R.M., 1966, ‘Tunisia: Sufism’, Moslem World, 56: 58

Trimingham, J.S., 1965, Islam in the Sudan, London.

Turner, V.W., 1957, Schism and continuity in an African society, Manchester .

-, 1967, The forest of symbols, Ithaca.

-, 1968, The drums of affliction, Oxford.

-, 1969, The ritual process, London.

Van Binsbergen,W.M.J., 1970a, Verwantschap en territorialiteit in de sociale structuur van het bergland van NW.Tunesie, thesis, University of Amsterdam.

-, 1970, Genealogical knowledge, manipulation and the Murphy-Kasdan-Randolph hypothesis, MS.

-, 1971a, Religie en samenleving: een studie over het bergland van NW Tunesie, thesis, University of Amsterdam.

-, 1971b, Ernest Gellner: Saints of the Atlas, in: Cahiers des Arts et Traditions Populaires, 4, 1971: 203-11.

-, 1971c, Popular and formal Islam, and supra-local relations: the highlands of N.W.Tunisia 1800-1970, seminar papcr, University of Zambia.

-, 1972 a, Bituma: preliminary notes on a pcssession Ihealing cult amons tlhe Nkoya, paper read at the Conference on the History of Central-African Religious systems, Lusaka; revised version in: Van Binsbergen, forthcomiug - a.

-, 1972b, Possesion and mediumship in zambia: towards a comt.a.rative approach, paper read at tlsc Conference on the History of Central-African religious systems, Lusaka; revised version in: Van Binsbergen, forthcoming- a.

-, 1974a, Ethnicity as a dependent variable: the ""Nkoya"" ethnic identity and inter-ethnic relations in Zambia, paper read at the 34th Annual Meeting, Society for Applied Anthropology; forthcoming in: African Social Researchh.

-, 1974b, Labour migration and tlhe generation conf lict : an essay on social change in Central Western Zambia, paper read at the 34th Annual Meeting, Society for Applied Anthropology; forthcoming: in: Cultures et développement, 1976

-, 1974a, Ritual, claa ss andurban-rural relations: elements for a Zambian case study, in: Cultures et développement, 1976, 2 (in press ) .

-, 1975b, Law in the context of Nkoya society, in: Roberts, S.A., ed., New directions in African family law, The Hague/Paris (in press).

-, forthcoming - a, Explorations into the religion of  Zambia.

-, forthcoming - b, Ritual, class and urban-rural relertions: the Nkoya of Zambia.

Werbner, R.P., and Garbett, G.K., Conference on regional cults and oracles, Association of Social Anthropologists, circular letter, 197 5 .

White, C.M.N., 1t 9, Stratification and modern changes in an ancestral cult, .Africa 19:324-31. Zaretsky, I., 1967, Bibliography on spirit possession and spirit mediumship, Evanston.

Appendix: a systematic comparison between Khumiri and Nkoya shrine cults (click to open)

[1]          This paper is based on fieldwork carried out in Tunisia, 1968 and 1970, and in Zambia, 1972-1974. In addition to my informants and the Tunisian, and Zambian, authorities, I am indebted to the following persons and institutions: to Hasnaui ben Tahar and Denes Shiyowe for excellent research assistance; to Henny van Rijn, my wife, for sharing much of the fieldwork and the analysis; to the University of Amsterdam, the Centre des Arts et Traditions Populaires (Tunis ), the University of Zambia, the latter’s Institute for African Studies, the Netherlands Foundation for the Advancement of Tropical Research (WOTRO), and the State University of Leyden, for financial support and research facilities ; to Douwe Jongmans, Marielou Creyghton, Jeremy Boissevain, Jaap Van Velsen, Klaas Van Der Veen, André Köbben, Matthew Schoffeleers, Terence Ranger, Maud Muntemba, Bob Papstein and Dick Werbner, who over the years have been my partners in stimulating discussions on parts of the present argument; to John Beattie for cowmen

[2]          Such standard reference books as the Encyclopaedia off the Social Sciences (Macmillan), or Kolb and Gould ‘s Dictionary of the Social Sciences, do not contain items on any of the central topics in this conference .

[3]          The concept of dyadic relationships, which has great applicability in Khumiri society (cf. Jongmans 1968, 1973) is taken here in the sense it has been introduced by Foster (1961, 1963).

[4]              In two adjacent villages where I obtained a close to complete insight in the complex genealogical networks between the inhabitants, 65 marriages were in existence in 1968. of these, 10 (15%) were between effective agnates (people who can trace their genealogical relationship in distinct, patrilateral links ; on the complex problems involved, cf. Van Binsbergen 1970a, 1970b); out of these 10 marriages, only 5 (8% of 65) were with FFBD, the other 5 involved more distant agnatic links (Van Binsbergen 1970 : 138 ) .

[5]          4A shrine is a particular spot in the landscape, where a concentration of ritual takes place, and which is marked as such by external, observable features defined within the local culture (Van Binsbergen 1971a : 72 ) .

[6]          This ritual is called uriya. More frequently it is directed to the Prophet Mohammed, who otherwise is not a subject of ritual action among the Khumiri, although he features in some pious songs.

[7]          Khumiri inheritance law differs on this point from Qoranic law.

[8]          In the Khumiri highlands, divination of causes of illness is performed by various types of specialists : prominent members of the ecstatic cult, particularly those attached to the cult ‘ s local lodge (cf. 2.3); Qoranic teachers (maddab), of whom a few live scattered over the Khumiri area; migrating diviners who once a year visit the area and who are supposed to descent from non-Khumiri saints; and finally diviners (takaza) whose technical skills have no special religious connotations, and who are mainly found near the weekly markets .

[9]          A small town 100 km south of the Khumiri highlands.

[10]         Statistical analysis suggests occurrence of faqirship amongst one ‘s close matrilateral kinsmen as the single main factor in the recruitment of faqirs - a fact which, given the patrilineal ideology of this society, reminds us of Fortes’ classic generalisation claiming "interests, rights and loyalties ( . . . ) that rely on religion ( . . . ) to be generally tied (...) to the complementary line of filiation"" 01953:34).

[11]         Not all of them were: e.g. Sidi Abdallah and Sidi Salah are Khumiri representatives of universal maghrebine saints ; cf. Marçais and Guiga 1929a5.

[12]         The ecstatic element in North-African religious brotherhoods is said to derive from three sources : early Islam in the Middle East (cf. Molé 1963) ; ecstatic cults in sub-Saharan Africa (cf. Brunel 1926; Trimingham 1965); and autochthonous ecstatic cults dating back to Antiquity (cf. Bertholon and Chantre 1913). In addition to anthropological studies, much more historical research is needed on this point.

[13]         For the early colonial era, cf. Ling 1967:59; for the post-colonial era, cf. Speight 1966.

[14]         Apart from the competition for honour and informal leadership of lower segments - which is rather futile in comparison with the enormous power invested in the local representatives of the central government and the national political party.

[15]         For a similar view of Luapula villages in N.E. Zambia, cf. Cunnison 1956.

[16]         In the area, the only other occasions bringing together this number of people, are girls’ puberty ceremonies, and burial.

[17]         By ecological I mean all activities (and the accompanying cognitive processes) by which Man acts upon, structures and transforms his natural, non-human environment, so as to create for himself lie material basis of human society (food, shelter, materials, skills, technology etc.)(Van Binsbergen 1971:239f;forthc.-a). 16a. Many cult songs are bilingual (e.g. Nkoya/Luvale, Nkoya/Mbunda); adepts and leaders of most cults are found among several of the many ethnic groups of Western Zambia, although most cults are considered to originate from one such ethnic group.

[18]         A much more developed example from outside the Nkoya area is the Nzila sect, the bureaucratised healing cult (claiming nearly 100,000 adepts), which developed out of the vision of Chief Chana in 1940 (Muntemba 1972), [ CHECK, DIT IS NOOT 17 NIET 16A, DIE IS NIET TE VINDEN ]

[19]         In this paper, the emphasis is on rural communities, I have decided not to discuss rural-urban ties. Both village shrine cults and non-ancestral cults of affliction play however an important role in Shaping rural-urban relations (Van Binsbergen 1972b;1976a; forthcoming - b); at least among the contemporary Nkoya. affliction diagnosed as ancestral makes urban migrants refer to their home villages ‘or treatment, thus binding them effectively to the economic and political process of the village in which they are forced td) keep a stake, e.g. because of the paucity of` urban retirement opportunities for low-class migrants. Non-ancestral affliction binds urban migrants to cult leaders both in town and in their home area (many rural cult leaders occasionally visit town to stage sessions; a few leaders live in town permanently); in addition, the urban cult sessions provide a major rallying point for urban migrants from the Nkoya area and from Western Zambia in general, and is therefore a major force in shaping these urban migrants into an urban ethnic group with frequent interaction and common identity among the members who live dispersed over a large city, These aspects hardly received attention in my Khumiri studies. [ CHECK IF PLACEMENT OF NOTE IS CORRECT HERE ]

[20]            [ check for repetition ] In this paper, the emphasis is on rural communities, I have decided not to discuss rural-urban ties. Both village shrine cults and non-ancestral cults of affliction play however an important role in Shaping rural-urban relations (Van Binsbergen 1972b;1976a; forthcoming - b); at least among the contemporary Nkoya. affliction diagnosed as ancestral makes urban migrants refer to their home villages ‘or treatment, thus binding them effectively to the economic and political process of the village in which they are forced td) keep a stake, e.g. because of the paucity of` urban retirement opportunities for low-class migrants. Non-ancestral affliction binds urban migrants to cult leaders both in town and in their home area (many rural cult leaders occasionally visit town to stage sessions; a few leaders live in town permanently); in addition, the urban cult sessions provide a major rallying point for urban migrants from the Nkoya area and from Western Zambia in general, and is therefore a major force in shaping these urban migrants into an urban ethnic group with frequent interaction and common identity among the members who live dispersed over a large city, These aspects hardly received attention in my Khumiri studies.

[21]            I have some doubt about the concept of regional cult, as it was defined by Werbner and Garbett (1975) for the purpose of this conference. Their definition runs along the following lines: a regional cult is a cult which (a) reaches beyond a political or ethnic community; (b) whose membership tends to be recruited across major socio-economic divisions; (c) whose staff’s capacity to pronounce on matters of moral and ritual concern is effective both within and between communities.

"A regional cult is thus intermediate in span and falls between the extremes of an exclusive, parochial cult whose congregation is drawn from a single ethnic, political or narrowly localised community, and a universal church which, in principle, recruits members irrespective of their meet specific communal affiliations".

Communal Shrine cults among the Khumiri (saintly shrines) and the Nkoya (village and valley shrines) have a maximum catchment area with a radius of about 10 km: the valley and adjacent valleys. This amounts to a maximum catchment area of over just over 300 km2 (= p * (10 km)²) . Is this enough to qualify as "intermediate in span"? I should think so. Yet in neither case is there (cf. criterion (a)) any reaching beyond the political and ethnic community, which in both cases is much wider (valleys being ecological and interactional, but not political or ethnic communities). In both non-stratified societies the major socio-economic divisions are along sex and generation; participation in the communal shrine cult includes all community members, so I suppose criterion (b) is met;. I Cannot see how criterion (c) is met in either case. The suggestion is that criteria (a), (b) and (c) represent different typological dimensions which only rarefy coincide. A similar argument can be presented for the ecstatic cults in both societies, although there Werbner and Garbett’s definition fits somewhat better. I therefore opt for the simpler concepts of communal cult (involving shrines associated with a residential group characterised by internal interaction and common identity), and non-communal cults (all others).

[22]            I have done something like this for communal shrine cults and political change in Central Africa (Van Binsbergen, forthcoming = 1981 ch. 3)

[23]         This is the extreme comparativist position, based on the perhaps controversial methodological view that, as explanation implies generalisation, the power of an explanation depends on the extent to which it can be generalised.

[24]         In the Khumiri highlands, the local inhabitants ‘ nominal adherence to the universal religion of Islam, in conjunction with the local segmentary autonomy, provided (together with clientship /herdsmanship) the main entrance for individual strangers to settle : they were welcome as specialists in the universal religion, could develop local esteem, security and ability t+) act as conflict-resolving outsiders on that basis, to the extent of being turned into local saints after their deaths .

return to the index page of Historic Berber culture

			The illustrations in the heading of this webpage: The background photograph shows a women's work group (mainly drawn from the family of chief Hillal bin Hassuna), from the village of Sidi Mhammad, harvesting rye near the shrine of Bu Qasbaya al-Kabir, late spring 1968; the same illustration repeated in the original black and white, bottom left of the page head; top centre left: the shrine (mzara) of the neighbourhood of Qa'a Raml, village of Sidi Mhammad, late spring 2003 -- the fresh pious gifts show that the cult is still alive and kicking -- note the car bumper which has come to replace cork plates as the shrine's traditional roofing; top centre right: the domed shrines (qubba) of Sidi Mhammad al-Wilda (the Son) in the village of Sidi Mhammad, late spring 2002 -- note how the excessive erosion of 1968 is no longer in evidence, due to the local villagers' decades of reafforestation efforts, in the context of the Tunisian state's unemployment relief work -- the photo is slightly misleading in that it just keeps out of view the extensive Qur'anic school complex which external, formal Islamic interests have erected just twenty meters north of the shrine of Sidi Mhammad, in the mid-1970s; right: a Khumiri warrior, photographed (no doubt under carefully arranged near-studio conditions( by Garrigues c. 1880, and reproduced in Bertholon, L., & Chantre, E., 1913, Recherches anthropologiques dans la Berbérie orientale, Tripolitaine, Tunisie, Algérie, 2 vols., Lyon: Rey -- a colonial anthropological study on Tunisia.