free web hosting | free website | Business WebSite Hosting | Free Website Submission | shopping cart | php hosting
Free webhosting


Part II

Wim van Binsbergen

homepage | index page 'Íslamic myth' | Part I

5. Aspects of the contemporary social and ritual organization of the region

The hanshir Sidi Mhammad is one of four valleys which the French colonial government in the 1880s united to form one chiefdom, cAtatfa. The other valleys are al Millah, al Mazuz and Shahada. The population of these valleys belongs to a number of different clans and lineage segments. The nature of these social groupings is much too complex for justice to be done to the subject within the scope of this article. I must limit myself to the following summary (see van Binsbergen, 1970 and 1971a for fuller accounts).

                         Every Khumiri places himself or herself in a genealogy based on patrilineal descent. These genealogies usually have a depth of four or five generations, and the participants regard them as repositories of the literal historical truth. In the Khumiri view, contemporary rural society is still governed by a structure of segmentary patrilineages, which is supposed to regulate rights to land, male residence, the nature and intensity of interpersonal ties, and the relations between people and saints. Residential units which are clearly visible in the landscape (from the level of the individual household, via such higher-level clusters as compounds, neighbourhoods, and hamlets, up to the valley and chiefdom level) are supposed to correspond with lineage segments at various levels as defined by the generation in the lineage genealogy. In the participants’ folk theory, therefore, all inhabitants of a valley, and all the residential units at various levels, could be fitted into one large genealogy. Since patrilineal descent uniquely and ideally defines membership of local residential groups and rights to local land, people who at a given moment in time happen to live at a particular spot are under strong ideological pressure to justify their presence there in terms of patrilineal descent from the local apical ancestor. At the ideological level the migration of individuals and groups, and the acquisition of rights to land by means other than filial inheritance, are not recognized. Yet, of course, the various patrilineal descent lines that are locally represented ( I do not call them lineages to avoid confusion with the participants’ ideologically distorted view of social groupings) do not, on the level of some analytical, objective historical truth, converge towards one and the same historical ancestor. Most local descent lines have only immigrated into the valley which they are occupying now in the course of the nineteenth and twentieth centuries; and most have acquired land rights not through filial inheritance but by other much less prestigious means: by matrilateral inheritance, gift, patronage, purchase, violent conquest, collusion with the colonial authorities, and theft. While there is no reason to assume that genealogical manipulation is a recent feature in Khumiriya, the insistence on ideologically acceptable rights to local land appears to reflect, partly, the dramatic increase in population pressure on the land since the beginning of the colonial period. this increase is due to a number of related factors: a five-fold numerical increase of the population in less than a century (the valley now supports 60 inhabitants per square kilometre); the alienation of land for the purpose of state exploitation of cork forests and for the benefit of a few colonist farmers; the politically inspired state alienation of pious-endowment land (habus) attached to shrined and religious brotherhoods; and the concentration of land in the hands of the chiefly family, which i shall discuss in more detail in section 7.

                         At any rate, the existing ideological pressure forces people constantly to revise their genealogies so as to bring these into line with the actual residential situation in the valley, and with the existing social relationships between the inhabitants. It may be attractive for a recent immigrant to trace descent to an undisputed local ancestor: but he will only be followed to do so if his presence is appreciated by those already firmly established in the valley. Alternatively, negative social and political relationships are inevitably expressed in terms of recent immigration: anyone will try to dispute his enemy’s right to local residence, and to participation in the local community. There are other reasons why genealogies are not complete or true reflections of historical relations. When genealogies serve such a clear purpose defined by the pattern of relationships that prevails here and now, there is no point in burdening one’s genealogies with whole series of collateral relatives who have left the valley long ago and who no longer take part in local affairs. As a result, genealogies mainly hark back to those past occupants of a valley who have living descendants there — unless an informant is strategically placed, e.g. as the youngest of an otherwise extinct generation who has personal recollections of the emigrants. Finally, genealogies are constantly revised so as to reflect the relative numerical, political and economic strength of contemporary kin groups who are presented as the descendants of particular ancestors in the genealogy. So as the relative importance of various clusters of agnates waxes and wanes on the contemporary scene, the genealogical relationship said to heave existed between their ancestors is revised accordingly. Segments descending from tow brothers (or less closely related segments that once were equals and allies and for that reason were considered to be ‘as brothers’) may assume a genealogical relationship of father and son, or even grandfather and grandson, if their subsequent fortunes at the local scene take a substantially different course.

                         Since genealogies are so manifestly the outcome of shifting contemporary relations and are only subject to discussion and alteration at abnormal moments, during open conflicts or when an anthropologist comes along with intrusive questions, the manipulation of genealogies by individuals produces genealogical fictions which seldom dovetail neatly. No two informants, not even full brothers, produce exactly the same genealogy of their own line of descent, or of that of their neighbours. And no genealogy taken down in the field can be said to be historically correct in all its detail.

                         In order to serve such ideological purposes the revision of genealogies has to be covert and, in fact, largely subconscious. The situation recorded for the West African Tiv (Bohannan, 1952), of lineage segments publicly revising their genealogies so as to bring them into line with their altered social relationship, is unthinkable in Khumiriya. The revision process works at incredible speed. It is common for a line of descent that has immigrated into a hamlet as recently as thirty or forty years ago to find itself firmly attached to the locally dominant genealogy, provided interpersonal relationships within the hamlet are harmonious.

                         Yet there will always be cases of immigrants whose arrival has been too recent to be included in genealogies in this fashion; or there will be ties too distant to be encompassed within a single master genealogy of the sort which the participants themselves consider historically correct. The main device by which Khumiri genealogies overcome this difficulty is by attaching one or more mythical ancestors to the head of each allegedly historical lineage chart. These ancestors are said to have lived in some unspecified past and, as founders of clans, allow people to claim common descent without having to invent specific and connecting descent lines. The clan name becomes a sort of surname which people adopt (and sometimes alter) without having to overhaul all the more recent elements in their genealogical knowledge. Clan names, and therefore mythical ancestors or clan founders, turn out to be attached to particular areas, for the most part irrespective of the specific genealogical position of the people to which each clan name is attached.

                         Thus summarized and greatly simplified, the social organization of contemporary Khumiriya can be said to rest upon three interconnected principles:

 (a)             A structure of shallow segmentary patrilineages,

(b)             A structure of residential units,

(c)             A structure of localized clans

A structure of shallow segmentary patrilineages

Shallow segmentary patrilineages are continuously re-defined in the process of migration and fission. In so far as it serves as an organizing principle in the Khumiri understanding of their own society, this ideology provides us with an explanation for the genealogical manipulations and the distorted perceptions of local residential history which are so widespread in this area. Therefore, this organizational principle also allows for the detailed reconstruction of actual historical events (i.e. actual residential movements of people in the past), provided we have a sufficient quantity of distorted material at our disposal to inspect and assess the many possible permutations. Genealogies are the most readily available, and the least specialized form of oral-historical evidence in this region. On the basis of some 200 genealogies collected in the late 1960s among inhabitants of the valley and adjacent valleys, supplemented by statements regarding the places of residence of all the individuals concerned and by more comprehensive traditions concerning migratory movements and the attendant social and political repercussions and disputes,[13] I was in fact able to reconstruct, more or less to my own satisfaction, the residential and migratory history of the people in the valley of Sidi Mhammad since c. 1800.

A structure of residential units

Residential units are clearly identifiable on the ground and, beginning with individual households, combine in a segmentary. pyramidal fashion to form compounds, neighbourhoods, hamlets, valleys and chiefdoms. The everyday social process that determines the economic, social and political structures in contemporary Khumiri life mainly takes place within these residential units. At all levels (except that of the individual households) they are heterogeneous as far as unilineal descent is concerned. For although these residential units are named after lineage segments and are considered to be founded by ancestors belonging to a comprehensive fictive patrilineage encompassing an entire valley, in fact most compounds, and all neighbourhoods and hamlets, comprise more than one patrilineal descent line. The social relationships that inform the continuous manipulation of genealogical ties so as to bring them into line with the participants’ patrilineal ideology are mainly acted out within these residential units. Moreover, the contemporary local ritual structures can be adequately described and explained in terms of these residential structures. From the most inconspicuous mzara concealed somewhere behind a cactus hedge to the qubbas that are the focus of massive festivals, the patterns of collective and individual pilgrimages, offerings and sacrifices, dedication of meals, and burials within any valley are entirely determined by the fact that each shrine is attached to a residential unit at one level or another (cf. van Binsbergen, in press). Thus Sidi Mhammad al Kabir serves as a characteristic attribute of the residential unit encompassing the total valley of Sidi Mhammad. The kurbi halfway towards Mayziya, however, may share the name of Sidi Mhammad, but it is only a characteristic attribute of one hamlet, that of Mayziya, and does not feature in the ritual activities of the inhabitants of other hamlets in the valley. the same applies to Sidi cAbd Allah at Fidh al Missay. Likewise, at the residential levels below the hamlet level there are a considerable number of mzaras whose name I have not mentioned here but which serve as the characteristic attributes of these lower-level units, and thus as foci for (typically lesser) dedications and offerings, and small-scale collective pilgrimages exclusive to the members of one compound or neighbourhood. In the collective pilgrimages which entire hamlets direct to a major shrines such as Sidi Mhammad al Kebir, the impact of these lower-level units becomes eminently visible, when from the various households women (under the supervision of one elderly woman from their midst who is referred to as the kabira) team up to form a group of pilgrims representing their compound; then various compound groups join into one neighbourhood groups as they proceed; until the various neighbourhood groups under their respective kabiras team up into one massive procession on their way to shrine. On their return home the same process can be observed in the reverse order. It is in the contemporary ritual structures that the social organization of Khumiri society becomes most manifest. Shrines are the symbols of residential units which pose as kin groups and desperately revise their genealogies in order to conceal the fact that they are not kin. This is in fact the underlying meaning of the participants’ characterization of the saints as jadudna. But in order to perform this function it seems imperative that the local saints should never themselves figure in the genealogies. They are primarily attached to residential units. If, through the inclusion of local saints in their genealogies, patrilineages were allowed to lay exclusive claims by birth-right to these saints, the integrative symbolic function which the saints now provide for the residential units as a whole would be jeopardized.[14]

A structure of localized clans

A structure of localized clans loosely categorizes the population and which is so flexible that it facilitates the transformation of lineage segments into residential units, and permits the latter to pose as lineage segments.


With these summary insights into the contemporary social organization of the region set against the background provided by a painstaking reconstruction of the settlement history of the valley of Sidi Mhammad since c. 1800, we may now return to the myth of Sidi Mhammad and see if it can be shown to contain some specific historical information in addition to the cultural and symbolic messages explored in section 3. We know that shrines and saints serve as symbolic foci for the collective identities and activities of contemporary residential units. They are more than likely to have done so in the past. If, in the past, the valley of Sidi Mhammad turns out to have been the scene of continuous migration, and if the existing residential units have therefore been subject to continuous alterations in their composition, the shrines and saints that characterized and symbolized those units are likely to have been distributed and redistributed as a reflection of these social processes of fission and fusion. The myth of Sidi Mhammad speaks of the geographical displacement of a saint, and of his attaining independence vis-à-vis another saint. Would it be too far-fetched to read into this myth the record of an actual migration of a social group that has taken the saint Sidi Mhammad as its focus, its characteristic attribute, and its symbol? And would it be possible, on the basis of the information available on the area, to pin-point this social group? To answer these questions, we shall now look at other myths which present relations between saints and explore the historical background of these myths.


6. Saintly myths and the history of shrines in Khumiriya

Though the myth of Sidi Mhammad and Sidi Salima may be the best-known myth in the region, it is by no means the only myth featuring saints.

                         One common type of myth seeks to explain the presence, within one valley or two adjacent valleys, of the shrines of several saints not having the same name. Although dead and invisible, the saints are more or less considered to be dwelling in the area, and when all ordinary human beings inhabiting the same valley should ideally fit into one genealogy, participants are inclined to apply the same model to saints. So the presence of lesser shrines within a valley is often explained in terms of the saints associated with these shrines being junior relatives of the valley’s main saint (the one with the most elaborate festival). the myth usually takes the extremely simple form:

                                 Sidi X was the younger brother/sister/son/etc. of Sidi Y.

It is important to stress that this myth forms a kind of productive model, in which any minor saint can be substituted ad libitum; in other words it is a model which informants are prepared to invoke as a standard explanation even if they cannot give any more specific, colourful details concerning the relationship between the saints involved.

                         There is another folk explanation of the relationship between saints which is even more significants because it introduces a non-kin connection which suggests that strangers or immigrants have been incorporated: it runs thus:

                                 Sidi X was the friend/servant/herdsman of Sidi Y.

The myth of Sidi Mhammad and Sidi Salima clearly forms an elaborate version of this type of myth. But also of Sidi Tuhami of the hamlet of Khamaysiya in the valley of al Millah, it is said that he was the servant of Sidi cAmara, that valley’s main saint. Likewise, Lalla Bu Waliya, a female saint associated with a mzara in Tra’aya-sud, is said to be the servant of Sidi Mhammad.

                         Sidi Mhammad and Sidi Salima ultimately became friends and neighbours. Two saints who according to a local myth, have always been friends were Sidi cAbd Allah and Sidi Bu Naqa.

So inseparable were they after their death their servant Hallal put the former on a male camel and the latter on a female camel (naqa), to travel to a place where he could bury them. Wherever the animals stopped Hallal would start to dig graves, but the animals would always get up and continue their journey before the graves were ready. People living in those parts would turn the un-finished graves into shrines, either for Sidi cAbd Allah or for Sidi Bu Naqa. finally the camel stopped among the Huamdiya clan in the Salul chiefdom south of cAin Draham. Here Sidi cAbd Allah was buried. the female camel stopped in the chiefdom of Homran, were Bu Naqa was buried.

This myth explains to the participants’ satisfaction the occurrence of a number os shrines having the same name throughout Khumiriya. We have already encountered Sidi cAbd Allah in Ulad bin Sayid, Fidh al Missay and Ramal al cAtrus. To this is now added one in Salul, while a fifth exists in Ulad Musa, in the valley of Babush. All these places are connected by ancient footpaths, along which pilgrims, traders and local people going to the regional markets must have travelled for many centuries. In addition to the major Sidi Bu Naqa shrine in Homran there is a minor mzara of that name east of Tra’aya-bidh, and others are likely to exist a few kilometres further to the east. Incidentally, Hallal was never raised to sainthood, but instead became the mythical ancestor of the Ulad Hallal at Huamdiya, another example of the implicit rule that saints do not appear in genealogies.

                         There is also a less ornate topographical myth to explain shrines having the same name. This myth lacks specific references such as those to the servant Hallal and to the animals, and constitutes rather a productive model similar to the one discussed above. It takes the following form:

Sidi Z travelled through the countryside. Wherever he sat down or slept, people created a shrine for him. Therefore today we find shrines for him at A, b, C, D, etc.

That this is in fact a productive model which people apply to any actual case of a number of shrines having the same name, irrespective of more specific mythical or historical knowledge, is clear from the fact that I have heard this myth applied not only to the multiple shrines of Sidi Mhammad and Sidi Bu Qasbaya, but also to those of Sidi cAbd Allah!

                         Finally, there is a local explanation for a number of shrines having the same name, which on the one hand forms a productive model and may be applied, just like the preceding explanation, to all cases with which an informant is confronted, but which on the other hand turns out to have very solid foundations in living memory. When people emigrate from one area and settle in the next, they cannot take with them the shrine that is the main characteristic attribute of the residential unity they are leaving behind. In many cases they will join an existing residential unit where they will be received as dependents, clients, herdsmen, etc. In those cases they will not be in a position to erect their own shrines, and instead will try to ingratiate themselves with their hosts, and with the latter’s saints, by directing ritual activities to the shrines in their new place of residence. when, however, they move to a relatively unoccupied are, or when they emigrate to an area not as dependents but as purchasers of land or even as violent invaders, then they will insist on erecting there a branch of the shrine of their area of origin — as a sign of their identity, as a focus for the new residential unit they are in the process of establishing, and particularly as a symbol of their recently won independence vis-à-vis the residential unit they have left behind. From the original shrine they take a few relics: the bones of the saint if these in fact can be found there, or else a musba or tassa. Around these, the new shrine structure is erected. The new shrine receives the same name as the original shrine; when both are in the same valley, the latter is distinguished by the addition ‘al Wilda’, the son.

                         There is ample oral-historical evidence that several of the shrines having the same name in the valley of Sidi Mhammad have been created in this fashion during the course of the nineteenth and the early twentieth centuries. In those case informants could mention the name of the historical people who actually created these branches, and could indicate some of the surrounding circumstances. This might seem paradoxical in the light of my previous statement that in Khumiri society migration, the precondition for the creation of filial branches of shrines) was not cognitively recognized. However, in the course of oral-historical fieldwork one builds up, with some informants at least, a level of trust that allows one of penetrate beyond their formal narrative image of their own society. Particularly if one’s informants are unusually intelligent and belong to generations otherwise extinct, and if the researcher can feed into his interviews bits of information that indicate that he has already advanced some way towards a more objective truth, and if, finally, one can play off such tensions and rivalries between residential units and lineage segments as exist in this highly competitive society, then, glimpses of the objective truth concerning events in the last century may yet be revealed. Much depends also on finding the proper operational translation for a research question. For instance, informants would not have a clear picture of the succession of cemeteries in the valley of Sidi Mhammad. But they would know where, ever since c. 1900, specific people have been buried; and thus the history of cemeteries (and the attendant shrines) could be gleaned from shift in lists of individual burials.

                         It became fairly well established that of the three shrines named Sidi Bu Qasbaya in the hamlet of Sidi Mhammad, two have been created in the early twentieth century, after a European colonist built his farmhouse on the original shrine and cemetery of that saint. The Sidi Bu Qasbaya shrine in Fidh al Missay already had been created in c. 1870 as a filial branch of that same shrine. Likewise, Sidi cAbd Allah in Ramal al cAtrus was created around 1850 upon relics taken from the shrine of that name in Fidh al Missay. Similar processes were recorded for some of the minor shrines in and around the hamlet of Sidi Mhammad. The most significant case, however, for the interpretation of the myth of Sidi Mhammad was the creation of Sidi Mhammad al Wilda on the basis of relics taken from Sidi Mhammad al Kebir. this event took place around 1900, when two of my informants were small boys and witnessed this activity. Both of these shrines were then kurbis; their transformation into qubbas was only effected in the late 1910s by a European contractor under contract to a local chief.

                         This solid piece of evidence suddenly transports the shrine of Sidi Mhammad al Wilda, and its associated saint, from the obscurities of antiquity into the more sharply delineated world of recent events. It strengthens our hope of penetrating the history of the original shrine of Sidi Mhammad al Kebir, and of understanding its relationship with the figure of Sidi Salima.[15]


7. The history of Sidi Mhammad

Sidi Mhammad al Wilda was erected on a flat stretch of land between the cemeteries of Sidi Rahuma and Sidi Bu Qasbaya following the migration of the shrine-keepers and their associates to the area one kilometre south of Sidi Mhammad al Kebir. This migration followed a dramatic and violent conflict that took place at the festival of Sidi Mhammad. For, before the migration of the shrine-keepers, the original shrine of this saint was not a place of death, but the site of one of the most famous zardas (saint’s festivals) in Khumiriya. On such occasions not only did the inhabitants of the local residential unit, and the out-married women who were under obligation to return annually to their shrine, forgather at the ritual centre, but they were also joined by pilgrims for the surrounding valleys; and even when interaction between the Khumiri groups was characterized by violence and feud, such pilgrims were assured of safe-conduct by virtue of the sanctions attached to the supernatural powers of the saint himself. At the same time the zardas were (and they still remain) the principal occasions when the local population revealed their strength, their alliances and the splendour of their saint and shrine. The ensuing sense of competition has been known to raise tempers, not only in the past but also in recent years.

                         During the zarda of Sidi Mhammad, c. 1900, a male pilgrim from Ulad bin Sayid insulted the local men by making sexual allusions concerning their wives and daughters. As a result, he was put to death. this bloodshed triggered further violence and disrupted social relations in the hamlet of Tra’aya near the shrine, to such an extent that the hamlet split in two parts, ‘Bidh’ (White) and ‘Sud’ (Black), and a part of the original population migrated to what is now the hamlet of Sidi Mhammad. This event by no means forms the explanation of the myth of Sidi Mhammad; but it indicates the existence of long-standing tensions between the groups living around the shrines of Sidi Mhammad and Sidi Salima, and the role which the shrines and their festivals have played in enhancing these tensions and bringing them to a critical point.

                         But who were the people who were in control of the shrine of Sidi Mhammad around 1900, and who, by the creation of Sidi Mhammad al Wilda, established the conditions under which the present hamlet of Sidi Mhammad could emerge, thrive, and become the site of a major zarda, whereas Sidi Mhammad al Kebir declined and became a mere cemetery?

                         It is an indication the great changes that have taken place in the valley of Sidi Mhammad, and in the chiefdom of cAtatfa as a whole, that the contemporary situation does not even suggest, at least on the surface, what I now take to be the correct answer.

                         Sidi Mhammad today is a hamlet dominated by the Zaghaydi clan. the Zaghaydiya ultimately trace their descent to the mythical ancestor Zaghdud, who is claimed to hail from the holy city of Kayrwan, in Eastern Tunisia. One of the major constituent lines of descent within this clan has supplied all the chiefs of cAtatfa since the office was created in the 1880s (with only a short interruption immediately after Tunisia became independent). Moreover, the confederation of clans which gave its name to the cAtatfa chiefdom was created in the 1870s by another member of that same line of descent. Today the most wealthy people in the valley of Sidi Mhammad, and indeed in the chiefdom as a whole, belong to this chiefly family. Most of the land in the valley of Sidi Mhammad belongs to people of the Zaghaydi clan, and particularly to the chiefly family. However, the expansion of the Zaghaydiya is relatively recent; it depended largely on their association with the French colonial power, and on their ability to convert this association into lasting economic power and influence through land ownership, education, and office in independent Tunisia (where members of this chiefly family are holding posts in l local government, unemployment relief work, etc).

                         The Zaghaydiya present a totally different picture from the other main clan in the chiefdom, the cArfawiya. This clan, an offshoot of the Drid tribe, traces its origin to the mythical ancestor cArfa. From times past, they have been associated with the religious brotherhood of the Shabbiya, and with tax collection. Contrary to the Zaghaydiya, the cArfawiya have a specific clan myth which centres upon the head of a partridge that is said (by the cook who had eaten it) to have been burned in the cooking-fire. The cArfawiya settled in the valleys of Sidi Mhammad and al Millah around 1800. In 1870 they created in the latter valley a lodge (zawya) for the Qadiriya brotherhood which later was moved to the valley of Shahada. Members of the cArfawi clan still control this lodge, the only one of its kind in the region of cAin Draham. Moreover, the cArfawiya are considered to be strongly represented among the lodge membership, and to be more expert than other groups in the ecstatic dancing that is the brotherhood’s main ritual in Khumiriya. Throughout the chiefdom, the cArfawiya command considerable religious prestige (further enhanced by their claims to a purer Arab descent than most Khumiris, by their predilection for horses, etc.); but their economic and political power is, these days, hardly comparable to that of the Zaghaydiya.

                         The Zaghaydiya and the cArfawiya ar by no means the only clans in the chiefdom of cAtatfa. Yet these two clans, whose interactions have constituted the main political and religious developments in the chiefdom for the past hundred years, have been so prominent that they have imposed a moiety-like structure upon all the valleys in the chiefdom except the southern part of al Millah. Most people would claim identity as either cArfawi or Zaghaydi — even those who belong to older descent lines that traditionally link up with mythical ancestors other than cArfa or Zaghdud. Their own mythical ancestors (such as Bu Maza, Bu Tara, Rashab, Bu Dabus) are then treated as descendants of either of the founders of the two dominants clans; for instance, the Tra’ayi and Mayzi clans in the valley of Sidi Mhammad today largely pretend to be members of the Zaghaydi clan.

                         At present one finds members of both the Zaghaydi and cArfawi clans residing in the valley of Sidi Mhammad, but the saint Sidi Mhammad is strongly associated in the popular mind with the Zaghaydi clan. Within this valley thecArfawiya are now exclusively associated with the western part of the hamlet of Sidi Mhammad, and with Fidh al Missay and Ramal al cAtrus; and here they are associated not with the saint Sidi Mhammad but with Sidi Bu Qasbaya and Sidi cAbd Allah. Even, since at present the cArfawi clan is most prominent in the valley of al Millah, whose major shrine is Sidi cAmara, it is suggested that the main saintly patronage of the cArfawiya who reside in the valley of Sidi Mhammad should lie with Sidi cAmara rather than with the saint Sidi Mhammad.[16] The people of the Zaghaydi clan, on the other hand, live closest to the four shrines of Sidi Mhammad; and the wealthy powerful Zaghaydiya in the hamlet of Sidi Mhammad, including members of the chiefly family, have a major say in the organization of the zarda. For the past forty years, the keepers of the shrine have been Zaghaydiya, close relatives or clients of the chiefs. The present-day fuqra specializing in the ecstatic dance for Sidi Mhammad largely have the same relationship vis-à-vis the chiefly family.

                         Yet the ecstatic dance, and the Qadiri brotherhood within which it is loosely incorporated, is primarily an cArfawi affair. What is more, Sidi Mhammad was originally an cArfawi shrine!

                         In fact, the present Zaghaydi control over the shrines and the cult of Sidi Mhammad dates back only to the 1920s. In half a century the Zaghaydiya, and especially the chiefly family, went through a dramatic expansion in the valley of Sidi Mhammad, both numerically and in terms of wealth, political power and ritual control. their latter-day kinship-based control over the shrine-keepers and the local fuqra have been the outcome of a concerted effort on the part of the Zaghaydi chiefs to break the ritual power of the cArfawiya, and to legitimate their own political and economic power by whatever symbolic support the ritual sphere had to offer. It was through the influence of a Zaghaydi chief that the original festival function of Sidi Mhammad al Kebir was transferred to Sidi Mhammad al Wilda. That chief would personally supervise the collective zarda rituals around al Wilda, which at that time still included the preparation and consumption of a huge meal for hundreds of pilgrims. Again, it was he who converted the site of Sidi Mhammad al Kebir into a cemetery; he even had himself buried directly in front of the entrance to the shrine of Sidi Mhammad al Kebir. By these means he drastically altered the ritual organization of the valley. Before that time the cArfawiya would be buried at the cArfawi cemetery of Sidi Bu Qasbaya, while the Zaghaydiya would be allowed to bury their dead at the Mayzi cemetery of Sidi Rahuma. The creation of the cemetery of Sidi Mhammad al Kebir ended Zaghaydi dependence on the Mayzi cemetery; it also sealed to defeat of the cArfawiya by the Zaghaydiya (for burial at Sidi Bu Qasbaya would be discontinued and henceforth all the cArfawiya within the hamlet of Sidi Mhammad would bury their dead at the Zaghaydi cemetery). However, the Zaghaydi chief could only do this after wrenching cultic control from the hands of the original cArfawi shrine-keepers, whose line of descent had created the original shrine of Sidi Mhammad al Kebir, who had administered the kurbi on the Hill-top until the bloodshed desecrated the zarda there, who had (as a cultic expression of the emigration from Tra’aya of both the Arfawi and Zaghaydi households) then created the shrine of Sidi Mhammad al Wilda, and who had continued to administer that shrine and the original one for another twenty years or so until the Zaghaydiya took over. The change-over is clearly marked in the succession list of shrine-keepers of Sidi Mhammad: after a number of close agnates of a cArfawi line of descent which succeeded each other according to a perfect patrilineal adelphic pattern from the 1870s onwards, in the 1920s suddenly non-cArfawi keepers crop up, all of whom have close relations with the chiefly family. This indicates that the cArfawi keepers lost the control over the cult and were economically and numerically brought to virtual annihilation within the valley, as the Zaghaydi chiefly family utilized their collusion with the colonial authorities to acquire rights over pious-endowment land that belonged to the shrine of Sidi Mhammad. The creation of expansion of the colonist’s farm in the western part of the hamlet of Sidi Mhammad, and the decline, therefore, of the shrine of Sidi Bu Qasbaya, furthered the downfall of the cArfawiya in the valley — a process that reached its culmination in the 1950s when the chiefly family obtained ownership of this farm as well.

                         The dominant Zaghaydi group, therefore, have a strong interest in denying the original cArfawi connection with the shrine of Sidi Mhammad. They laid a dense smoke-screen of historical distortion around what I now take to be the objective historical facts. And significantly, all informants, and, for a long time, I myself, found it difficult to step out of the illusion that the present-day associations between local groups and shrines (shrines that gave the impression of having been there for ever) should be project wholesale into the distant past.


8. Conclusion: The history of Sidi Mhammad and Sidi Salima

We are now finally ready to glean the historical message from the myth of Sidi Mhammad and Sidi Salima. at this point let me remind the reader of the anti-climax which I have already anticipated in my introduction.

                         As my painstaking reconstructions of the residential history of the valley of Sidi Mhammad during this century and the last bear out, a number of distinct groups from the cArfawi clan settled along the Wad al Kebir around 1800 in what today are the hamlets of Fidh al Missay, the western part of Sidi Mhammad, Ramal al cAtrus, and both Tra’ayas. They hailed from the area around Sidi cAbd Allah in Salul, and along with their awareness of belong to the Drid tribe they brought with them the cArfawi myth of origin featuring the burned partridge. So closely associated is the cArfawi clan with the valley of Sidi Mhammad that the mountain slope west of the Wad al Kebir facing the hamlet of Sidi Mhammad is still called Raqubat cArfa, after their clan founder. Many informants make specific reference to this place name in their version of the myth of Sidi Mhammad and Sidi Salima (elements 6 and 10).

                         One of these immigrant cArfawi lines of descent was that of the original pre-Zaghaydi shrine-keepers of Sidi Mhammad. In contrast with their fellow-clansmen, they were pacifists. On various occasions during the turbulent nineteenth century, they would intervene in the battles which the militant and expanding cArfawiya fought with earlier inhabitants of the region; carrying the flags of their shrine, the shrine-keepers would come to the battlefield and exhort the parties to end hostilities. Among all the lines of descent in the region, these shrine-keepers come closest to the type of pacifist saintly lineages which Ernest Gellner (1969)[17] describes for the High Atlas, some 1,500 kilometres to the west.

                         The prominence of the partridge in the myth of Sidi Mhammad and in the ecstatic song associated with him; the fact that, unlike the other clans in the area, the self-perceptions of the cArfawi immigrants supported their identification with the Ulad bin Sayid on the basis of common affiliation with the Drid tribe; and the occurrence of shrines for Sidi cAbd Allah both among these cArfawi groups in the valley of Sidi Mhammad, and among the Ulad bin Sayid — al these items of evidence lead to the conclusion that the myth of Sidi Mhammad and Sidi Salima symbolically embalms the historical interactions that occurred between the cArfawiya of Tra’aya and their close neighbours, the Ulad bin Sayid of Khadayriya, during the first half of the nineteenth century. Their common association with the Drid tribe enabled the immigrant cArfawiya to find hospitality and patronage among the Ulad bin Sayid living around the shrine of Sidi Salima. It is most likely that the early cArfawi immigrants in the Tra’aya area received not only land to the south of the Wad Ghanaka to settle on, but also the right to bury their dead in the cemetery of Sidi Salima. However, as the immigrant group expanded, they asserted their own distinct identity vis-à-vis their hosts, and created their own shrine. Partly because of its strategic location in the ecology of the region and partly because of the backing with the guardian lineage received from their non-pacifist clansmen, within a few decades this shrine became one of the major shrines of Khumiriya, worthy of a myth that is known throughout the region. This transformation of the cArfawi/Ulad bin Sayid relations from one of dependence to one of equality took place in the first half of the nineteenth century, and the shrine of Sidi Mhammad al Kebir must have been built by about 1850, that is, by the time the eldest remembered cArfawi guardian of that shrine was born. the bones which, half a century later, my informants saw dug up and transferred to another site to create the shrine of Sidi Mhammad al Wilda. , must have been those of a man, very likely called Mhammad, who lived and died in Tra’aya in the first half of the last century.[18] True to type, he does not occur in the genealogies of the guardians line of descent. Nor would he ever have sat on the Hill-top as a herdsman of Sidi Salima, who by that times must have long since rested at the cemetery that bears his name. It is likely that the partridges that alighted on Sidi Mhammad’s shoulders flew out of the Arfawi clan myth rather than from Heaven. And the cArfawi connections itself has been all but concealed in the course of half a century of Zaghaydi expansion. It is my contention that the devious ‘mountain paths’ of my historical reconstruction have reconstruction have brought the myth of Sidi Mhammad and Sidi Salima into the ‘plain’ of history in the sense in which ‘history’ is commonly understood by scholars today.



Beeker, M.C. (1967) ‘Mozaïek van het wonen’, University of Amsterdam, Drs. Soc. Sc. dissertation.

Bel, A.(1938) La Religion musulmane en Berberie, vol. I, Paris: Paul Geuthnmer

Blok, A. (1979), ‘Rams and Billy-goats: Breaking the Mediterranean Code of Honour’, paper read at the conference on ‘Religion and Religious Movements in the Mediterranean Area’, Amsterdam: University of Amsterdam and Free University.

Bohannan, L. (1952), ‘A Genealogical Charter’, Africa, 22: pp. 301-15.

Bourdieu, P. (1965), ‘The Sentiment of Honour in Kabyle Society’, in J. Peristiany (ed.), Honour and Shame, London: Weidenfeld & Nicolson, PP. 191-242.

Brunel, R. (1926), Essai sur la confrérie religieuse des cAîssâoûa au Maroc, Paris.

Cuisenier, J. (1962), ‘Endogamie et exogamie dans le mariage arbe’, L’Homme, 2: pp. 80-1205.

Davis, J. (1977), People of the Mediterranean, London: routledge & Kegan Paul.

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

Dermenghem, E. (1954), Le Culte des saints dans l’Islam magrebin, paris: Gallimard.

Eickelman, D. (1976), Moroccan Islam, Austin and London: University of Texas `press.

Favret, J. (1968), ‘Relations de dépendance et manipulation de la violence en Kabylie’, L’Homme, 8, pp. 18-44.

Geertz, C. (1968), Islam Observed, New Haven and London: Yale University Press.

Gellner, e. (1969) Saints of the Atlas, London: Wiedenfeld & Nicolson.

Gibb, H.A.R., and Kramers, J.H. (eds) (1974), Shorter Encyclopaedia of Islam, Leiden: E.J. Brill.

Hartong, A.M. (1968), ‘De Geschiedenis van het sjeikaat Atatfa op Basis van de Orale Traditie’, Catholic University of Nijmegen, Drs. Soc. Sc. dissertation.

Heusch, Luc de (1972), La Roi ivre ou l’oirigine de l’etat, paris: Gallimard.

Jongmans, D. (1968), Meziaa en Horma’, Kroniek van Afrika, 3: pp. 1-34.

leach, E. (1976), Culture and Communication, Cambridge: Cambridge University Press

Leach, E. (ed.) (1967, The Anthropological Study of Myth and Totemism, London: Tavistock.

Lévi-Strauss, C. (1958), Anthropologie structurale, Paris: Plon.

Lévi-Strauss, C. (1964), Le Cru et le cuit, Paris: Plon

Lévi-Strauss, C. (1966), Du miel aux cendres, Paris: Plon.

Lévi-Strauss, C. (1973), anthropologie structurale deux, Paris: Plon.

Marçais, W., and Guiga, A. (1925), Textes arabes de Takroûna, paris: Editions Ernest leroux.

Miedema, A.W.F. (1967), ‘Verslag Leeronderzoek tunesië, University of Amsterdam, Antropologisch-Sociologisch Centrum, typescript.

Miller, J.C. (1976), Kings and Kinsmen, Oxford: Clarendon Press.

Miller, J.C. (ed.) (1980), The African past Speaks, Folkestone: Dawson.

Montet, E. (1909), Le Culte des saints musulmans dans l’Afrique du Nord, Geneva: Librairie Georg & Cie.

Reefe, T.Q. (1977), ‘Tensions of Genesis and the Lunda Diaspora’, History in Africa, 4: pp. 183-206.

Souyris-Rolland, M. (1949), ‘Histoire traditionelle de la Kroumirie’, Institut des Belles Lettres Arbaes, 12: pp. 126-65.

Van Binsbergen, W.M.J. (1970) Verwantschap en Territorialiteit in de Sociale Structuur van het Bergland van Noord-West Tunesië, University of Amsterdam, Drs. Soc. Sc. dissertation

Van Binsbergen, W.M.J. (1971b) ‘Ernest Gellner: Saints of the Atlas’, Cahiers des arts et traditions populaires, 4: pp. 203-11.

Van Binsbergen, W.M.J. (1979) Explorations in the History and Sociology of Terrritorial Cults in Zambia’, in J.M. Schoffeleers (ed), Guardians of the Land. Essays on Central African Territorial Cults, Gwelo: Mambo Press, pp. 47-88.

Van Binsbergen, W.M.J. (1980) Popular and Formal Islam , and Supralocal Relations: The Highlands of North-Western Tunisia, 1800-1970’, Middle Eastern Studies, 16: pp. 71-91.

Van Binsbergen, W.M.J. (1981) Religious Change in Zambia: Exploratory Studies, London and Boston: Kegan Paul International

Van Binsbergen, W.M.J. (forthcoming), Shrines and Ecstasy in the Social Structure of North-Western Tunisia

Van Binsbergen, W.M.J. (in press), ‘The Cult of Saints in Northwestern Tunisia: An Analysis of Contemporary Pilgrimage Structures’, in. E. Gellner and E. Wolf (eds), Religious of the Mediterranean, Berlin: De Gruyter.

Van Binsbergen, W.M.J. van (1971a) ‘Religie en samenleving: een studie over het bergland van Noord-West Tunesië, University of Amsterdam, Drs. Soc. Sc. dissertation.

Vansina, Jan (1965), Oral Tradition, London: routledge & Kegan Paul.

Westermarck, E. (1926), Ritual and Belief in Morocco, vols. I and II, London: Macmillan.

Willis, Roy (1976), On Historical Reconstruction from Oral-Traditional Sources: A Structures Approach, Evanston, II: Northwestern University. Twelfth Melville Herskovits Memorial Lecture.

homepage | index page 'Íslamic myth' | Part I

[13]    Part of this material was collected by P. Ernsting, P. Geschiere, C. Holzappel, G. von Liebenstein, P. Tamsma (deceased), and myself, in the course of a collective project under supervision of K.W. van der Veen in March/April 1968. I am grateful to these colleagues for their permission to use this material. Earlier accounts of genealogical and oral-historical research in the chiefdom of cAtatfa include that Hartong (1968), while I also gleaned some information from Beeker’s (1967) preparatory study for a housing project (which never materialized) in the hamlet of Sidi Mhammad.

[14]    On the other hand, when a clan founder who figures in a genealogy has some association with sainthood (e.g. in the clan of Ulad al Hajj — ‘Descendants of the Pilgrim’ — in the valley of al Mazuz), he has no shrine locally and is never the subject of a cult.

[15]    The details which informants could supply with regard to the history of Sidi Mhammad al Kabir and Sidi Mhammad al Wilda contrast sharply with the absolute lack of specific historical information concerning the two remaining kurbis of the same saint: the one adjacent to Sidi Mhammad al Wilda, and the one half-way to the hamlet of Mayziya. Mechanical application of either of the two productive models for shrines having the same name led some informants to suggest that Sidi Mhammad might have rested at the sites of these kurbis in the course of his wanderings through the region. But in general the informants were remarkably taciturn on the subject. Elsewhere, I offer a reconstruction of the history of these two kurbis, suggest that they were originally named after a totally different saint who was associated with a clan that had prevailed in the area before the immigrations of the shrine-keepers and their associates from Tra’aya (van Binsbergen, 1971a: pp. 281 ff.);.

[16]    Yet the fact that Sidi cAmara is not called Sidi cAbd Allah, and my reconstruction of the valley’s residential history, suggest that SidicAmara was not originally an cArfawi shrine, but one created by pre-cArfawi members of the Mayzi clan, to whom the cArfawiya had come as client immigrants.

[17]    Cf. van Binsbergen (1971b)

[18]    Why has not the shrine of Sidi Mhammad al Kabir been called Sidi cAbd Allah like the other shrines established by the cArfawiya? The fig-tree in the ecstatic song of Sidi Mhammad cannot be found on the Hill-top today, and although there are traditions of it having been destroyed by lightning at the beginning of this century, it is most likely a vestige of Sidi cAbd Allah bu Karma, in Ulad bin Sayid. The shrine of Sidi Mhammad might originally have been dedicated to Sidi cAbd Allah, only to be renamed after Sidi Mhammad once the Tra’aya cArfawiya had produced from within their midst a saintly man of the name of Mhammad.

homepage | index page 'Íslamic myth' | Part I

page last modified: 2000-05-17 20:02:12