THE CULT OF SAINTS IN NORTHWESTERN TUNISIA
An analysis of contemporary pilgrimage structures
Part I (chs. 1-5)
Wim van Binsbergen
2. REGIONAL AND HISTORICAL BACKGROUND
3. SEGMENTATION IN KHUMIRIYA TODAY
4. SHRINES IN KHUMIRIYA
5. SAINTS AND THE LIVING
click here in order to access Part II
6. SEGMENTATION AND TYPES OF ZYARA
7. LOCAL ZYARA IN THE VALLEY OF SIDI MHAMMAD
8. ORIGINAL AND PERSONAL ZYARA IN THE VILLAGE OF SIDI MHAMMAD
In this chapter I shall present a description and analysis of the cult of local saints, as major aspect of contemporary popular religion in the highlands of Khumiriya, north-western Tunisia. This paper is therefore a contribution to the ethnography of religious behaviour in general and that of rural North Africa in particular. As is the case in much of religious anthropology, studies of popular Islam have tended to concentrate on systems of belief and symbolism, with excursions into the relation between religion and the wider social, economic and political context in which that religion occurs. The behavioural aspect of religion has been somewhat neglected, and as a result for some of the most pertinent questions of contextual religious analysis we have had to content ourselves with tentative answers largely founded on intuition and persuasion; the necessary empirical data have often been lacking. A major problem in this connexion is that an empirical, quantitative description of religious behaviour such as I shall offer towards the end of this chapter remains meaningless without an adequate discussion of the symbolic and social-organizational aspects of such behaviour.
Having elsewhere dealt with the historical aspects of saintly cults and the interplay between popular and formal Islam in the Khumiri region (cf. Van Binsbergen 1971a, 1980, 1980b), I shall here largely limit myself to the contemporary situation concerning pious visits (zyãra) to shrines associated with named local saints touching on local history only in so far this helps to explain the nature of territorial segmentation today, and refraining from a discussion of such significant aspects of Khumiri religion as: the veneration of trees and sources; veneration of saints through other rituals than pious visits; the ecstatic cults that are loosely organized in religious brotherhoods and that, although implying saints, form a popular-religious complex somewhat distinct from zyara; the symbolic deep structure of such key concepts as sainthood and baraka; and finally the formal Islam of the Quran, the mosque, the pilgrimage to Mecca. Even so the ethnographic argument will be too lengthy to wallow for a more than cursory discussion of the many wider theoretical implications of the Khumiri data (cf. Van Binsbergen 1971a, 1976, in preparation).
2. REGIONAL AND HISTORICAL BACKGROUND
Khumiriya is a mountainous area in north-western Tunisia, situated between the Tunisian-Algerian border (which is hardly a social and cultural boundary), and the towns of Tabarka and Janduba, The regional capital is the small town of cAin Draham, where the regions only and recently-built mosque is found.
Until the late 19th century, the narrow, densely-forested valleys of this remote region provided a relatively prosperous livelihood for a tent-dwelling population engaging in semi-transhumant animal husbandry (cattle, sheep, goats) and small-scale agriculture (wheat, rye, olives). Each of the scattered homesteads consisted of a core of close agnates, with their wives, children and non-agnatically related adult male dependants (herdsmen, who often became sons-in-law). These residential and productive units existed at the basis of a segmentary system, whose explicit ideology was one of patrilineal descent but in which, in fact, factional allegiance, geographical propinquity, and genealogical manipulation were equally important structuring principles. Localized clans, tribes, and confederations of tribes formed the highest levels of the segmentary model. The segmentary organization regulated: rights over pastures, forest areas and springs; special patronage links between social groups and invisible saints, associated with the numerous shrines scattered over the land; and burial rights in local cemeteries situated around a saintly shrine although, given the large number of shrines and the very small number of cemeteries per valley, most shrines had no cemetery around them.
On all segmentary levels, complementary segments were in competition with each other over scarce resources, women, and honour. The armed conflicts to which this competition frequently gave rise, were in two ways mitigated by the cult of saints.
First, each higher-level segment (encompassing the majority of the population of a valley) would have a twice-annual saintly festival (zarda) near the shrine of its patron saint, located at some conspicuous point in that valley. On this occasion, all members of the local segment (i.e. all inhabitants of the valley) would make a collective visit to the shrine, and would for several days stay near the shrine, chatting, feasting, and being entertained by dancing and singing. Members of feuding segments in neighbouring valleys were likewise under obligation to make a pious visit to the shrine concerned, attending this festival, and sharing in the collective meal there. Temporary lifting of segmentary opposition was achieved not only through this ritual commensality but also through a safe-conduct for all pilgrims, sanctioned by the invisible saint. Also women who, originating from the local segment, had married into a different valley, were under obligation to make the pious visit to the shrine on the occasion of the saintly festival.
Secondly, the major shrines those that had a twice-annual festival catering for an entire valley were administered by specialist shrine-keepers. The latter were not considered saints in themselves, but they were pious, pacifist men who had placed themselves outside the feuding system and who, on the basis of a saintly safe-conduct and by virtue of the respect that the shrines flags commanded, were often successful in quenching violence between segments.
The colonial period in Tunisia, which began with the French conquest of Khumiriya in 1881, brought tremendous changes in the social, economic and religious structures of the region. It took the colonial state a quarter of a century to impose its monopoly of violence, but from the beginning of the twentieth century an effective stop was put to feuding as the main motor behind segmentary dynamics. Movement of the population was further restricted by state exploitation of the extensive cork-tree forests, the establishment of settler farms (which in Khumiriya however remained a much more limited phenomenon than in the fertile Tunisian valleys to the south and the east of this mountainous area), and the concentration of land rights in the hands of a few state-appointed chiefs and their families, who were in collusion with the colonial administration. Pressure on the land was exacerbated by dramatic population increase, and massive erosion through over-exploitation of the vulnerable soil system proved inevitable. The economic opportunities in the French-created garrison town of cAin Draham, even after its development into a regional capital and a tourist resort, could not compensate for the decline of the local subsistence economy; neither could, during the colonial period, labour migration directed to areas of capitalist farming, and to urban areas, in Tunisia and Algeria. The re-afforesting projects and the unemployment relief work undertaken since Tunisia became independent (1956), did not alter this state of affairs substantially. The ethnographic present of the late 1960s offers the picture of a destitute peasant population, which within the rigid confines of its villages of immobile stone houses and small and fragmented fields keeps going a transformed neo-traditional social and ritual organization, and a no-longer viable local subsistence economy ineffectively supplemented by unemployment relief projects.
3. SEGMENTATION IN KHUMIRIYA TODAY
The model of a segmentary lineage system has remained the standard idiom by which participants structure their social environment,, distinguish between residential groups, and explain relationships between these groups. In the face of the realities of peripheral capitalism, this lineage model became devoid of such economic and political significance as it had in nineteenth-century Khumiriya. It no longer effectively governs the everyday ongoing social process in the villages. Moreover, as the population has become totally sedentary, and pressure on the land increased, the idiom of patrilineal descent is no longer a device for segmentary mobilization in the competition over scarce resources, but has become merely a folk idiom to describe the pattern of organizational alignment of bounded territorial units such as are manifestly visible in the Khumiri countryside today and a means to claim legitimate membership of such unites, i.e. rights of residence and rights in land.
Diagram 1. Segmentation in Khumiri society (click on thumbnail to enlarge)
From the lowest level upwards, we find (cf. diagram 1) households, compounds, sub-neighbourhoods or hamlets, neighbourhoods, villages, valleys, chiefdoms. Each of these is clearly marked, and distinguished from complementary units at the same segmentary level, by unmistakable features in the landscape: the walls of dwelling-houses and the open spaces between houses; the cactus fences between compounds and hamlets; the pastures, fields, shrub-covered fallow areas, and patches of forest between neighbourhoods and between villages; and the steep, forested mountain ranges between valleys and between chiefdoms consisting of a number of valleys.
Most of these units are designated by names of derived from human proper names: Dar cAli (cAlis House), Mhamdiya (Descendants of Mohammed), Ulad Ibrahim (Descendants of Ibrahim), etc. While these labels in fact function as names for residential units, and as toponyms, their evocation of a historical or mythical ancestor from which all born members of that unit are claimed to descend, enables Khumiri participants to represent their territorial organization today by a patrilineal genealogy encompassing an entire valley and even chiefdom despite massive oral-historical evidence at my disposal which clearly establishes that, at least in the 12 km2 that formed the core of my research area, few compounds and hamlets, and no neighbourhoods, villages or higher-level territorial segments, are composed of a homogeneous set of agnates descending from one common ancestor. On the contrary, the population belongs to more than a dozen mutually unrelated patrilineal descent lines, most of which immigrated into their present day territory in the course of the nineteenth and twentieth century; only by virtue of genealogical manipulation can they manage to identify as agnates.
Khumiri territorial segments have distinctive features beyond their visible boundaries and their proper names evoking ancestors. The extent to which the model of territorial segmentation sketched here is not just a researchers construct, but a living reality to the participants, is clear from the fact that at each level of territorial segmentation a segment has a characteristic attribute which defines it against complementary segments at the same level. Like the unit boundaries, these attributes are clearly visible in the landscape, and they are a result of human activity. Each household is characterized by its own dwelling-house, which defines the basic unit of human reproduction, since by containing the family bed it sets the scene for sex life, child-birth and child-rearing. A few dwelling-houses combine so as to form one compound; this territorial unit is defined by the storage table, which marks the compound as a basic unit of food processing and consumption. Each hamlet or sub-neighbourhood consisting of a small number of compounds, is characterized by its own threshing-floor, which defines the hamlet as a minimal unit of agricultural production. Neighbourhoods, consisting of a small number of hamlets, each have their own springs, use of which is private to the members of that neighbourhood. The spring defines the neighbourhood as a unit whose members share (for such purposes as water hauling, grazing, collection of firewood hunting) an overall productive interest in the surrounding countryside, even though the neighbourhood is internally divided into smaller complementary segments with relation to those aspects of production and reproduction that require more prolonged, complicated and socially more intricately-organized tasks. Finally, villages, consisting of a small number of neighbourhoods, are characterized by their own mens assembly: a wind-swept open space overlooking the valley and its main shrines. Here the adult male inhabitants of the village assemble towards the evening, to discuss the ongoing social and political process and to entertain each other with tea-drinking and card-playing. If the village has a store, it is located adjacent to the mens assembly. The mens assembly defines the village as the social unit of sufficient scope and at the same time of sufficient intimacy, to accommodate the ongoing face-to-face social process between people who have widely divergent and conflicting economic interests, as members of lower-level segmentary units. At the mens assembly people meet most of whom, while not strangers to each other, do not automatically share a day-to-day routine of dwelling and working together; thus the mens assembly provides a social and political arena, a more or less external yet inescapable standard for the evaluation of wealth, honour, and propriety, and as such the wider social framework of the interactional processes on which, within the lower-level segmentary units, the organization of production and reproduction depends.
Khumiri territorial segments thus are not just significant units in the organization of geographical space, they also structure the social and economic space in a way that reflects the vital processes going on in this society. The characteristic attributes by which each segmentary level is marked are, as it were, chosen with great wisdom, and their very nature is suggestive of the social and economic significance of the segments at various hierarchical levels. Not surprisingly, in Khumiri symbolism the storage table, the dwelling-house, the threshing-floor, the spring and the mens assembly constitute powerful images, around which an important part of the local world-view condensates and finds expression. What is more, each of the characteristic attributes mentioned is conceived as a diffuse, nameless but somewhat personalized, supernatural entity, a distinct power which appears in the dreams of the human members of the segment with which it is associated, and which can mete out benefits and punishment depending on the degree of propriety and respect people display in the specific activities involving that characteristic attribute. Nor are these activities of an exclusively utilitarian nature: dwelling-house, threshing-floor, spring and mens assembly are in themselves subjected to ritual actions, particularly the burning of incense and the sprinkling of chicken blood. The most important symbolic aspect of these characteristic attributes, and one that in the peoples eyes sufficiently explains the animistic overtones alluded to here, is that (as latent or primordial shrines) they are all carriers of baraka, the Grave or Life-force through which, under the catalytic effects of morality and good social relations, Man succeeds in sharing the non-human power of Nature and of the Divine.
These characteristic attributes with their rich symbolic elaborations are the visible beacons in a structure of territorial segmentation. But although segmentary dynamics have been stagnant as compared with the turbulent pattern obtaining in the last century, the system of territorial segmentation is by no means entirely static today. Despite rural decline and the pressure on the land, demographic and economic processes are at work which over time propel some lower-level units to higher levels, and vice versa. A compound, while retaining its proper name and ancestral association, may be seen to wax into a neighbourhood and even a village in the course of half a century or less. In those cases the named units, as they break through from one segmentary level to a lower or higher one, will shed the characteristic attribute appropriate to the former level and will adopt one appropriate to the new level. Thus the construction, and the sinking in decay, of dwelling-houses, threshing-floors and mens assemblies, and shifts in patterns of water hauling from one spring to another, all mark, again in a way that is visible in the landscape, the waxing and waning of territorial segments.
This is the moment to introduce shrines into our increasingly complex picture of territorial segmentation in contemporary Khumiriya.
4. SHRINES IN KHUMIRIYA
Shrines exist in Khumiriya in a number of variants. I shall leave aside such non-man-made salient features in the landscape as remarkable trees, rock formations and ferruginous springs, which tend to be venerated without being clearly associated with saints. All other shrines are man-made, and considered to be intimately associated with saints: deceased human beings whose baraka was and is such that they continue to wield power in the world of man. The association between shrine and saint is conceived in either of the following three ways:
a. the shrine was erected upon the saints grave;
b. the shrine was erected upon a spot that had a special relation with the saint during his lifetime or shortly after his death: as the place where he rested in the course of his wanderings, or where his body was temporarily put before definitively being put into the grave; and finally
c. the shrine has been secondarily erected upon relics brought from a shrine as explained under a or b. For each shrine there tends to be some disagreement among participants as to which option (a, b or c) applies in its particular case. The historical dynamics underlying these patterns fall outside our present scope.
Saintly shrines comes in variety of material forms. All mimick more or less the human dwelling-house. Many do so in a very crude form, and consist only of a semi-circle of large rocks covered by another rock or by a slab of cork. This is the type commonly called mzara, although this term (meaning that which is visited) in principle applies to all shrines. In some shrines the inner room within the ground-plan of rocks is more spacious and of more or less rectangular shape; they may be covered by an elaborate reed roof supported by forked poles carrying a roof-beam. This is the type called kurbi, a word otherwise reserved for human dwelling-houses constructed out of arboreal material. The most elaborate type of shrine in Khumiriya is the qubba: a square, stone building with plastered white-washed walls, a domed roof and horned ornaments on the four corners, as commonly found throughout the Islamic world.
All saintly shrines certain minor pious gift: small amounts of incense wrapped in paper, candles, incense-burners and candle-sticks locally made out of fired clay, and household refuse such as broken teapots and spoons purposely taken to the shrine as token offerings. In addition, the shrines associated with saints, in the local hierarchy of saints?, are considered to rank high, often contain stone balls (kurra: the saint is said to have carried them in his life-time, as proof of his sainthood); elaborately decorated flags donated to the shrine as votive gifts; and a wooden chest in which these flags are stored along with other pious gifts, including coins.
Although for the sake of simplicity saints are described here as male, participants acknowledge the existence of female saints. A valleys major saints usually are male. Many saints bear ordinary personal names: Massauda (Aisha, etc.) preceded by the reverential term of address Sidi (master, sir: elder brother), Lalla (madam, miss, grandmother; elder sister) or Jaddi (grandparent). A large number of saints however do not bear human names but derivatives of words denoting natural species: Bu-Kharuba (Man with the Carob-tree), Bu-Qasbaya (Man with the Reed), etc. The Khumiri saintly cult has an understream? of totemism which is also manifest in saintly legends and taboos; but this, however interesting, falls outside our present scope.
Neither can I go into detail here with regard to the relationships deemed to exist between saints. Various structuring principles are invoked to establish some degree of order among the large number of local saints with which each Khumiri participant is familiar. First, there is a general hierarchy of saints, ranking from Sidi cAbd al-Qadir al-Jilani (who throughout the Maghrib is considered to be the most powerful saint), through a small number of major saints of more than regional significance (e.g. Sidi cAbd as-Salam ben Mashish), to the greatest Khumiri saints (the ones whose shrines are best known and whose festivals are best frequented: Sidi cAbd Allah bi-Jamal, Sidi Mhammad, Sidi bu-Naqa, Sidi Bu-Kharuba, Sidi Ben-Mtir), the lesser saints that are only known within a valley and adjacent valleys, and finally the least powerful saints, the ones that are only known and venerated at the village, neighbourhood or even compound level.
This hierarchy very roughly corresponds with the material form of the principal shrines associated with those saints. Whereas the top-ranking international saints do not even have shrines within the region (they are known through hagiographic legends, and as saints featuring in the songs that pertain to the ecstatic ritual of the brotherhoods), the greatest regional saints have long-established qubbas, those immediately below them tend to have large kurbi shrines or large rock mzaras, whereas the smallest mzaras and miniature kurbi shrines tend to be associated with the least important saints.
Besides this overall hierarchy, saints associated with shrines within the same valley, or in adjacent valleys, tend to be linked to each other in hagiographic legends that claim specific relationships to exist between these saints: they are described as unrelated equals (neighbours, friends), as non-kin involved in a master-servant relation, or most frequently as close agnatic kinsmen: father and son, brothers, brother and sister.
The erection of a shrine upon relics brought from an older shrine often creates a situation where, within a valley or adjacent valleys, a number of shrines are associated with and named after one and the same saint. In that case the main shrine (the one that is the most elaborate, and that has the greatest festival) is considered to be the original shrine although objective historical research would not always bear out the participants view on this. This shrine is called the Elder, al-Kabir, whereas the other shrines bearing the same name are called the Son (al-Wilda). Thus in the valley of Sidi Mhammad four shrines of the saint Sidi Mhammad exist: Sidi Mhammad al-Kabir is a qubba located on a hill-top overlooking the valley, whereas one qubba and two kurbi shrines, all three called Sidi Mhammad al-Wilda, are found at a distance of 1 to 1.5 km south of Sidi Mhammad al-Kabir. In the same valley, four shrines associated with the saint Sidi Bu-Qasbaya exist, all of them fairly large mzaras: the shrines of Sidi Bu-Qasbaya al-Wilda are situated at 0.3 and 1.5 km south and 0.5 km north of the parental shrine.
Here we encounter a most interesting phenomenon, which occurs time and again in saint worship featuring localized shrines: the material multiplicity of shrines associated with one and the same saint tends to create several more or less autonomous cultic foci, despite the fact that the participants are fully aware that all these shrines the same saint is venerated. Thus the various shrines of Sidi Mhammad and Bu-Qasbaya are each in their own right objects of ritual attention. Having a relationship with a saint does not mean that one can venerate that saint at just any shrine associated with him; one has also specific relationships with shrines. One cannot however visit any of the three shrines Sidi Mhammad al-Wilda unless a part of a ritual cycle, which, within the same week or so, also includes a visit to Sidi Mhammad al-Kabir; and the rules of etiquette, which apply in man-saint relationships just as in man-man relationships, would suggest that one visits Sidi Mhammad al-Kabir first. The point is that the shrine , as a material entity, takes on a personalized and autonomous aspect more or less independent from the invisible saint to which it refers; for no participant would maintain that the saint venerated at the shrine of Sidi Mhammad al-Wilda is a son of the saint of the hill-top it is the shrine itself which is the child of the other shrine, and which functions as an irreducible focus of ritual action rather irrespective of the saint with which is associated. this is summarized in the Khumiri maxim: baraka wahada; nzuru kull (it is the same grace, but we visit them all). And it is precisely the shrines capability of taking on such cultic autonomy which enables them to function as beacons in the segmentary structure, even when so many shrines bear the same name.
The reader may have noticed that for the highest territorial levels no characteristic attributes have been mentioned. Major shrines, with or without adjacent cemeteries, function as such. As in the nineteenth century, every Khumiri valley has a major shrine which serves as its characteristic attribute, and which provides a focus for ritual interaction and identification for people whose life-world is contained within the same steep mountain ranges, even though their day-to-day economic, social and political lives, as members of different villages, only infrequently intersect. But there is more. While the attachment of more or less utilitarian characteristic attributes (dwelling-house, threshing-floor, spring, mens assembly) to territorial segments could be seen as a spilling-over, into the symbolic order, of the essentials of the economic and social process, this system is again duplicated in this sense that lesser shrines, in addition to these utilitarian attributes, can be seen to function as ritual attributes of lower-level segments, from the compound level onwards. There are too many territorial segments at the lower levels to make it possible for each segment to be uniquely and exclusively associated with one local shrine. Patterns of shrine ritual are however such that each segment above the household level can be said to be characterized by a fairly unique pattern of saint veneration, in which a number of shrines, venerated with different frequency and intensity, combine in a manner that is manifestly and characteristically different from the combination obtaining in complementary segments. In ways which will become increasingly clear in the course of my argument, shrines are intimately associated with segments; and as can be shown on the basis of a detailed reconstruction of the residential history of the valley of Sidi Mhammad and adjacent valleys since c. 1800, the creation of filial shrines of the saints Sidi Mhammad and Sidi Bu-Qasbaya is a direct reflection of the fission, migration, and relative waxing and waning of social groups in that area since the middle of the last century. These processes occur throughout Khumiriya, and invariably find expression in the geographical distribution, and nomenclature, of shrines.
However, the fictive genealogy of humans, encompassing all living inhabitants of the valley via the ancestral toponyms of their villages and neighbourhoods, is never systematically mirrored by a fictive genealogy encompassing all saints and shrines in a valley easily a score or more. The multiplicity of shrines associated with the same saint, and the non-kin relations supposed to exist between many saints whose shrines are situated near each other, render such a saintly genealogy impossible. Shrine and segment are united not through a saintly parallel of human genealogical fictions, but through patterns of pious visits establishing relationships between saints and the living.
5. SAINTS AND THE LIVING
Let us therefore now turn from saint-saint relationships to the relationships that the people of Khumiriya claim to exist between living men, and saints. There is no doubt as to the human nature of saints. However exalted their powers and grace are, the legends about them depict them as recognizable human beings, whose exploits of piety and wonder-working often contain a touch of humour and human weakness. The extremely complex and protean semantic and symbolic properties of sainthood in Khumiriya cannot be adequately summarized here. For instance, to stress that saints (as indicated by their most frequent designation: uli) are Allahs friends and derive their baraka from Him, would underplay the fact that for most practical and ritual purposes Khumiri saints (not unlike the several shrines with which they are associated) are conceived as autonomous supernatural beings, whose dealings with living humans hardly require Allahs rubber-stamp.
Saints have the power to open up the potentialities of nature and human life for those humans who approach them in the proper manner, i.e. respectfully, sincerely (qalb bahi), and with pure intention (niya). There are few provinces of life that are considered to be outside the power of saintly intervention. Saints are invoked to send rain, to assist in the reproduction of domestic animals, to cure madness and reproductive troubles in humans, to enhance the general economic and physical well-being of the family, to control and ward off jnun (spirits of the wilds), to enhance the baraka of the house, the threshing-floor, the spring and the mens assembly, to protect people who depart on a long journey, to help people in their careers, to render supernatural sanctions to oaths, to inflict misfortune on humans at the request of their human rivals, etc.
Much of this saintly intervention is taken for granted, as the automatic result of the routine aspects of the saintly cult in which every Khumiri is involved: the frequent invocation of the names of local saints, the regular dedication of a meal to a specific saint, and the pious visit (zyara), at least twice a year, to the local shrine or shrines of that saint. At the latter occasion a small offering of incense and candles is left at the shrine, and specially prepared and dedicated oil cakes are consumed, which after having been consecrated at the shrine, are full of the saints baraka. This ongoing routine of the saintly cult is characterized by great spontaneity, fondness and trustful reliance implied in the main descriptive (as distinct from addressive) kinship term Khumiri people employ for their local saints: jaddi, jadda (my grandfather, my grandmother). Although immensely powerful, the saint is not usually thought of as a stern figure of authority, but rather as a grandparent who, like a real grandparent, can afford to spoil his grandchildren, the living humans, since their disciplining is left to an intermediate generation. This quality of fond intimacy stands out clearly when people recount hagiographic legends about their saint, share a meal dedicated to him or her, or when women, in the course of zyara, shed their socially-imposed reticence, and in near-ecstasy dance near the shrine, fondle and kiss the walls and the sacred objects there, and exclaim jaddi, jaddna (grandad, our grandad).
While the saint, deceased and invisible, is considered a grandparent, the kinship term jadda carries an interesting additional connotation: it also means lineal or collateral ancestor in general. Supposed (often erroneously) to be buried at the main shrine that carries his or her name, the Khumiri saint is considered to have lived in the same area in some undefined past, and to be, somehow, among the set of local ancestors. but never is the saint the imputed apical ancestor of a social group, to whom descent is traced through a genealogy. Likewise, the ancestors that gave their names to social and territorial units at various levels of segmentation, are never saints. The two sets of personalized historical symbols do not overlap. In rare cases a saint is claimed to have been a brother of a local apical ancestor, but it turned out to be impossible to let participants pinpoint any living lineal descendants of the saints venerated at local shrines; even when my own historical research convinced me that at least one of these saints, Sidi Mhammad, had actually lived in the area during the nineteenth century, and I thought I could identify his living descendants whose saintly origins had gone lost under the historical and ideological constructions of the contemporary participants.
Outside the ongoing routine of the saintly cult, there are three complementary modalities fro the relationship between man and saint, in addition to the trustful intimacy of the grandparent idiom.
First, in particularly important matters the implicit reliance on saintly intervention tends to give way to explicit supplication. Reminding the saint of the supplicants ritual prestations in the past, and stressing the (fictive) kinship relation between man and saint, the supplicant describes his or her plight and entreats the saint to intervene. Such supplication normally takes place at the saints main shrine, in the course of zyara. All the predicaments summed up above may apply. Normally supplication is made to one of the saints associated with the territorial segment to which the supplicant belongs. In rare cases, however, typically having to do with illness and impaired human fertility, supplication may be made at distant shrines, associated with one of the regional saints that are well-known throughout Khumiriya. On such occasions the usual, small pious gifts are augmented by more substantial offerings, such as: an expensive, elaborately adorned candle; a flag; a meal dedicated to the saint and eaten at home; a similar meal but prepared at the shrine and distributed gratis among passers-by; and, as the highest prestation stipulated in the Khumiri saintly cult, the sacrifice of a domestic animal (chicken, goat, sheep, cow, or bull in a dramatically increasing order of cost, prestige and supernatural pay-off).
Secondly, the prestations accompanying such supplication often assume a conditional aspect. The saint whose special intervention is requested with regard to a specific problem, is promised a substantial offering, only to be made if the saintly intervention turns out to be successful: if a previously barren woman produces a child, if a mental patient regains sanity, etc. Often these conditional promises take on the nature of a gamble. thus saintly protection over a herd of cattle or a brood of hens is ensured by promising the saint a male specimen of that years calves or chicks as a sacrifice; if no males are produced, the saint has to accept that his intervention will go unrewarded that year.
Supplications, particularly if of a conditional nature, introduce a contractual element into the man-saint relationship, that stands in some tension with the inclusive, generalized pattern of the grandparent model. Here the saint appears more as a patron. However, both as a patron with whom one has struck a dyadic, conditional contract, and as a grandparent, the man-saint relationship carries, as a third modality, many obligations for the people involved. However much a saint is supposed to love his living protégés and clients, however much he is prepared to intercede on their behalf, every saint insists on respectful treatment. The same baraka that can, positively, release the possibilities of nature and human life to the peoples benefit, is sure to inflict material misfortune, illness and death, should the people fail in respect, and neglect their general and contractual obligations vis-a-vis a saint. On the basis of these sanctions, the saint protects the integrity of his shrine, the sacred objects and pious gifts it contains, and the immediately surrounding area. The dead that may be buried there, remain undisturbed; and ;the trees, plants and animals there are taboo. He also protects his shrine-keepers, and pilgrims in the course of zyara. he does not allow people to terminate their relationships with him: whoever has entered, at some point in his life, into a relationship with a saint, is under a life-long obligation to make the twice-annual zyara to his shrine and to dedicate meals for him. The saint is supposed to jealously guard his human following against the claims of other saints. Thus the cult of saints acquires an internal momentum of its own which allows it to express and underpin, at its turn, non-religious aspects of life in Khumiriya.
homepage | click here in order to access Part II
 Field-work was conducted in north-western Tunisia in 1968, 1970, and 1979. Although the 1979 field-trip has convinced me that the religious patterns described in this paper have by and large persisted through the 1970s, the ethnographic present in this chapter refers to the late 1960s. I am indebted to the Municipal University of Amsterdam, and to the Free University, Amsterdam, for grants towards my 1968 and 1979 field-trips respectively, and to the Musée des Traditions Populaires, Tunis, for local support. I am moreover indebted to: the people of Khumiriya, Hasnawi b. Tahar, Douwe Jongmans, Jeremy Boissevain, Klaas van der Veeen, and Henny van Rijn, for substantial contributions to my analysis of Khumiri popular religion. An earlier version of this paper was written and presented in 1980, when I was a Simon Visiting Professor at Manchester University; I am indebted to the participants in the anthropology seminar, and particularly to Emrys Peters, Richard Werbner and Kenneth Brown, for helpful criticism made on that occasion. Finally I wish to thank Daan Meijers and Jojada Verrips for organizing the conference out of which the present volume has emerged; Ernest Gellner and Katie Platt for going out on their way in order to accommodate this chapter in that volume; Ria van Hal and Mieke Zwart for typing successive drafts; and F. de Jong for advice on transliteration.
 For the rendering of place-names (including the name Khumiriya), Arabic terms and plurals, cf. Van Binsbergen 1980b: 71, n.7. The system adopted is merely intended to approximate the Khumiri dialect and obviously obscures many of the orthographic and phonetic distinctions. Long vowels in Arabic words are indicated by a stroke whenever the word appears for the first time. In Khumiriya the personal names Muhammad and Mhammad are clearly distinct, with the first a in Muhammad tending towards the Italian a, in Mhammad towards the French è. Earlier ethnographic sources on the cult of saints in Khumiriya include: Dornier 1950; Demeerseman 1938, 1939-40, 1964; Dallet 1939-40; Ferchiou 1972. I shrink from citing here the enormous literature on maghrebine rural popular religion. For a recent bibliographical survey with particular reference to Tunisia, cf. Louis 1977. Studies of maghrebine rural religious behaviour applying the canons of modern social science are, however, extremely scarce. A useful, though more islamological than anthropological, recent survey of popular Islam is: Waardenburg 1979.
 The contentious model of territorial segmentation presented here is argued at great length in Van Binsbergen 1970 and 1971a. This model could not have been formulated but for Gellners (1969) stimulating study of saints and segmentation in the Moroccan High Atlas; cf. Hammoudi 1974 and Van Binsbergen 1971b. For another interesting case of the distribution of religious centres following secular segmentation, cf. Evans-Pritchard 1949, ch. 3.
 While this reflects the historical ideal, the breaking-up of commensality between co-residing kin has led to a situation where households, rather than compounds, are in the possession of their own storage table.
 For a definition of shrine, and a theory of shrines in relation to social organization and the natural environment, cf. Van Binsbergen 1981.
 In Khumiriya, springs emanating from soil with a high iron content contain reddish foam; these springs, which are relatively rare, are invariably the object of a cult.
 The puzzling status of shrines 3 and 4 as filial branches of shrine 1 is discussed in Van Binsbergen 1971a: 281f and 1980b: 71, n. 15.
 For a striking parallel in Andes popular religion, cf. Sallnow 1981.
homepage | click here in order to access Part II
page last modified: 2000-05-17 19:59:01