SOCIO-RITUAL STRUCTURES AND MODERN MIGRATION AMONG THE MANJAK OF GUINEA BISSAU
Ideological reproduction in a context of peripheral capitalism
Wim van Binsbergen
In the wake of Marxs classic analysis of the ideological dimension of the capitalist mode of production (Marx 1973; Marx & Engels 1975), a number of leading ideas have developed in the Marxist approach to religious phenomena: religion is seen as (a) ideological reproduction, (b) a structure of material production and exploitation sui generis, or (c) as a structure of ideological production (cf. Van Binsbergen & Geschiere , 1984a: 270f).
Ideological reproduction has received the most attention in theoretical and descriptive analyses so far. From this perspective, religion is seen as the ideological projection, into the celestial and the unreal, of processes of control, appropriation and exploitation that constitute Mans social life and particularly relations of production such as actually existing between humans. By reflecting these relations, and by endowing the phantasms (ancestors, deities, spirits of the wilds, etc.) that constitute these reflections, with a unique, exalted sense of reality and power, the relations of production are underpinned and carried over to new generations (e.g. in rites of passage), and to other parts of the world (e.g. through the spread of world religions, in conjunction with the spread of capitalist and bureaucratic secular structures).
In the simplest form of this pattern there exists a certain correspondence between the relational structure underlying the relations of production, and the relational structure defining the religious sphere: e.g. authority relations between elders and youths, or between the sexes, in real life, may be reflected in local ideas concerning the relations between deities and human beings. Here the Marxist approach differs only in idiom from classic structuralfunctionalist approaches as developed in mainstream anthropology of religion.
However, relations of production in contemporary societies are usually complex and internally differentiated. They tend not to pertain to one unique mode of production, but to combine a limited number of different modes, each with its specific internal logic as revolving around the central relation of exploitation that constitutes that mode of production; modes of production are linked to each other through a social and historical process of articulation. In such articulation the central relation of exploitation that characterises a dominant mode of production seeks to impose itself upon other, preexisting modes in such a way as to make the latter subservient to the reproduction of the former. Class alliances between the exploiting classes in each of the various modes of production involved constitute a standard form through which articulation is effected. In this complex situation (for which often the term social formation is used; cf. Terray 1969) religion has many options besides simply reflecting, in some onetoone correspondence, the relations of production that make up one of the constituting modes. Various articulated modes can be reflected within one religious system, which then becomes virtually as heterogeneous (in terms of socioritual organisation, conceptualisation and history) as the relations of productions that are involved; if this is the case, not one set of symbols, collective representations concerning the unreal, causality, misfortune etc. permeate the total religious sphere, but a limited number of different sets.
These sets are mutually irreducible, and the logic of each may tune in with the logic underlying one particular constituting mode of production within the social formation. However, besides such multiple correspondence, the religious sphere may contain elements which question, protest against or negate, rather than reflect, relations of production in any of the constituting modes. Finally, in a social formation religious elements may not just display specific relations (of reflection, protest or negation) with specific constituting modes of production such ideological relations may also be developed vis-à-vis the total structure of articulation that makes up the social formation as a whole. Thus, certain religious institutions and religious movements in nineteenth and twentieth-century Africa have been claimed to reflect, within a given social formation, neither an encroaching capitalist mode of production, nor preexisting modes upon which capitalism tried to impose itself, but the very process of the articulation of these modes (van Binsbergen 1981: 42f, 258f).
However, to the extent to which the religious sphere is not a simple ideological reflection of relations of production, but often assumes a great deal of autonomy vis-à-vis such relations, religion can be more than ideological reproduction. It may take on an impetus of its own, and (in the hands of elders, kings, priests, cult leaders) may stipulate a circulation of producers and an appropriation of their surpluses which begin to constitute relations of exploitation in their own right, sui generis. Territorial and regional cults in South Central Africa have been described in such terms, both by Marxist and by nonMarxist writers; but hundreds of other examples from many historical periods and other parts of the world could be quoted as cases in point.
This capacity of religion to give rise to forms of production and exploitation that do not manifestly spring from nonreligious relations of material production and that more or less create their own (semi)autonomous field, or region (in Werbners sense; cf. Werbner 1977) could only be realised because religion is not only a structure of ideological reproduction but also a structure of ideological production: it is not only capable of reflecting and reiterating the logic and the concepts that underlie material relations of production, but is also eminently capable of producing new logics, new concepts, new notions of causality, or presenting such existing ideological elements in a new light.
This form of ideological production is well documented, in Africa and elsewhere, for the case of exceptionally gifted religious innovators, prophets, preachers. Attempts to relegate the latters activities in the field of symbolic and conceptual production (i.e. innovation) to their specific class situation within complex and changing social formations may have been illuminating, but they do contain a certain onesidedness. Contrary to such structurecentred determinism which abounds both in Marxist and non-Marxist social science, religious innovation represents forms of experimentation and free variation which are inherent in the very nature of symbols and the religious order, and not to be explained away by reference to whatever broad groups, classes and historical processes to which the individuals involved may belong.
Moreover, it would be a mistake to think that ideological production in religion only occurs in the context of the inimitable activities of these great religious personalities. Ideological production is a constant and ubiquitous aspect of religious phenomena. All members participating in a religious system are involved in such ideological production in a variety of ways. It is already a case of ideological production when the standard, overall causal explanations of misfortune as defined within a certain religious system, are invoked by the participants in their attempts to explain the details of a specific case that befalls them. Since religion by definition deals with the unreal and is largely concerned with nonempirical referent, the participants interpretations of particular empirical facts in the light of culturespecific religious notions tend to display much more divergence, individual idiosyncrasies and creative vagaries than is commonly assumed by anthropologists of religion. Given the human tendency for symbolic and philosophical experiments, consensus and hence uniformity and unanimity in the religious sphere are mainly achieved (as a more or less exceptional state cf. Fabian 1984:....) when religious elements are subjected to social control. Admittedly the medium of internalisation safeguards a measure of uniform reproduction of religious form and content by a participant without necessitating the constant scrutiny by other participants. It is my contention, however, that the bulk of religious uniformity is achieved, in the African case at least, as an effect of the ad hoc social control mutually exerted by participants upon the overt, interactional, empirical expressions of their religion: verbal and musical utterances that (as forms of interaction involving more than one participant) are made, commented upon, and possibly sanctioned; concrete material objects (shrines, paraphernalia, offerings, payments) and dramaturgical arrangements (rituals, seances) that can be seen and discussed by others. Such continuity and uniformity as a local religious system may display, is primarily anchored in these empirical referents. This is why my casestudy of Manjak religion, in the present paper, will primarily describe these empirical, visible aspects. Most of a religious system however goes beyond them, in the way of implicit meanings, symbols, imagery, notions of causality that are only imperfectly phrased (if at all), and that underlie the material objects and dramaturgical arrangements in ways most participants would be unable to spell out and most researchers would be unable to grasp except through several years of fieldwork. In these intangible ideological aspects there is to the distress of anthropologists looking for structure room for immense free variation and lack of continuity creativity, in other words. In the field of divination, we may find that the participants apply, simultaneously, rival interpretations of the same empirical referents (illness, death, ecological and meteorological disaster); and even if we succeed in explaining this rivalry as a reflection of various individuals of groups antagonism in the economic or political field, the essential leeway provided by the very nature of ideological production should not be explained away in the analysts attempts at social-structural contextualisation. Likewise, the modern study of ritual would stress the creative communicative patterns in ritual, where officiants and clients often belonging to different linguistic and ethnic groups struggle to arrive at some revelatory or therapeutically effective message which, while partly using a recognizable selection of preexisting symbolic means, in its specific combination and dramaturgical presentation could be called unique to the event at hand, and therefore essentially new and unpredictable.
On the face of it, there would be little that is specifically Marxist in such an approach to religion in terms of ideological production. It is rather in the mainstream of cognitive and symbolic anthropology, particularly in the more recent praxeological variant. On closer analysis, however, a number of particularly interesting research themes open up here: the relationship between ideological and material production; the relationship between ideological production and ideological reproduction; the conditions under which the ideological sphere either manages to realise its autonomy or becomes dependent upon such forms of material production as would physiologically, if not logically, appear to form a precondition for all symbolizing; the extent to which the laws that may turn out to govern ideological production (some of these laws have been discussed, under totally different headings, by praxeologists, or by structuralists seeking to formulate something like a universal grammar of symbols and their transformations) are comparable to the laws which Marxist analysis has sought to formulate for material production and exploitation; the extent to which changes in the ideological field may historically be related to changes in material production and reproduction etc.
If the Marxist approach to religion is to make progress, it should begin to address these research questions in earnest.
Within a wider institutional and policy setting prompting research into the therapeutic particularly psychiatric effectiveness of autochthonous West African religion, the above outline of possible themes and relations informed my recent field work among the Manjak of Northwestern Guinea Bissau. Against the background of presentday village society, its productive system, social organisation and political structure, my research was directed at contemporary economic and symbolic structures involving deities and ancestral shrines, and oracles as administered by specialists who often combine divination with somatic curative action.
In the present paper I shall concentrate on the position and religious activities of Manjak labour migrants who, hailing from the administrative divisions of Calequisse and Caió in Cacheu district, spend very substantial portions of their lives in urban centres in Senegal and France, while maintaining close ritual and therapeutic ties with their area of origin. These ties involve a spectacular expenditure of time and foreignearned money on the part of the migrants concerned, and bring out clearly the exploitative nature of local gerontocratic power. Against a more general background of the articulation of capitalism and several noncapitalist modes of production, I shall attempt to answer as a central question: in what respect can these migrants rituals be interpreted as xideological reproduction?
While thus my emphasis is on ideological reproduction we shall briefly consider to what extent can these ritual structures really be relegated to such an underlying pattern of relations of production; they might as well be considered as structures of exploitation sui generis, without specific and detectable links with such material structures of exploitation as make up the local economy and social organisation. Finally, while a change in ideological content and function (such as the institution of divination, and the specialists administering it, appears to have undergone among the Manjak in recent years) represents an obvious case of ideological production, one major set of data gathered in the course of my research will largely remain outside the present argument: the way in which the Manjak rituals at shrines and oracle huts can be said to assume therapeutic effectiveness, by creatively presenting to their migrant clients revelatory insights and guidance that may contain solutions for the spiritual predicaments the migratory experience had landed them in (cf. van Binsbergen & de Jong, forthcoming). For the structure of my argument this has the unfortunate effect that the emphasis, in the theoretical introduction, on ephemeral and praxeological aspects of ritual (as distinct from analysis in terms of enduring social, economic and ritual structures) will not yet be backed, in this paper, by an extensive casestudy, but should merely be read as a statement of intent for future work.
The Manjak (Manjacos, Yagos) ethnic group is found on the peninsula defined by the Cacheu river, the Atlantic Ocean and the Mansoa river. They are the dominant ethnic group in the districts (seccaos) of Calequisse and Caió, both belonging to the Cacheu region. Occupying the central part of the peninsula, they are virtually closed off from the Atlantic by their neighbours the Feloop (a subgroup of the Diola) to the West and Northwest, while the Braam (Mancagne) form the Manjaks neighbours to the southeast. Further eastward, the Braam give way to such ethnic groups as the Papél, Balanta, Mandinka (Malinké) and Fula (Peul, Pular). To the northeast, beyond the old harbour town of Cacheu (once the colonial capital of Portuguese Guinea), lies the inaccessible area of the Cobiana ethnic group.
There is a close affinity between the Manjak, Braam and Papél, in language, agricultural production, religious system and hierarchical sociopolitical organisation. All three ethnic groups display the remnants of small precolonial kingdoms that used to enjoy considerable colonial protection in the Portuguese era and that were dissolved after the PAIGC (Partido Africano da Independencia da Guiné e do Cabo Verde) proclaimed territorial independence in 1973. Significantly, the areas involved were among the latest to be liberated from Portuguese occupation.
Given a similar ecology, the Feloops system of production has many common features with the Manjaks, but the former lack a history of statehood; their language is not intelligible to Manjak speakers. Neither is the Cobiana language; while the Coboiana group remains one of the least studied ethnic groups of the country, and no definite pronouncement could yet be made concerning their historical forms of sociopolitical organisation, there are indications that they form a surviving pocket of an older population preceding Manjak and Feloop settlement in the area.
On the peninsula, the Cobiana area was liberated at an early stage of the liberation war, and from here considerable guerilla activity was waged against Portuguese strongholds at the towns of Cacheu and Canchungo (then called Teixeira de Pinto). The rural areas in the rest of the peninsula, and foremost the Manjak population, have since retained a certain aloofness visÖvis PAIGC politics. Office in local party branches is largely held by nonlocals, and also in nationallevel politics are the Manjak underrepresented. The onetime activities of a Dakar-based political party, opposed to the PAIGC and mainly organised along Manjak ethnic lines, has however not led to marked animosity or antagonism on the part of the national political centre as regards the Manjak.
In the religious domain, such autochtonous forms of religion as will be discussed in this paper still form the dominant idiom with a remarkable degree of interethnic participation across linguistic and socio political boundaries. Of the world religions, only Christianity (in the form of Roman Catholicism) has managed to superimpose itself upon (rather than replace) these authochtonous forms. The inroads of Islam, so conspicuous elsewhere in GuineaBissau and neighbouring countries since the nineteenth century, on the peninsula have remained limited to a handful of trading families at the district centres, although Muslim presence in the town of Canchungo has already warranted the building of a mosque there.
Agricultural production among the Manjak combines a number of main types of cultivation: an annual cycle of paddy-rice cultivation on irrigated fields adjacent to brackish, mangrovecovered rivulets cutting deeply inland (a spectacular form of cultivation found, with minor variations, all over the Upper Guinea Coast stretching from the Gambia to Sierra Leone); annual cultivation of dry forest gardens and small garden plots inside the village, on which bananas, cassava and yams are cultivated; and finally orchards, situated in or near the villages, and yielding palm kernels (from which palm oil is prepared), palm wine, cashew nuts, cashew wine and lemons. Fowls (chickens, guineafowls) form the main domestic animals. Nowadays goats and pigs are rarely raised in the villages. The few head of cattle found there today are invariably owned by the elderly male heads of extended families. There is only an underdeveloped local, regional, national or international market of agricultural produce. Most families experience great difficulties in keeping up their daily food supply in which rice is the staple. The nearfamine conditions which have existed in GuineaBissau for a number of years now, are also encountered here. Fish and shellfish, either caught by female members of the household or bought at the local market, form the most frequent source of animal protein. Hunting is insignificant nowadays.
Stores, either stateowned or private (and in the later case mostly run by Muslim traders from the eastern part of the country), very occasionally offer rice for sale and act as local marketing venues for cashew nuts. A trickle of local (i.e. district level) trade in food crops, palm wine and domestic fowl (exclusively used for ritual purposes, which however may include human consumption; see below) as a source of cash is somewhat supplemented by petty commodity production: pottery, basketry, bandweaving, the preparation of salt, cashew wine and palm oil. A varying but significant proportion of the local households are involved in this petty commodity production at a small scale. In addition to the local market concentrated at the praça (the district centres main square, which also serves as physical marketplace and where all state services are located: school, clinic, party branch office, staff houses etc.), these products sometimes find their way to the regions central market held at fivedays intervals in Canchungo, to the national capital of Bissau, and via, the smuggling circuit, to Senegal (especially palm oil). Allweather dirt roads ensure the communication between Canchungo on the one hand and Calequisse and Caió on the other; between the latter district centres, which are only 15 km apart, the only direct connection is by canoe. Excellent tar roads connect Canchungo with Cacheu, Bissau and the rest of the country, the eastern part of which has a much more developed economic circulation in the hands of Muslim, primarily Fula, traders. Transport at the peninsula is provided by Manjak and Fulaowned pickup trucks, which (except in the frequent times of national petrol shortage) run regular services across the country, and of which at least one comes to either district centre every day.
In this way a significant volume of commodities is brought into the Manjak area: canna (rum, produced mainly in state distilleries in the capital and further to the east); clothing, utensils, furniture, building materials, medicaments, and some rice and preserved food stuffs, from Bissau, the east, or Senegal; pigs, goats and cashew wine primarily from the Balanta-dominated Bula region north of Bissau; and some cattle from the east again. Although my research did not include a quantitative assessment of production and circulation, it seems safe to conclude that much more is being imported into the districts of Calequisse and Caió than is exported; and most of the imports serve a local consumption instead of being an investment in local production.
To some extent, this imbalance might be attributed to intervention on the part of the national state. On the one hand the state has peopled the two district centres with officials whose modest salaries are locally spent, primarily on items of consumption; on the other hand the state operates (via some of these officials: the local Comité dEstado) a system of price control which, especially with regard to such vital commodities as rice, textile and canna, may influence the balance of rural trade (although not, as a rule, in the interest of rural areas). However, the number of resident state officials is very limited (it lies in the range of one hundred for both districts combined); and much of the local flow of trade is effected outside state control. Migrants cash income, realised outside the Manjak rural area but spent inside, is therefore the main explanation for the imbalance.
The Manjak area has long been recognized in the literature as remarkably migrancyprone, as compared to other parts of Guinea Bissau.People from Calaquisse and Caió are found, in capitalist employment or in informal sector relations of production leaning on to capitalism, in all urban centres of Senegal, as well as in France, where they are particularly numerous in the automobile industries around Paris. Today, especially on weekdays outside the planting season (June September), Manjak villages do in many respects convey the impression of a typical labour reserve: a preponderance of the elderly and of young children, a slight underrepresentation of women in childbearing age and a marked absence of youths and adult men. While permanent inhabitants of the village go about their daily productive activities, they contrast strongly with a leisured minority of visiting migrants, conspicuous in their blue jeans, fancy shirts, fine shoes and sun glasses if they are men their elegant Senegalese bubu dresses and turbans from the same bright material if they are women. Surrounded by choice symbols of their migrant status (a wireless set, a stereo cassette recorder, an industrial workers hard hat or a wrist wallet), they recline under the eaves of a house, engaging in conversation with such local relatives as can be spared from domestic or agricultural work. Or, even more typically they are seen performing a ritual at one of the ancestral shrines or deities shrines with which Manjak villages abound, or carrying full bottles of canna to the sacred groves just outside the village, or waiting at the praça for a roundtrip by pickup truck to Canchungo, where they will buy another sacrificial animal and yet more canna for rituals at their home village. On weekends, this small outlandish group is eclipsed by the more numerous locals who, from their jobs or secondary school in Bissau, Canchungo or elsewhere in Guinea Bissau, take every opportunity of visiting their home village. These weekend commuters are on the average younger, their much shabbier attire leaves no doubt as to their residence within the national boundaries, and once back home they are much more readily reintegrated into the social and productive activities of the family. It is not from them, but from the distant migrants to Senegal and France that Manjak villages finance their import imbalance largely used for the purpose of ritual obligations as we shall see.
The production described above is realised in a rural society whose most conspicuous unit is the village; local social organisation further comprises, at levels above the village, initiation regions (each consisting of a handful of neighbouring villages), and the now defunct kingdoms (each consisting of several initiation regions). Internally, each village is segmented into up to a dozen ward, each consisting of compounds occupied by extended families. Manjak ritual organisation largely revolves around shrines distributed over these social units at their successive levels, and therefore both social and ritual organisation will here be described in the same section of my argument.
The main feature of the initiation region is a Sacred Grove: a stretch of meticulously preserved virgin forest access to which is restricted to men who have gone through the initiation rites which are held, for every initiation region, once in about twenty years. In the Sacred Grove the initiation regions central deity is venerated. While that deity has a specific proper name for each initiation region, its essential identity is that of the Land in general. In everyday verbal usage, this deity is equated with the sacred grove where its shrine is located. The central concept is the Manjak religious system, uchaay, commands a complex semantic field comprising, among others, such meanings as God, Land, deity, spirit, devil, sacred grove, forest; in the remainder of this paper, I shall translate the term by Sacred Grove, implying all the nuances spelt out here.
The cult of the Sacred Grove reaches its paroxysm during the two months period of initiation, when all uninitiated young men above the age of seven or eight years go through a training ordeal inside the Sacred Grove under the direction of a number of initiation specialists recruited from among the mature and elderly men of the initiation region. Not only uninitiated actual inhabitants of the initiation region but all uninitiated youths hailing from the villages concerned and presently living in other initiation region, in towns in GuineaBissau, in Senegal and even in France are called upon, and an amazing number still heed the call. Moreover, all initiated men from the initiation region are expected to be present and to make substantial offerings of canna and sacrificial animals during at least some days of this two-months period. The villages then teem with hundreds of returning relatives and visitors, and all resources are drained in order to provide these masses with meals and shelter.
The cult of the Sacred Grove is however far from confined to the time of initiation; instead, it is a continuous, daily concern of all initiated men in the villages of the initiation region. Elderly men convene at the Sacred Grove virtually on a daily basis, in order to pour libations, to sample the palm wine and canna left over once the Land has had its share, to sacrifice animals and consume their meat, and to perform chicken oracles. They converse on social and ritual matters, and in general have a good time together. In these congenial surroundings, sheltered from the gaze of women and boys, the mature men daily engage in a process of social interaction in which honour and power are assessed and redistributed, and claims to office are made, supported or rejected. The high frequency of social and ritual action at the Sacred Grove is guaranteed by the fact that any officering at a lower shrine wherever in the initiation region (i.e. royal shrine, village shrine, ward shrine, a compounds ancestral shrine or an oracle hut)) must invariably be reported at the Sacred Grove, along with a suitable gift of palm wine or canna. The very many rituals which can be seen to be performed in a village from day to day always have an invisible complement at the Sacred Grove. Initiated men visiting a village will always first retire, with their hosts and a suitable libation, to the Sacred Grove, which thus in many ways forms the ritual equivalent of a traditional mens club or public house in Western Europe. When particular important matters are at hand, e.g. rain ritual or the election to a major office within the community, activities at the Sacred Grove also take on a clublike, corporate aspect in that all men concerned are under the obligation to be present there, failing which one has a considerable fine to pay, again in the form of canna or palm wine.
The most important rituals however, involving the most expensive sacrificial animals (fullgrown pigs, heads of cattle) tend to be directed exclusively at the Sacred Grove, without lesser shrines being involved. Women and noninitiated men (e.g. non-Manjak, in the Portuguese era and in the course of my research even Europeans) who are not entitled to enter the Sacred Grove, may appeal to a local elder to sacrifice there on their behalf. These rituals invariably have to do with the discharge of contractual relationships humans enter into with deities, foremost with the Sacred Grove. The form and rationale of these rituals is best described when, below, I discuss Manjak oracles and sacrifices.
While in the older works on the Manjak the kingship is presented as the pivot of sociopolitical organisation (cf. Carreira 1947), the dismantling of this institution since 1973 has been so effective that little more than vestiges of it remain in the sphere of production and land tenure. The kingship still has specific incumbents, and in at least one case a traditional king has managed to attain a formal position of power within the new political and judicial structures controlled by the PAIGC. In the ritual sphere, royal families continue to attend to their royal shrines, which are located not in the Sacred Grove but in a less secluded place adjacent to their compounds. This royal cult however no longer mobilises people from all over the territory of the former kingdom, and such partial control as the kings appeared to have had over the cult of the Sacred Grove in the past, has now disappeared.
Under the luscious beauty of their giant kapok trees and mango trees, Manjak villages stand out as extensive and wellshadowed parklike arrangements, separated from the surrounding forest by a broad circle of paddyfields affording wide view. In addition to this physical delineation, a village is characterised by the following features: it has a village headman; a part (kor ) traditionally set apart to accommodate this official; and a central open place (benii ). At the benii, always marked by some particularly imposing sacred kapok trees, we find the village shrine, likewise called benii: a thatched hutlike construction without walls; in its centre we see a small miniature palissade within which the shrines spirit (uchaay ) is said to dwell although at other times this spirit is said to house in the surrounding kapok trees. Earth from within this palissade is the main substance used in amulets worn by the villagers.
Although there do not appear to exist corporate rituals focussing on the benii (instead of the Sacred Grove), the benii is the scene of important rituals staged for individual villagers (in times of illness, or when twins come of age). Also is the benii the scene of the villages major burial ground. Here the highly respected members of the guild of gravediggers perform their duties; administering the final ritual interaction between a human being and the Land (notably: burial), they are best seen as a prominent type of landpriests. After a burial, inquests are also held at the benii. The empty bier, carried on the heads of bearers supposed to be in trance, is then used as a divination instrument answering questions concerning the cause of death, possible sorcery connotations, and the distribution of the inheritance.
Since major ritual, judicial and localpolitical functions are discharged at the level of the mens assembly in the Sacred Grove (catering for several neighbouring villages), the social significance of the village as a social unit today appears to be less central than its conspicuous delineation in the landscape would suggest. The village headman used to be prominent in the old royal hierarchy before it was dissolved. Even today this officebearer is the guardian of the Cassara shrine (see earlier note) that, covered with cloth of a colour peculiar to each village, features in annual ritual competitions between villages belonging to one ancient kingdom. The allocation of land, a privilege of the king as the greatest landowner, and the organisation of royal tribute labour, used to take place primarily at the village level; and still today, after the dissolution of the kingdoms, a villages fields lie next to one another, quite distinct from the fields of neighbouring villages with whose inhabitants the village in question may entertain very close social, marital and ritual ties. All this suggests that as a unit of social organisation the village lost most of its functions with the breakdown of the ancient political system, which in terms of control over land and the appropriation of surplus labour could well be considered a tributary mode of production.
With the abolition of tribute labour, the virtual dissolution of the ancient royal hierarchy and the appropriation of ancient royal land by individual elders, the vital unit in the production system today has become the pekiin or ward. Each ward (cf. figure 1) occupies a contiguous stretch of land within the villages residential area; boundaries between ward are marked by roads, orchards, gardens or fences. The ward is a strictly exogamous unit. The recruitment of its male members has vague patrilineal connotations, but there is no claim of common apical patrilineal ancestors shared by all members of the ward. It is usually possible to point out one agnatic core within a ward, but this does not preclude that other inhabitants generally and publicly known to be matrilateral or even affinal relatives of the members of this agnatic core enjoy full rights of membership of this residential unit. Except for ritual purposes (the veneration of ancestral shrines, and the very restricted rights of pouring libations there) little stress is laid on genealogical knowledge.
Meanwhile, the pekiin is a productive unit in that the cultivation of the main crop, paddy rice, is realised in collective labour by all pekiin members under the forceful direction of the pekiin head who, it must be admitted, usually prides himself in being an untiring cultivator himself. The planting season (JuneSeptember) has a major rallying aspect: not only do all actual residents of the pekiin take part in production, but in addition virtually all migrants normally dwelling elsewhere in GuineaBissau, and a large proportion of those in Senegal, in this period return to the pekiin and (as young men and women) take a lions share in the inconceivably heavy toil in the flooded paddyfields.
Finally, the pekiin is a socioritual unit in that it tends to have its own central place, called benii like the central place at the village level, and physically and functionally hardly distinguishable from the latter: sometimes major rituals are performed at the pekiins benii, and people may also be buried here. The main difference between village benii and ward benii is that the latter turns out to cater for a much smaller group of people. At the ward benii also the slid drums are kept which, collectively owned by the members of that social unit, play an important role in funerary ceremonies, including, several years or decades after someones demise, the public erection (accompanied by a major sacrifice and extensive libations) of an ancestral shrine at one of the compounds in the ward.
Although the fact that the village is segmented into ward is unmistakable at the most superficial inspection of the residential space, the functional distinction between these two levels of social organisation (village and ward) is so blurred that as a researcher one is tempted to regard the village simply as a maxineighbourhood, or the ward as a minivillage. In a continuous process of fission and fusion, waxing and waning, ward would appear to grow into villages and villages to decline into ward, with supposedly a redistribution of social, ritual and political features which however is not yet adequately documented in my data and which may have become just as blurred due to the dissolution of the old kingdom organisation.
In addition to the type of open, thatched, shrine described above for the benii, at the ward level two other types may be found. First there are shrines of land deities located away from the central place and lacking the corporate connotations of the benii shrine; these shrines may look like miniature huts, but they may also have a more rudimentary shape: marked by nothing more than a shrub, a wood log, or a simple shallow hole in the ground suitable for libation. Guardianship of these shrines is owned by individual by virtue of their being pekiin headman. Sometimes such a shrines deity is merely considered the special guardian of the ward in which the shrine is located; others however are venerated far beyond the ward, as benevolent spirits specialising in granting rain or human fertility in exchange for animal sacrifices.
Secondly most ward contain one, and seldom more than one, oracle hut (pubol ), constructed and owned by an individual oracle priest (napene), who usually is not the pekiin headman. These oracle huts and the divination that takes place there are so crucial to the religious system of the Manjak that they deserve a section of their own (see below).
With regard to the other forms of agriculture than paddyrice cultivation, and to petty commodity production as summarised above, Manjak relations of production are regulated not at the level of the pekiin, but at that of the extended families. Each pekiin consists of up to half a dozen of such families, each characterised by their own dwelling compound. The extended family is bound together by coresidence and commensality. It is headed by a male elder, who owns the familys life stock if any, and who administers the familys rice granary. This elder also officiates at the ancestral shrines that (in the form of a collection of armthick wooden sticks planted in the ground) are found in every Manjak compound.
Women make a vital contribution to production (agriculture, petty commodity production, and domestic work in general) at the level of the extended household, and since marriage constitutes the main procedure to gain control over an adult womans labour power, some remarks about the Manjak marriage system are in order at this point.
While the ward is exogamous, there is a marked degree of intravillage endogamy, and most marriages are contracted within the initiation region. Given the extent of migrancy, this statement must be modified so as to include marriage partners not actually dwelling in the rural area, but having their village home in a particular village and initiation region. Continuity in marriage patterns, and the tendency for initiation regions to coincide with matrimonial areas (Meillassoux 1964) within which the biological reproduction of the population largely takes place, is reflected in the practice of daughters marrying into their mothers village, thus, as the Manjak say, returning the gourd.
Marital payments are slight, often not exceeding a few liters of liquor presented by the soninlaw to his wifes father or guardian. After a transitional period (up to a few years) in which the wife stays in her fathers ward where her labour (and part of that of her visiting husband) is controlled by her family, marriage is virilocal; rapidly the wife is incorporated into her familyinlaw, to such an extent that she will stay there until her death, even in times of absence of her husband, and after his death. Her labour power is controlled by the elder of her husbands ward who delegates most of this control to his senior wife. Under conditions of migrancy, husbands aspire to the creation of neolocal nuclear households away from home: in Bissau, but particularly in Senegal and France; polygamy, as is widely practised, enables men to combine urban and rural marital interests and aspirations.
Oracle huts are located at some distance from their owners compounds, often set apart from the latter by a fence. They are very different from pekiin shrines: they have thick clay walls and a narrow entrance, and in their dark main compartment (the other, smaller compartment being reserved for the invisible oracle deity) easily up to six people may be seated. Here also the altar is found, where surrounded by a collection of shells, antilope horns, gourds etc. a libation basin can be seen, retaining a semifluid sediment of earlier libations of palm wine, canna and blood of sacrifical animals.
The divinerpriest (napene ) caters for individual clients from anywhere except his own ward. There are considerable differences between napenes. Some only act as diviners revealing the causes of a clients misfortune and stipulating necessary ritual action (sacrifice at the Sacred Grove, the erection of an ancestral shrine, etc.), without themselves engaging in somatic treatment. Others combine divination and treatment, thereby laying a personal claim on the clients material resources in excess of the chicken and the bottle of canna that are the inevitable expenses of divination. Although in theory the divinerpriest should spend all revenue from treatment (often a considerable sum per case) on sacrifices and libations for the benefit of the oracular spirit, in practice much of this money is invested on such secular items as a corrugatediron roof, clothing and electronic consumer goods. Methods of divination also vary: while divination always includes a chicken oracle (the inspection of a fowls entrails), some napenes combine this with direct pronouncements allegedly uttered by the spirit in some ventriloquial spirit language unintelligible to ordinary human beings. Some divinerpriests require the client to come to the oracle with already a clear assessment of his own predicament, while others do not want the client to give any personal information on the case, and instead base their diagnosis and direction solely on divination and revelation. But whatever the specific forms (whose description and analysis falls outside the present argument), the napenes art occupies an absolutely central function in the religious system of the Manjak. In order to explain why this should be so, more should be said about the Manjaks view of ritual obligations.
With the exception of simple greeting rituals which travellers and migrants perform at the local shrines, most rituals are a response to specific misfortune, and form part of the following chain of interpretation. First the problem at hand (drought, epidemic, infertility, illness) is to be interpreted as the manifestation of a specific deity or ancestor seeking ritual attention. If an ancestor is thought to be involved, it is usually one who has not yet been honoured by the erection of an ancestral shrine a costly affair which is always postponed until the ancestor shows his impatience through the sending of misfortune. If a deity is thought to be involved, the misfortune is most often attributed to that deitys impatience to see one living up to the terms of a contract one (or ones forbears) has entered into with that deity: in the past one has asked health, fortune, offspring, a nice job in Senegal or France, from the deity, in exchange for the promise of a major sacrifice; and even if the deity can be said to have granted the request, the promised sacrifice is never made before several years have gone by and before a specific case of grave misfortune or ill health has convinced one that the deity is getting impatient.
Whatever somatic treatment a napene may offer (and such treatment implies that he induces his client to enter into a sacrificial relationship with his own oracular deity in addition to any other deities with which the client may already be engaged in contractual relationships), the napenes first task is invariably to provide an answer to the question as to which specific deity or ancestor causes the clients specific misfortune, and to stipulate how this invisible being can be placated. Since every human being is entangled in a close web of ritual obligations vis-à-vis several members of the preceding generation, and in any number of (partly long forgotten) contracts with deities, such a diagnosis is no easy matter; but at any rate it can safely be assumed that no human being is ever completely innocent of ritual neglect.
Napene cater for misfortune that is considered to be the personal affliction of one individual. The secluded intimacy of the oracle hut allows for private conversations where, in addition to the client, only the latters spouse and/or very close kinsmen are allowed to be present. Napene consultation poignantly reflects Manjak notions of privacy and secrecy (which at times may drive a researcher to his wits end).
However, in some cases (covering only a definite minority of all sacrificial events going on in a village or family), one can dispense with the napenes services. Prominent elders (particularly the officiants of the cult of the Sacred Grove, the priests of lesser land spirits that have their shrine in some ward, and ward headmen in general) are sufficiently competent and confident in ritual matters to stage rituals without first consulting a divinerpriest. This pattern applies in the following cases: collective instead of individual misfortune (drought, epidemics); minor sacrificial contracts and/or minor cases of ritual neglect typically involving junior members of the village; and more serious cases of individual misfortune involving an elder himself. In these cases, the bypassing of the napene does not mean that no divination is carried out. Rather, the elders stick to the minimal divinatory requirements that attend all sacrifices (including those stipulated by napenes): a chicken oracle has to assess whether the deitys or ancestors general feeling in the matter at hand is positive (white, in the Manjak oracular symbolism the colour black signifies the alternative case); after which the urine oracle has to show whether the deity or ancestor, having already expressed his overall agreement, subsequently accepts or rejects the specific sacrificial animal selected for him (in the case of acceptance, the animal urinates immediately before one proceeds to killing it).
These minimal divinatory requirements may indicate the uncertainty, powerlessness and tediousness that pervade relations between Man and the supernatural among the Manjak. The supernatural is difficult to approach. While many sacrifices are rejected (as shown by the animals failure to urinate), some sacrifices may be, at best, just tolerated by the supernatural, but they never ingratiate Man with the supernatural there is always the danger of falling short to unknown and demanding expectations on the part of some forgotten deity in distant parts. Ritual among the Manjak is a thoroughly joyless, miserable duty, in which one never reaches a state of blissful accomplishment. Although humans engage in contracts with deities as if they were equals, these relations ultimately convey a sense of onesided dependence on some whimsical and tyrannical power a striking reflection of the model that underlies Manjak gerontocratic relations between youth and women on the one hand, and elders on the other. Therefore, for Manjak women and youth ritual contains a double bind: it does not release them from the clutches of everyday life (in which they are dependent upon and exploited by gerontocratic elders), but rather reinforces their predicament, first because no ritual can be completed without an elder officiating in it, secondly because the relations with the supernatural can be said to be an ideological reproduction of gerontocratic arrogance. This is the reason why I do not consider Manjak rituals to have much therapeutic value for others than the elders themselves.
There are historical indications  that when elders stage rituals without consulting a napene, we have to do not with elders usurping the napenes professional prerogatives, but with elders insisting on their historical ritual competence in the face of a recent expansion of the napenes competence and prerogatives. Of old, the napene formed part of the kingdom hierarchy; as members of a hereditary guild, their activities were to a considerable extent controlled by and subservient to the politicoritual powers of the king. Elders at the village and ward level were also part of the same hierarchical differentiation of functions. With the dissolution of royal power the hierarchical structure collapsed (no doubt under the additional influence of individualising tendencies brought by ever increasing capitalist encroachment through labour migration and the cash economy), and the napene more than ever before took on the characteristics of divinatory and therapeutic entrepreneurs, catering for misfortune that was more than ever before conceived as a strictly individual matter. In this they were less than before checked by the cult of the Sacred Grove (which was no longer associated with royal power), while also the social control exercised by the napene guild organisation was slackened.
Having thus summarised contemporary social, economic and ritual structures among the Manjak, we may now ask ourselves what is the place in this village society, of migrants who have their places of work in Senegal or France. This discussion will highlight migrants rituals as a major form in which the articulation of capitalism to preexisting modes of production presents itself in this West African periphery in a way that reproduces not so much capitalism, but the local modes.
Manjak rural society today can still serve as a textbook example of a viable gerontocracy. The codes of gerontocratic power continue to be respected not only by young men residing in the village but also by those living as labour migrants at distant places of work under relations of production and under social conditions very different from those prevailing in the village. An amazingly large proportion of these migrants keep up contacts with home. They send remittances, clothes, building materials and electronic consumer goods, try to attend the local initiation festival as held once every twenty years, and also in other years visit their elders, bringing canna and more endurable gifts. This vitality of the Manjak gerontocratic system is puzzling. Marriage payments among the Manjak are too insignificant to form the basis of the elders power over young men, as they do in many other African societies. (cf. Rey 1971; Geschiere, in press). One could further invoke in this connection a number of socioeconomic explanations which have been advanced for other African cases involving a high rate of migrancy: the elders control the youths access to land, and the latter cannot risk rural ostracism given the insecurity of their urban footholds as migrants; the migrants wives and children are left in the care of the village elders, in a subtle captivity ensuring the migrants continued respect and financial remittances, etc. In the Manjak case explanations of this nature appear to lack conviction. Many Manjak migrants have acquired Senegalese or even French citizenship, and have thus a rather secure foothold abroad. Many are less than committed to the independent state of GuineaBissau, its disaster economy, and its ruling party the PAIGC, and would not dream of retiring in their home area although many ultimate do retire there. Many consider village life an ordeal that one can only endure for a few weeks a year, if that; and although many have left (some of) their wives and children at home, others have their dependents safely outside the elders control, in relatively comfortable houses in Senegal and France (cf. Diop 1981).
It is not simple economic necessity that drives Manjak men back into the arms of their elders. The initiation, as youths, into the cult of the Sacred Grove, their more gradual and less dramatic exposure to the cults of lesser deities and ancestors, and the ensuing socialisation into notions of obligation, neglect, dependence and fear, may have much to do with the migrants continued observance of rural ties that are not an obvious asset to them. However, I would shrink from invoking such an ideological factor as an independent variable, and would rather admit that my exploratory research among the Manjak does not yet allow me to provide a full analysis on this point. Social control mutually exerted by Manjak migrants at their distant places of work, sometimes taking a formal organised form, may provide part of the explanation, particularly as regards those migrants who keep up such intensive ties with their village that every year they participate in the planting season.
Many migrants in Senegal, and all migrants in France, do not retain productive rural ties of this nature, and their home visits tend to be at intervals that are much longer than one year. What is very striking in those cases is that, when a home visit finally occurs, there is invariably a major ritual obligation that in the migrants mind forms the most obvious reason for the arduous and expensive trip. Migrancy is of course a condition conductive to all sorts of somatic and mental trouble . When (often after vain appeals to Western doctors, psychiatrists, social workers etc.) these complaints are put before some Manjak diviner residing at the distant place of work, they are interpreted as spiritual or ancestral manifestations due to ritual neglect or obligations. Sometimes money sent home for the purchase of canna and sacrificial animals may be considered a sufficient remedy, but in most cases the migrant sees no alternative but to return home personally. He will bring gifts for his living rural kinsmen, but his main expenditure apart from his return ticket will be ritual: literally dozens of liters of canna, several pigs and/or goats, a considerable number of chickens to be used for oracles. Most of these items are only available at Canchungo if at all. One does not buy them all in one go, but item after item, making the expensive and tedious taxi ride to Canchungo time and again, in the course of weeks that may easily become months, as ones ritual obligations find ever new and unexpected extensions through the divinations at the oracle huts and the Sacred Grove of the home area.
The expenses incurred would be truly astronomical if the migrant were to abide by the official exchange rate of the GuineaBissau currency. Many migrants however manage to change their French or Senegalese currency at the black market (the border town of São Domingo, where longdistance taxis enter from Senegal and deliver the migrants at the pedestrians ferry to Cacheu, is famous in this respect), thus reducing their costs very considerably. But even so, the migrants home visit is traumatic. He is constantly aware of being at the mercy of oracle priests, elders, taxidrivers, Canchungo traders, and the Guinean economy as a whole a painful contrast with the relative comforts of his distant place of work. As his ritual obligations turn out to ramify in unexpected directions, his time budget and finances begin to give out. Nor do the many rituals in which the migrant is involved, in any symbolic and psychological way seem to create a marked catharsis, some redeeming reimmersion in the culture and society he was born in. On the contrary, the prolonged dependence, for the fulfillment of ritual obligations, on both human and supernatural authority figures against whose whims and directions not the slightest appeal is possible, in addition to rural health conditions and the effects of ritual overconsumption of alcohol, create a state of stress from which the migrant can only recover after having left his village home again. He is discharging a painful and costly duty which has little intrinsic gratification to offer. The fact that his coming home has primarily been defined in terms of ritual obligation and neglect, spoils what might have been a rural vacation into a race for spiritual and financial survival. The migrants gain seems to be not so much that he is confirmed as a member of the rural society and culture, but that he earns the right to leave again and to stay away for some years at least...
It is no longer necessary to argue that labour migration is a particularly effective form in which the articulation between capitalism and noncapitalist modes of production is brought about. It is more opportune here (concluding a descriptive argument that started out with a theoretical statement on religion as ideological reproduction and production) to ask ourselves how, precisely, and primarily at which point do migrants rituals in Manjak society today affect the structure of articulation of modes of production and from there to proceed to an assessment of the limitations of the answers the preceding argument will suggest.
My summary of rural production and economic circulation in contemporary rural Manjak society had a purpose beyond ethnographic redundancy: it enables us to distinguish, albeit tentatively, the articulated modes of production that make up this social formation. Obviously the encroaching capitalist mode of production, whose local protagonists are the migrants, does not confront one monolithic noncapitalist Manjak mode of production, but a complex of articulated structures, in which at least two constituent modes of production are immediately manifest: one more properly domestic mode, revolving on the central exploitation of youth and women by elders; and another one, revolving on the exploitation of producers engaged in a domestic mode of production, by royal courts. The latter type of exploitation, amounting to a tributary mode of production, used to be effected through the appropriation of both land and surplus labour.
Both modes of production have their structures of ideological reproduction peculiar to that mode: ancestral shrines underpinning the elders authority in the case of the domestic mode; and royal shrines underpinning royal legitimacy and exalted power over the fertility of the land in the case of the tributary mode. But this does not exhaust the complexity of Manjak religious structures the cult of the Sacred Grove, the cult of the benii shrines, of other lesser deities at the ward level, of the oracular spirits associated with the pubols, and finally the cult of Cassara, are not easily fitted into such a suspiciously attractive picture of onetoone correspondence between production unit and ritual expression.
Underlying these several cult complexes, two major types of idiom can be detected: a land idiom and a sky idiom. The sky idiom is only represented by the cult of Cassara; its features are the motton of a High God, emphatic moral concern (a preoccupation with sorcery), somewhat colour symbolism, movable shrines, an annual calendar, and association with the village level which as have argued is more a social and political than a productive unit. The land idiom is represented in the other cults mentioned; its features are a pantheon of leccer deities localized in the landscape, virtual absence of moral concerns (deities are allegedly prepared to enter into any sort of contract, harmful or not to humans, with or without commotations of sorcery  and when they punish they do not do so for moral indignation but for wounded pride), restricted colour symbolism (the only significant colours are balck and white in which divination is encoded; further the greyish red colour of clay pervades everything connected with these cults), immovability, a twentyyears cycle (in the case of the Sacred Grove) or no conspicuous cycle at all (in the case of the other shrine cults), and rather than the specific association with one specific level of sociopolitical organisation, a hierarchical structure encompassing all levels. The latter feature is further brought out by the fact that all rituals at lesser shrines are echoed by ritual at the Sacred Grove. Manjak believe, moreover, that the lesser deities themselves report all rituals directed at them, to the Sacred Grove.
The difference between these two idioms is so striking that one might be surprised to find them in one and the same culture. We have here a clear example of two ideological logics that are mutually so irreducable that one would be tempted to connect each with a different logic of production and exploitation within a social formation composed of several articulated modes of production.
It requires no great effort of imagination to identify paddyrice cultivation as the counterpart, in the sphere of material production, of the land idiom. The objection that we have already identified a domestic mode of production underpinned not by the Sacred Grove and lesser deities but by ancestors, is easily resolved when we call to mind the striking distinction, in the regulation of agricultural production, between paddyrice cultivation (nowadays organised at the ward level), and other forms of agricultural production (organised at the compound/nuclearfamily level). This distinction will appear to be even more relevant once we realise that irrigated rice cultivation (although these days organised at the ward level) requires the concerted efforts of a much larger community even than the ward, not so much for the annual preparation, planting and harvesting at the individual paddyfields, but particularly for the maintenance of the complex irrigation system. These concerns go far beyond the very small group that is bound by common ancestors, and must ultimately be attended to at even a higher level of social organisation than the village: the initiation level, normally composed of villages whose paddyfields are located in the same valley or at the same rivulet. There really seems to be a case for the identification, in the Manjak social formation, of a third mode of production, like the properly domestic one based on the exploitation of women and youth by elders, but differentiated from the latter in that the classes involved in the rice mode of production are not primarily defined by kinship and domestic roles as discharged in closelyknit extended families, but by membership of broad age groups with specific tasks in the ricegrowing process. The elegance of this argument is further enhanced by the fact that, on the ideological level, the paroxysm of the cult of the Sacred Grove is initiation, which regulates the relationships between age groups.
I do not think that we have to construct, within the local economy, again a fourth mode of production for which the sky idiom of Cassara could then be argued to serve as a structure of ideological reproduction. The Upper Guinea coast has been exposed to intercontinental circulation since the fifteenth century. During the sixteenth century the coastal area was a major source of slaves for the transatlantic trade (cf. Rodney 1970; Curtin et al. 1978: 231f). Although much more specific data are needed, I would suggest that the Cassara complex should be taken, not as a reflection of a separate mode of production, but as an ideological expression of the articulation of the local social formation to an encroaching mercantile capitalism.
Migrants, in this picture, represent the articulation of the emerging social formation with a later development of the capitalist mode of production: industrial capitalism, revolving on the exploitation of labour by capital. Born and bred in the Manjak villages but selling their labour power at a distant capitalist labour market, they perfectly fit in the picture of peripheral capitalism: their labour is overexploited (Meillassoux 1975), first because its surplus value is appropriated by capital, and secondly because their own domestic community does not enjoy the full interest on the investment it has made in the biological reproduction of these workers since they were conceived and born.
Through autochthonous rituals, these migrant workers are brought to spend a large part of their capitalist wages in their home communities. At first sight, it would look as if thus the overexploitation of their labour is reduced: after all, in this way the domestic community reaps some of the fruits of its investment in these migrants unproductive years as infants and boys. One could even attempt to estimate the value of pigs and canna involved, but unfortunately the migrants exchange fiddles make this a spurious exercise.
Such a financial cost/benefit analysis would, however, distract us from the crucial question: what, in the sphere of material production and the attending relations of production, is really being reproduced by the ritual structures in which the migrants are so active?
Part of the canna and most of the sacrificial animals blood runs away in the ground, and the rest of drink and meat is consumed by the elders, with some limited share for other villagers if the ritual happens to take place not in the Sacred Grove but inside the village. Biological or physiological reproduction of the labour power of the elders and possibly other villagers hardly seem to be the point here although the nutritional value of meat and canna cannot be denied. What these rituals reproduce, to an excessive extent, is a relation of gerontocratic exploitation. They make the elders and their prerogatives eminently visible, both in a direct form (elders officiate at the rituals and consume drink and meat), and in a symbolic form (yielding to the demands of deities and ancestors means yielding to powers that closely resemble living elders). Giving in to the demands of deities and ancestors, the migrants in their rituals in fact ideologically reproduce the two local modes of production that, after the dismantling of the tributary mode, are still viable: the domestic mode and the rice mode. Instead of being the local propagators of capitalism in any direct sense, or even becoming the agents of modernisation and liberation, the migrants apparently have no choice but to be the overzealous servants of the ideological structures (including such concepts as deities and ancestors, and such attitudes as fear and insecurity) upon which the ancient modes of production are partly based. It is as if the migrants relative and temporal immunity from gerontocratic control during their stay at distant places of work, has to be bought by ostentatious symbolic submission to this control during the short time they spend at home.
Peripheral capitalism in itself is not being manifestly reproduced in this context or it should be that the migrants potlatchlike (cf. Diop 1981) ritual displays of wealth and ritual zeal induces other potential migrants to actually depart for Senegal and France, or induces elders to grant their permission for such departure more readily. This however is again a wrong interpretation. The migrants are not welcomed home as people who have made the grade abroad, but as pitiable patients who come to seek ritual redress, and as negligent observers of ritual obligations who come to make up for their shortcomings. There is no triumph whatsoever in the migrants excessive ritual action. Moreover, a ritual display of wealth is scarcely necessary: the material display in the form of expensive outlandish clothes, stereos, pictures showing welldressed people in a wellfurnished home abroad, does already enough to create incentives for potential migrants and their elders.
A further look at the napenes role in the migrants ritual activities might suggest that such an interpretation in terms of ideological reproduction of ancient but still vital modes of production might yet have its limitations:
The napenes art today is not totally subservient to the upkeeping of the ritual structures of the land idiom and the ancestral idiom. In addition to ushering people into chains of ever more expensive rituals directed at the Sacred Grove, the benii, the ancestral shrines etc., the divinerpriests try very hard to make the client enter in a specific, expensive relationship with his own oracular spirit, with whom the client usually had no previous relationship or contract. In this way, the oracular spirit becomes not the servant and messenger, but the business rival of the Sacred Grove and the other, lesser spirits. In recent decades, divinerpriests fees seem to have increased, their private secular investment of these fees has become common practice, and a number of divinerpriests are alleged to have begun to experiment with lucrative types of somatic treatment for which they do not have the proper traditional training nor the solemn initiation, and which is no longer effectively controlled by the napenes guild. In the hands of the napene, Manjak ritual structures appear to have taken on the characteristics of an exploitative structure sui generis. Of course this structure can only thrive at the fringe of the more general religious notions and actions that make up the various cults, but it is no longer wholly dependent upon the latter but has taken on a dynamic of its own. Migrants caught in this structure are not just engaged in the ideological reproduction of ancient modes of production they are also, in part, directly exploited by (and thus can be said to reproduce) specialists ritual structures that no longer bear a particularly close relationship with the ideological dimensions of Manjak modes of production.
Finally, in order to do justice to the serious therapeutic concerns of some of the napenes with whom my research has brought me into contact, I should like to stress that this divinatory racket variant is not the only possible limitation that the notion of ideological reproduction encounters in the Manjak case. As a client, even more than as a researcher, I have seen certain napenes creatively manipulate the symbolic and dramaturgical material that is present in the Manjak religious system today, with such virtuosity and profound human concern that, rather than confirming their clients in some form of exploitation (by elders, napenes, distant capitalists, or some articulated combination of them all), they bring about genuine revelation and liberation. A Marxist approach ultimately aimed at the liberation of consciousness, should be prepared to acknowledge such similar potential in other, authentically African, forms of ideological production.
Amin, S. (ed. ) (1974), Modern migrations in IVest Africa, London: Oxford University Press.
Amselle, J.L. (ed.) (1976), Les migrations africaines, Paris: Maspero.
Auge, Al. (1975), Theorie des pouvoirs et ideologie, Paris: Hermann.
Banton, M. (ed.) (1966),Anthropological Approaches to the Study of Religion, London: Tavistock.
Bare, J.F. (1977), Pouvoir des vivants, language des morts, Paris: Maspero.
Bergues, El. (1973), Limmigration de travailleurs africains noirs en France et particulierement dans la region parisienne, Population, (Paris), 28, 1: 59-79.
van Binsbergen, W.M.J. (1981), Religious change in Zambia, London/ Boston: Kegan Paul International.
van Binsbergen, W.M.J. (1983a), Religie en etno-psychiatrie in Guinee-Bissau; Voorstel voor een explorerend onderzoek, Leiden: African Studies Centre.
van Binsbergen, W.M.J. (1983b), Project religie en etno-psychiatrie in Guinee-Bissau: Tussentijds rapport, Leiden: African Studies Centre.
van Binsbergen, W.M.J. (1984), Zusters, dochters: Afrikaanse verhalen, Elaarlem: In de Knipscheer.
van Binsbergen, W.M.J. (in press), Can anthropology become the theory of peripheral class struggle?, in va,n Binsbergen & Hesseling (in press).
van Binsbergen, W. M. J. & P. L. Geschiere 1984a, Marxist theory and anthropological practice, in van Binsbergen & Geschiere 1984b, pp. 235-89.
van Binsbergen, W.M.J., & P.L. Geschiere (eds) 1984b, Old modes of production and capitalist encroachment, London/ Boston: Kegan Paul International.
van Binsbergen, W.M.J. & Hesseling, G.S.C.M. (eds) (in press), Aspecten van staat en maatschappij in Afrika, Leiden: African Studies Centre.
van Binsbergen, W. M. J., in collaboration with J. de Jong (n. d. ) Divination and mental health in northwest Guinea-Bissau: An exploratory study, Leiden: African Studies Centre, forthcoming.
van Binsbergen, W.M.J. & Meilink, EI.A. (eds) (1978a), Migration and the transformation of modern African society, Leiden: African Studies Centre.
van Binsbergen, W.M.J., & Meilink, El.A. (1978b), Migration and the transformation of modern African society: Introduction, in van Binsbergen & Meilink (1978a): 7-20.
van Binsbergen, W.M.J., & Schoffeleers, J.M. (eds) (1984a), Theoretical explorations in African religion, London/ Boston: Kegan Paul International.
van Binsbergen, W.M.J., & J.M. Schoffeleers (1984b), Theoretical explorations in African religion: Introduction, in van Binsbergen & Schoffeleers (1984a): 1-49.
Bonte, P. (1975), Cattle of God, Social Compass 22: 381-400.
Brito, E. (1952), A poligamia e a natalidade entre os grupos etnicos Manjaco, Balante et Brame, Boletim Cultural da Guine Portuguesa (Bissay) 7 ,
Buijtenhuijs, R. (1984), Dini ya Msambwa: Rural rebellion or counter-society?, in van Binsbergen & Schoffeleers (1984a): 322-45.
Carreira, A.A.P. (1946), Alguns aspectos do regime juridico da propriedade imobiliaria dos Manjacos, Boletim Cultural da Guine Portuguesa (B
Carreira, A.A.P. (1947a), Vida social dos Manjacos, Bissau: Centro de Estudos da Guiné Portuguesa, Bissau.
Carreira, A.A.P. (1947b), Ceu, Deus e a Terra (lenda de Manjacos), Boletim Cultural da Guine Portuguesa (Bissau) 2,6:461-3.
Carreira, A. A. P. (1948), A habitaçao dos Manjacos de Costa de Baixo, in Mota & Neves 1948.
Carreira, A.A.P. (1953), O levirato no grupo etnico manjaco, Boletim Cultural da 29: 107-11.
Carreira, A.A.P. (1956), Movimento natural da população não civilizada da circumscrição administrativo de Cacheu, Boletim Cultural da Guiné Portuguesa, Bissau, 9: 7-90.
Carreira, A.A.P. (1960), Regiao dos Nlanjacos e dos Brames: Alguns aspectos da sue economia, Boletim Cultural da Guine Portuguesa (Bissau) 15.
Carreira, A.A.P. (1961a), Mutilaçoes etnicas dos Manjacos, Boletim Cultural da Guine Portuguesa (Bissau) 16, 61: 83-102.
Carreira, A.A.P. (1961b), Simbolos, ritualistas a ritualismos animo-feiticistas na Guine Portuguesa, Boletim Cultural da Guine Portuguesa (
Carreira, A.A.P. (1967), Manjacos-Brames et Balantas: Aspectos demograficas, Boletim Cultural da Guine Portuguesa (Bissau) 22 , 85-6: 41-92.
Carreira, A.A.P., & Marques, 1947
Collomb, H., & B. Diop (1969), Migration urbaine et sante mentale, Waltham (Mass. ): African Studies Association.
Correia, C.B. (1958), A ceramica na vida social dos Balantas et Manjacos, Boletim Cultural da Guine Portuguesa (Bissau) 13, 50: 133-48.
Coulon, C. (1984), Prophets of God or of history, iE van Binsbergen & Schoffeleers (1984a):346-66.
Crespo, M. Pereira (1955), O rio Cacheu, Boletim Cultural da Guine Portuguesa (E
Curtin, P., Feierman, S., Thompson, L., & Vansina, J. (1981), African History, London: Longman, 2nd edit.
Devisch, R. (1984), Perspectives on divination in contemporary sub-Saharan Africa, in van Binsbergen & Schoffeleers (1984a): 50-83.
Diallo, J.-C. (1964), Le probleme des refugies de la Guinee-Bissaü en Casamance, Senegal dAujourdhui 12: 11-17.
Diarra, S. (1966), Problemes dadaptation de travailleurs africains noirs en France, Psychopathologie africaine, 2, 1: 107-26.
Diarra, S. (1968), Les travailleurs africains noirs en France, Bulletin de lInstitut fondamental dAfrique noire, serie B, XXX, 3: 884-1004.
Diop, A.M., (1981), Tradition et adaptation dans un research de migration senegalaise: La communaute manjak en France, these 3eme cycle, Ecole des Hautes Etudes en Sciences Sociales, Paris.
Diop, B. (1968), Sur la transplantation negro-africaine en France, Psychopathologie africaine, 4, 2: 227-76
Fabian, J. (1984), Religious pluralism, in van Binsbergen & Schoffeleers 1984a: 138-63.
Fortes, M. (1959), Oedipus and Job in West African religion, Cambridge: Cambridge University Press.
Geertz, C. (1966), Religion as a cultural system, in Banton (1966): 1-46.
Gerold-Scheepers, T.J.F.A., & van Binsbergen, W.M.J. (1978), Marxist and non-Marxist approaches to Migration in tropical Africa, in van Binsbergen & Meilink (1978a): 21-35.
Geschiere, P. (in press a), Bruidsprijzen en kapitalistische expansie bij de Maka, Te Elfder Ure, 33.
Girard, J. (1969), Genese du pouvour charismatique en Basse Casamance (Senegal, Dakar: Institut Fondamental de lAfrique Noire (IFAN)
Houtart, F. ( 1980 ), Religion et modes de production precapitalistes, Brussels: Edition de lUniversite de Bruxelles.
de Jong, J. (1983), Proposition dun programme de la sante mentale dans le cadre du plan quadriennal 1983-1987, Bissau: Ministry of Health and Social Affairs.
Jong, J. de, & R. Buijtenhuijs (1979), Guine-Bissau: Een bevrijdingsbeweging aan de macht, Wageningen: de Uytbuyt.#
de Jonge, K., J. van der Klei, H. Meilink & R. Storm, Les migrations en Basse Casamance (Senegal), Leiden: African Studies Centre.
van der Klei, J. M. (1984), Articulation of modes of production and the beginning of labour migration among the Diola of Senegal, in van Binsbergen & Geschiere (1984b): 71-93.
Lopes, E.A. Correia (1943), Manjacos: III - Vida material Mundo Portugues, 10.
Lopes, E. A. Correia (1945), Observaçoes sobre os clans no paper Manjaco, Mundo Portugues 12:139.
Marx, K . (1973), Capital, New York: Monthly Review Press .
Marx, K., & Engels, F. (1975), On Religion, Moscow: Progress Publishers .
Mbenque, RI.M. (1971), Lintegration des refugies dans le departement de Ziguinchor, Mernoire de stage, annee scolaire 1970-71, Dakar: Ecole Nationale dAdministration (Dakar).
Meireles, A . Martins ( 1948 ), A habitaçao dos Manjacos de Caio, in: Mota & Neves 1948.
Meireles, A. Martins (1949), Baiu (Gentes de Kaiu), II. Kabaxe Boletim Cultural da Guine Portuguesa (Bissau) 4, 13: 7-25.
Meireles, A. Martins (1952), Instituiçoes de direito penal consuetudinario dos Manjacos de Caio, Confer. Inter. dos Africanistas Ocidentais, 2a. Confer., Bissau, 1947, Lisboa, 1952, vol. 5, pp. 391-430.
Meireles, A. Martins (1960), Mutilaçoes etnicas dos Manjacos, Bissau Centro de Estudos da Guine Portuguesa (Bissau), publ. 22 .
Meillassoux, C. (1964), Anthropologie economique des Gouro, Paris/ The Hague: Mouton.
Meillassoux, C. (1975), Femmes, greniers et capitaux, Paris: Maspero.
Mendy, F. (1969), La contrebande en Casamance: Son importance, ses causes, les remedes, memoire de stage, Dakar: Centre de Formation et Perfectionnernent administratifs.
Minces, J. (1973), Les travailleurs etrangers en France, Paris: Seuil.
Mitchell, R.C., & Turner, EI.W., (1966), A comprehensive bibliography of modern African religious movements, Evanston: Northwestern
Mota, A. Teixeira da (1954), Guine Portuguesa, 2 vol., Lisboa: Agencia geral do Ultramar.
Mota, A. Teixeira da, & M. Ventim Neves (eds) (1948), A habitaçao na Guine Portuguesa, Bissau: Centro de Estudos da Guiné Portuguesa, Memorias do Centro de Estudos da Guiné Portuguesa no. 7.
Murdock, G.P. (1965), Social structure, New York/ London: Free Press/ Collier-Macmillan, repr.
Nogueira, A. (1947), O irã na circunscripçao de S. Domingos, Boletim Cultural da Guine Portuguesa (Bissau) 2, 7: 711-6.
Papyle, H . ( 1973 ), Les travailleurs etrangers en France: Essai dune bibliographic en langue française, numero special, Hommes et Migrations, 120
Pires, A. Rodrigues (1948), A habitaçao dos Manjacos de Pecixe, in: Mota & Neves 1948.
Poelhekke, F.G.M.N. (1979-80), Guinee-Bissau, The Hague: Staatsuitgeverij .
Reis, J. Estevao dos (1947), Manjacos de Calequisse (apud analise de T. hlota), Boletim Cultura1 daGuiné Bussay 2, 6:574.
Rey, P.P., (1971), Colonialisme, neo-colonialisme, et transition au capitalisme, Paris: Maspero.
Rey, P.P. (1973), Les alliances de classes, Paris; Maspero.
Rodney, W. (1970), A history of the Upper Guinea Coast, 1545 to 1800, New York/ London: Monthly Review Press .
Santos Reis, C., & F. Courtinho da Costa (1961), A alimentaçao dos Manjacos, Boletim Cultural da Guine Portugesa (Bissau) 16, 63: 377-503.
Schoffeleers, J.M. (ed. ) (1979), Guardians of the Land, Gwelo: Mambo Press.
Sow, A.P.B. (1969), Les refugies dans la region de Casamance memoire de stage, Dakar: Centre de Formation et de Perfectionnement administratifs.
Sundkler, B.G. (1961), Bantu prophets in South Africa, London: Oxford University Press, 2nd edit.
Terray, E. (1969), Le marxisme devant les sociétés primitives, Paris: Maspero.
Werbner, R.P. (ed.) (1977), Regional cults, Mew York: Academic Press.
 This clearly is a Marxist rephrasing of Geertzs famous definition (1966) .
 For Marxist analyses of the religious dimension of pre-capitalist modes of production, cf. Bonte (1975), Houtart (1980), Auge (1975), and Bare (1977).
 A classic statement of correspondence along structural-functionalist lines is Fortes (1959). On the significance and limitations of the correspondence thesis in the anthropology of religion, cf. Werbner 1977; van Binsbergen 1981.
 Cf. Rey 1971, 1973; van Binsbergen & Geschiere, in 1984b; on the ideological problems of such an approach, as well as its potential, cf. van Binsbergen (in press).
 Cf. van Binsbergen 1981; van Binsbergen & Geschiere (in 1984a: 274-8) succinctly propose an ethnographic and historical method (such as used in van Binsbergen 1981) for the analysis of religion in complex social formations.
 Cf. Werbner 1977; Schoffeleers 1979; van Binsbergen 1981.
 From among the massive literature on prominent African religieus innovators, cf. Sundkler (1961), Mitchell & Turner (1966).
 This theoretical caveat, representing a different position from the one I took earlier (cf. van Binsbergen 1981), ties in with such work as Buijtenhuijs (1984) and Coulon (1984).
 My emphasis on such experiments might seem to introduce, idealistically, an autonomous field fo human intellectual activity, independent from the constraints of material production. But on the contrary, Mans ability to experiment and to create provides the very basis for technological and organizatory innovation, and hence for the emergence of new relations of production and new modes of produc- tion; just as everyday material production provides the most obvious context (in the way of models, contexts, challenges) for symbolic and philosophical thought, experiment, and innovation.
 For recent studies in this vein, cf. Fabian (1984), Devisch (1984), and general comments in van Binsbergen & Schoffeleers (1984a).
 After preparatory trips in November 1981 and November 1982, fieldwork was carried out (mainly through the medium of the Creole language, the national lingua franca i.e. a Manjak interpreter was used only in the first few weeks - in the Cacheu region, GuineaBissau, from April to August 1983, at the request of the Ministry of Health and Social Affairs, and within the wider framework of that Ministrys mental health planning policy (cf. de Jong 1983). In addition to his extensive administrative and logistic support, and the contribution of relevant clinical case material not used in the present argument, the psychiatrist J. de Jon g for about two weeks shared in the field-work so as to augment the data with depth interviews of rural psychiatric patients. Further, I wish to register my indebtedness to local authorities in the Cacheu region; to the people of Calequisse; and to the African Studies Centre, Leiden, who funded the project and graciously granted me leave of absence in order to carry it out. For preliminary accounts of the Manjak research, cf. van Binsbergen (1983a; that paper contains a very extensive bibliography of possible use to other researchers) and (1983b). A more extensive report is in preparation (Van Binsbergen & de Jong, n.d.). Given the limited period of field-work the present analysis must be considered as provisional, awaiting further research.
 For an interesting account of Manjak migration as seen from France, cf. Diop (1981). Regrettably, this author has never set foot in the Manjak area of origin; yet his reconstruction, based on data gathered in France, of contemporary Manjak village society is basically sound. A spate of publications provide a background on West African labour migration to France; cf., e. g., Bergues (1973); Diarra (1968); Diop (1968); Minces (1973); Papyle (1973). Van der Klei (1984) is a penetrating analysis of migration and modes of production in a neighbouring area not unlike the Manjaks.
 A 1950 census estimated the number of Manjak in Guinea-Bissau at just over 70,000, or 14.3% of the territorys population of just over 500,000 (Poelhekke 1979-80:21). Allowing for a substantial increase of the national population as well as increased Manjak migration to Senegal and France, the present-day Manjak population in GuineaBissau could be estimated at 90.000-100.000, r,lost of whom live in the Cacheu region. In addition, an increasing number of Manjak have come to form a permanent (as distinct from migrant) population in Senegal (Casamance and Cap Verde regions, including Dakar), especially since the Guinean liberation war. The ethnographic literature on the Manjak is surprisingly abundant and interesting; cf. Brito 1952; Carreira 1946, 1947a, 1947b, 1948, 1953, 1956, 1960, 1961a, 1965, 1967; Correia 1958; Crespo 1955; Diop 1981; Lopes 1943, 1945; Meireles 1948. 1949, 1952, 1960; Mota 1954: 308f; Nogueira 1947; Pires 1948; Reis 1947; Santos Reis & Courtinho da Costa 1961. On the Manjak in Senegal, cf. Diallo 1964; Mbengue 1971; Mendy 1969; Sow 1969.
 My oral historical data, tying in with the exceptional position of the Cobiana central deity of Mama Jombo as described by Carreira (1961b: 529); cf note 23.
 Cf. Carreira (1956; 1960); Diop (1981), and the references cited in that work.
 In Creole, the Manjak word uchaay is usually translated by iran, while some informants, extensively exposed to Roman Catholicism, may prefer the translation shatan. The latter word plainly means devil, while iran refers to any spiritual being that represents a mode of existence independent from humanity (therefore, souls of the departed are not irans): the spirits associated with sacred groves, clan shrines and diviner-priests shrines, but also sorcery familiars. Usually, the concepts uchaay and iran are not supposed to encompass the notion of a EIigh God, called Deus in Creole, and Nasin Batsi (Y.ing of the Sky) in Manjak. This distinction between deity and EIigh God, so well geared to Christianity, may well be due to Christian influence. Likewise, Roman Catholic missionaries have equated Christ with the Cassara figure from the Manjak pantheon. Cassara is the messenger of Nasin Batsi who once a year (in April) visits the world in order to expose, chase and kill the witches. Cassaras material incarnation is a ceremonial bier covered with cloth in a colour peculiar to each village. During the week of Cassara villagers enjoy collective meals at the dancing-ground adjacent to the headmans house. Cassaras bier (with which entranced youths dance and run about on this occasion) is considered to be an oracle capable of detecting and chasing witches. Anyone dying in this period will not be mourned: he or she is considered a witch slain by Cassara.
 Thus, during the initiation ceremonies in the Sacred Grove of BoTimat, near Calequisse, in April/ June 1983, at least fifty youths participated, including a considerable number from Senegal, and even a few boys flown iI1 straight from France where they had been born.
 Palm wine is perishable and only available during part of the year (October-May) . Rituals are performed the year round, and therefore canna is an essential substitute for palm wine, even though the latter, as direct, unadulterated produce tapped straight from the trees, is held to be symbolically superior. For ritual purposes one liter of canna (a standard minimum quantity for most ritual occa- signs) is equated with five liters of palm wine; as a commodity, palm wine is somewhat cheaper than this ritual equation suggests: while a liter of canna sells at peso 180-200, a liter of palm wine only fetches peso 20-25. For the peso exchange rate, cf. note 27.
 The only case of intra-neighbourhood marriage that came to my attention was in the village of Bajob, in a context dominated by Christian and cosmopolitan influences which did not preclude that rumour attributed such great misfortunes as befell the spouses, to their breach of marriage taboos.
 The Manjak pekiin may be an example of the deme, a concept Murdock tried rather in vain to introduce in social anthropology (Murdock 1965).
 It is unclear yet which factors govern burial at either the village benii or the ward benii.
 For sacrifice prompted by sacrificial obligations stemming from a humans contract with a deity, the Creole language uses the apt expression of torna boka, to return the mouth (which has made the promise) .
 Mama Jombo, the main Sacred Grove of the Cobiana ethnic group, turns out to be much more easily approachable than the Manjak equivalents, e. g. the Sacred Grove of Calequisse. The minimal divinatory requirements as discussed, for the Manjak, in the t ext, scarcely apply at Mama Jombos shrines (she boasts five, all situated closely to one another). This accessibility may be an important reason (but what causes it in the first place?) for Mama Jombos very extensive inter-ethnic and international clientele, including not only Guineans from the peninsula (among whom Manjak) and beyond, but also Diola villagers from Senegal, and Europeans. Efficiently, the priests deal with the clients in the Creole language - while Manjak is the only language tolerated by the Manjak deities at their shrines. On one occasion, I found the waiting area of the Cobiana shrine so crowded that the impression was conveyed of a modern bureaucratic institution.
 My own oral-historical data, in conjunction with Carreiras account (espec. 1947) .
 Cf. Collomb & Diop (1969); Diarra (1966).
 Such diviners are found in the major centres of Manjak migration both in Senegal and in France, and not by accident. Thus, in April 1983 I personally witnessed, in the town of Canchungo, part of the initiation (bupene) of a diviner who was specifically initiated so as to take up, immediately afterwards, a divination practice in France.
 According to the official exchange rate, one Guinea-Bissau peso equalled c. US$ 0.025 (first half 1983); exchanged on the black market for French and Senegalese currency, its value is said to decrease to c. US$ 0.003. Full-grown pigs, priced at peso 10,000 and more in Guinea-Bissau and even heads of cattle, which may be twice as costly -, may thus come within the reach of the average returning migrant.
 Cf. Amin (1974); Amselle (1976); Gerold-Scheepers & Van Binsbergen 1978; van Binsbergen & Meilink 1978b.
 On closer analysis, many cases of misfortune involving lesser defies may turn out to have rather more extensive sorcery connotations than would meet the eye at first. In a way, sorcery interpretations would seem to form an embarrassing and hidden secondary layer underlying the participants more overt (but still rather secretive) interpretations in terms of ritual reglect. But even so the deities involved are largely considered to be morally indifferent - except the Sacred Grove of Bekasha, near Caiomét (between Caió and Calequisse), which punishes sorcerers and whose priest provides a somatic treatment for bone fractures thought to be caused by sorcery.
 For a related argument, cf. van der Klei 1984.
 For theoretical elaboration, and descriptive parallels from Zambia, cf. van Binsbergen 1981.
 While locally the ritual structures of the Manjak bear close resem- blance to those elsewhere on the Upper Guinea coast, e. g. among the Feloop, other Diola groups, the Balanta, Braam etc., only the Manjak seem to incorporate these structures so effectively in the context of migrancy. Apart from the fact that Manjak are more than the other groups in Guinea-Bissau (but not much more than c.q., the Diola of Senegal) involved in labour migration, (cf. de Jonge et al. 1978) I cannot explain this state of affairs; and of course, this explanation is too partial as long as their relatively excessive rate of migration itself is left unexplained. A comparative analysis of migra- tion and ritual on the Upper Guina coast would be most illuminating in this connexion. It would also help to assess the present-day situation of migrants rituals among the Manjak as either transitional and ephemeral (with the napenes sui-generic exploitation as a develop- ment that may increasingly dominate the Manjak ritual scene in future), or rather (as I suspect) as more ancient and permanent. Diop (1981) does give a history of Manjak migration, showing that it is by no means a phenomenon that started only a few decades ago. The case material I collected also suggests that the migrants rituals as described here have a considerable history.
 Meanwhile it is remarkable that my data do not reveal the slightest trace of prophetism, which elsewhere in Africa is such a significant ideological response to labour migration, and which has been recorded in the nearby Casamance region of Senegal (Girard 1969).
|page last modified: 13-02-01 11:58:52|