Version classiqueVersion mobile

A Complex Polity

Anne-Marie Peatrik

Part two. Multivocal generations: Weaving kinship and politics

Chapter four

Composite Kinship

Texte intégral

1. The vocabulary of kinship: terms and attitudes

1Some remarkable features can be observed in the vocabulary of kinship, not least of which its heterogeneous appearance. The generations marked by an identical logic to that of the set system are combined, at the level of Ego and adjacent generations, with designations that differentiate between blood relations. A distinction made between maternal and paternal relatives, on the one hand, and between parallel and cross cousins on the other, results in the separation of three or even four categories of cousins, then disappears from alternate generations, while indicating preferential marriages.

2Maternal and paternal grandparents and their grandchildren call each other jûûjû (fig. 16). The latter are considered to be the rebirth of the former; relations between them are warm and the affection that suffuses their relationship can be openly expressed. Beyond that, individuals are designated by new terms, or terms modified by the addition of prefixes and a suffix marking the growing distance between generations. The reciprocity in naming continues. In G+3, ngûkû, which is the diminutive of kiojûûjû in G–3, reintroduces a slight gap: the ties between great-grandparents and great-grandchildren are tinged with avoidance. This is because, since they do not belong to alternate generations, they are like adjacent generations; in the name of the newborn great-grandchild, the parents give a necklace to the great-grandmother in order for her to be able to touch him. In this series, the alternate generation of the alternate generation (G–4) is designated in rather a particular manner, indicating a category of gender: cionkoolo means “eldest mother” or “eldest mum”. However, the occasions when these terms can be used are rare or hypothetical and among the elderly informants, only experts in the language can enjoy listing them. Thus, kiûjûûjûi in G–5 involves a grammatical operation: to the alternate term kiûjûûjû, in G–3, is added an augmentative suffix - i that increases its seniority to the point where it surpasses the “eldest mother”. This table of appellations according to generational depth even ends up having a paradoxical effect, since it might suggest that genealogical memory stretches far back in time. In truth, however, informants’ memories hardly reached any further back than their grandparents. What should be read into this vocabulary of kinship is the influence of generational rythm and the imprint of its cycles.

Fig. 16 — Kinship terms and generation levels.




cionkoolo (eldest mother)














ngûkû (diminutive of kiûjûûjû)

3Baaba (pl. aba) is the term most often used to designate a father in both address and reference, in the family and in the system (fig. 17). The older and younger brothers of the father are designated by technonyms, baaba mûkûrû (“father older”), baaba mwanake (“father younger”); when used as adjectives, -mûkûrû and -mwanake serve to qualify anyone older or younger, in a position otherwise defined, and they can also be found in other compound words. When they have been initiated, sons and daughters call their father baate, a term imbued with particular solemnity that conveys quite clearly the tenor of the relations between a father (or the two parents) and his children: they are marked by coldness, distance and silence. A father only addresses his children through the intermediary of a third person, and vice versa. When used to refer to someone, baate is the diminutive of the term that designates a father: ithe (pl. baîthe). This term enters into compound words of the type ithenthaka, the “father of the son”, which is used when the son has been initiated and can also designate a father-in-law, in the sense of the father of the spouse. For a father, his son is a muthaka (“Bushman”, warrior) whom he can also call baaba (“father”) by virtue of the fact that every man is meant to bring about the rebirth of his father through a son, who is therefore like his father. While he is a child, a son calls his mother ntee; when he grows up and becomes more reserved, he uses îcia in both address and reference. This is one of the many Kimeru terms derived from the verb gûciara, “to give birth”. The older and younger sisters of the mother are, like all women of that generation, ciomûkûrû (“mother older” than her own mother) or ciomwanake (“mother younger” than her own mother). Diminutives are frequently used in the family context: nkoolo for ciomûkûrû, and kaake for ciomwanake. They are the sign of overt affection but are sometimes used to express a subtle difference in the position of the person thus designated: calling a woman kaake instead of ciomwanake brings her closer by taking away a little of her generational position. Ina also means “mother” and can be found, for example, in inonthaka—mother of the (initiated) boy—which can designate any mother or mother-in-law (mother of the spouse); this synonym is probably of Nilotic or Cushitic origin (Huntingford 1955: 34).

4Ego’s brothers and sisters, and the patrilateral parallel cousins, are mutaanocia or mwareocia. These cousins are like brothers and sisters, one cannot marry them. However, the etymology of these terms contains certain particularities. Mutaanis the noun that designates the initiated young man. Mware- means “girl” in the sense of “sister” or “daughter”, and only the possessive adjective allows one to distinguish between the two:

mware wacia: my sister

mware okwa: my daughter

aare bacia: my sisters

aare bakwa: my daughters.

5Moreover, siblings and patrilateral parallel cousins are designated with reference to the mother, as expressed by the suffix –cia. This is quite understandable in the case of siblings, but more surprising for half-brothers and half-sisters and even more so for patrilateral parallel cousins. The synonyms that are sometimes used, mutaanoina and mwareina, follow the same sense, since ina, like îcia, designates the mother. However, a modification appears with the female Ego, who uses mwareocia or mwareina to designate a sister but can also use mwareoba, a term which in this case refers to the father (-oba); mwareoba is always used to designate the daughter of the father’s brother. In short, unlike her brothers, a woman can make a distinction between a sister and a half-sister or patrilateral parallel cousin (fig. 18).

6Other particularities can be grasped at the level of the cross cousins on both sides and the matrilateral parallel cousins. The children of sisters are ndûgû and can intermarry; some of the informants explained that since the intended are of different clans, there is no reason to oppose their marriage, while others considered that this makes a good marriage, because it is good to share the same maternal uncle. This enlightens the particularity of the term a woman uses to designate her sisters, which is different from the term she uses for her half-sisters and patrilateral parallel cousins: their children can marry each other. The complete designation ndûgû wa nda imwe, literally “ndûgû of the same womb”, condensed into ndûgû wa ndemwe or ndûimwe, refers to the mother who has given birth to two daughters.

  • 1  Following standard notation, and designating kin in relation to Ego: D = daughter; S = son; Z = si (...)

7One cannot marry one’s matrilateral and patrilateral cross cousins; avoidance is particularly marked with regard to these relatives, but their mode of designation depends on a structure that can only be understood with reference to the previous and succeeding generations. The mother’s brother is, in both reference and address, muntuetû (diminutive muntuû), “our man”: this appellation combines the substantive muntû (pl. antû) common to Bantu languages and the collective possessive adjective etû-, “our”, which marks both proximity and ambiguity. The sister’s children are mwonomware (pl. anabamware), that is to say “children of the girl”, mwono being an altered form of mwana, “child”. A niece (ZD)1 is also mukenye wacia (“our non-initiate girl”), while the nephew (ZS) is nkaabu or even kîmole. These last two terms, particularly salacious, designate both the penis of an uncircumcised boy and the end of the foreskin that is cut off and thrown on the ground by the circumciser during initiation, and which the warriors stamp on while calling on the vultures to come and fetch it. Here, of course, we have entered the register of joking relationships, and only a mother’s brother and a sister’s son amuse themselves by calling out to each other in such a manner. The sister’s son may behave with a degree of ribaldry towards his uncle’s wife; he calls her bankithoi or munyori, terms of studied rudeness, since he is treating the maternal uncle’s wife as a woman with whom he might have sexual and marital relations. That being said, there is a time and place for joking, and when more formal relations prevail between a maternal uncle and a nephew who has been initiated, the reciprocal term baite can be used; the term also has currency in the political register between two generation sets between which power has been transmitted. This digression allows us to better address the issue of matrilateral cross cousins. The term muntuetû (maternal uncle) is also used to designate his son and grandson; the daughter and granddaughter of the mother’s brother are îcia, “mother”. The merging of generational levels can be observed in reverse symmetry at the level of the patrilateral cross cousins. The father’s sister is mwarewaaba (the “father’s daughter”), and her children are mwonomware waaba (pl. anabamware waabaa), the “children of the father’s daughter”. When, in turn, a brother designates his sister as a “daughter”, transforming a degree of collaterality into a degree of filiation, to paraphrase F. Héritier (1981: 48), he is effecting a shift that propagates to the next generation: the children of this sister, mwonomware (or “child (ren) of the daughter”), come under the relations that prevail between an uncle and his uterine nephews.

Fig. 17 — Main terms of kinship: male Ego

Fig. 17 — Main terms of kinship: male Ego

Fig. 18 — Main terms of kinship: female Ego

Fig. 18 — Main terms of kinship: female Ego

8This way of naming cross cousins by raising the matrilateral cross cousins and lowering the patrilateral cross cousins is a specificity of the Omaha kinship terminology, congruent with a regime of patrifiliation. Here, however, it is accompanied by a rule of preferential marriage. Cross cousins are considered blood relations with whom marriage is forbidden for as long as the Omaha generation shifts are apparent. When they cease to be effective, and no particular term is used to designate the descendants of cross cousins, it is said that “ndûgû is back” or “ndûgû starts again”. Ndûgû is, let us recall, the term used between the children of sisters (MZch), indicating that a marriage is possible, and even appropriate. The reuse of this term expresses the idea that a marriage can take place, an alliance can be formed between those who have become like the children of sisters. Depending on the sex of the person who returns, either there is a form of marriage with the descendants of MBD—the former muntuetû again gives a girl— or the current is reversed, the side that was formerly the “wife-taker” becomes the “wife-giver”: this is a form of marriage with the descendants of FZD. As a third and final meaning of the term, the clan of the mother, that of the mother’s mother, and that of the father’s mother are also considered ndûgû; one can take a wife from these clans provided that she is from a different mûciî (family) to the women just mentioned.

9In short, the term ndûgû designates potential allies who, in one way or another, include all the cousins except for patrilateral parallel cousins: the various different possessive adjectives or suffixes that are added mark the limit between what is impossible and what becomes possible, and the ndûgû in G+2 born of cross cousins are as if announced by the ndûgû in G 0. This broad and yet precise sense of the term ndûgû and its role as a semantic switch explain why, when the informants talk about it, they affirm one thing and its opposite, in good faith. In the field covered by the term ndûgû, however, one designation does not appear: the exchange of sisters, which is antinomic to the preferential marriage between children of sisters, is pointless and excluded. Asymmetry is sought and alliances are orientated.

10That forms of preferential marriage should appear in kinship is hardly surprising, insofar as the generational system functions along these lines. Yet other features are more unexpected. Marriage between children of sisters has been noted in eastern Africa, among the Gurage of south-western Ethiopia (Needham 1969), for example, or the Giriama of the coastal hinterland of Kenya (Parkin 1991: 236-238), but to the best of our knowledge, it has never been recorded as preferential marriage. The distinction made between siblings and MZch, or even FBch for a woman, is reminiscent of a Sudanese type of terminology. It is associated with a typical Omaha terminology in that it concerns the relation constituted by the siblings of opposite sex, but it serves to delimit the prohibitions beyond which an alliance can be formed. The presence of a kinship terminology that disregards the levels so carefully maintained by the generation sets is equally astonishing. While the lowering of sisters by one generation does not really conflict with the generation-based approach—it is, after all, the fate of the underaged in the generational system—raising or immobilizing the children and grandchildren of the sister’s brother calls for study, since it means taking the system backwards or suspending the flow of time. We must therefore examine more deeply what happens in kinship and alliance, where many of the habitual points of reference seem to be muddled, to enable us to untangle the combinations and interactions between this level and that of the generational system.

2. The uthoni alliance and bridewealth

11Nthoni is a term that one quickly learns to spot in conversations. It may be said of an ill-mannered person that (s) he lacks nthoni. Misbehaving children are scolded with the admonition Ikwa nthoni! (“Behave yourself!”), which brooks no reply. A speaker seeking to attract attention and obtain silence shouts out to everyone within hearing: Nthoni! (“Quiet please!”). One passer-by is hailed with a cry of Wee muthoni! (“Hey you, muthoni!”), while another individual mentions, in his explanations, the help provided by his muthoni. In this second form, muthoni (pl. athoni) no longer designates a psychological quality and an attitude (nthoni), but potential or real allies. Thus, any representative of the opposite generational division (the other gîtiba) can also be qualified as muthoni, that is to say an individual whose sister or daughter one is married to or could marry, or who could himself marry one’s sister or daughter. The man who is hailed in this way could be neither an ally nor even a member of the other generational division, but by calling him muthoni one establishes proximity while still maintaining a certain distance. Whence its use as an introductory ploy in certain circumstances. In the second occurrence, the speaker is designating an ally in the strict sense of the term, with whom relations said to be uthoni have been established: by definition, these relations are imbued with respect and restraint, but any failure to observe the codes of good manners, through either shortcoming or excess, indicates a lack of nthoni. The relation of alliance shows the extent of the good manners; under such conditions, it is difficult to classify the combinations of attitudes and relations. The vocabulary of alliance (fig. 19 and 20), described in the following paragraphs, contains specific appellations alongside technonyms and terms used in the contexts of consanguinity and the generation-set system.

12The first steps towards establishing uthoni relations are undertaken when a warrior and a non-initiated woman are planning to join together in matrimony. The suitor, or his father, takes a bundle of khat to the girl’s father to enter into conversation, thereby signifying that he “desires the alliance” (kwenda uthoni). By accepting the gift, the girl’s father, who has been notified in advance by a close relation, indicates that he is willing to address the subject. This is an important step, but it is not decisive: the khat only serves to “get the father to talk”; the discussion may end too soon and the suitor’s attempt come to nothing. Sometimes the girl uses this opportunity to make known that the project is really not to her liking. Her father and mother also give their opinion, but “arranged” marriages against the will of the betrothed are generally avoided: they are seen as a source of instability and the possible failure of promising relations which it would be inappropriate to seal under such auspices. One exception to this rule is the marriage of a daughter to a polygynous man, to which the father has been enticed into consenting by the offer of far more than the customary amount of bridewealth: in this event, instead of being conducted in the open as is usually the case, the transaction is carried out more discreetly, and the customary process, which can take several years to come to completion, is abbreviated. The girl’s brother sometimes endeavors to speed up the decision in one direction or the other, but in the event of disagreement, the father’s opinion is decisive, and there can be no appeal to the council of Fathers of the neighborhood or the territorial section.

Marriage prohibitions and preferences

13The preliminary discussions are the moment when the kinship ties of the betrothed are checked: as a result the marriage may appear impossible or, on the contrary, favorable. Of course, the betrothed are of the appropriate generation sets: nobody would dare to become a kioncho—an individual guilty of generational incest—and identifying their sets is one of the preliminaries of courtship. To which is added the control exercised by the warriors over the company kept by their companions and by the category of girls that they consider to be theirs. The betrothed, on the other hand, may be unaware of the ties that might oppose their marriage. It must be verified that the planned marriage will not transgress the taboo of family incest that would make them win’janga. The clan is always presented as being exogamous, but given the polysemy of the term, the vagueness of the limits and the liability of clan affiliations, the effective patrilineal ties are always carefully examined: the mûciî on both sides must be carefully checked, because this is the level at which real consanguinity appears. If the betrothed have the same clan patronym but come from mûciî without any known relationship as far back as can be traced, then the marriage can proceed. It is accepted without enthusiasm, with the observation that there were better matches to be made, because marriage is the perfect opportunity to establish new relations or renew existing ones. It is the women who express the strongest reservations because, they say, in the event of domestic violence and resulting injury, nobody will help the victim or claim compensation for her. The exogamy of the clan can also be understood in the sense of exogamy of the territorial division: people with the same clan patronym but from different territorial divisions can marry; it is less unsuitable than marrying someone from the same clan patronym and territorial division.

14In addition to the identification of an effective patrilineal, patrilateral and matrilateral link, as we will see, which prevents any marriage on pain of committing incest, the doubling of an alliance concluded by a brother (his brother marrying another sister) is so inconceivable that it is not even explicitly forbidden. When invited to explain the reasons for an assumption that never fails to surprise them, the informants mentioned two incompatibilities. The principle of levirate marriage could not be applied, because the surviving brother would have two sisters as wives, which is a form of incest. And then, what would happen to the children of the two sisters? In this, we can see that the ndûgû relation is also at the heart of this evidence.

  • 2  Lambert (1945, typescript: 272) reports that depigmentation (rwanga), leprosy (ndwanga or mutiri) (...)
  • 3  There is another category of individuals in this same region, called nkio, “Those who have a tail” (...)

15The gichiario tie is another source of marriage taboo. This term, again derived from the verb gûciara “to give birth”, designates a form of blood brotherhood established between individuals of two clans and two territorial sections, at an epoch that might be far-removed but the memory of which is handed down in the families concerned. As a general rule, each clan has two or three recognized gichiario clans. The gichiario cannot fight or marry; they owe each other hospitality, mutual aid and assistance, in the event of the premature death of one of the parents, among other things. Nobody can refuse to help a gichiario, and since every gichiario is a potential scrounger, gichiario tend to avoid each other in their day-to-day lives. The significance of marriage taboo is not the same as that of family incest in the strict sense of the term, although the term win’janga is used to qualify those who are guilty of it. The elders delight in telling horrible stories in which lovers who did not break off their relationship after being informed of a gichiario between their respective clans were crucified coupled together by the side of the road; people threw stones at them and they were then left there, without food or drink, until they died. Nevertheless, this type of marriage did occur; it took place in the register of the abnormal, and the bridewealth was not paid. It is also said that a depigmentation of the skin, if not leprosy2, is one of the consequences of transgressing gichiario taboos. Persistent rumors in the north of Tigania relate that marriages between gichiario did take place, with unexpected and probably revelatory consequences. A long time ago, a number of gichiario from the Imenti found refuge with their gichiario among the Tigania. Despite the taboo, some of them intermarried; they became nthea, that is to say people who are “spotted” like a giraffe, a metamorphosis that raises the question of marriage in terms of the three clan colors, since the transgression of the gichiario ties resulted in the appearance of a new category of color, the “Spotted”. In this case, the skin disease that was held to be the consequence of the transgression led to the appearance of a new category, because of the number of people involved3.

Fig. 19 — Allies: male Ego

Fig. 19 — Allies: male Ego

* Cio. .: uses the name given to the spouse by her father, or even muka or mwekûrû

Fig. 20 — Allies: female Ego

Fig. 20 — Allies: female Ego

* The prefix Ba- is followed by the name of the gift they exchanged
Mwuntû wa…: uses the gaaru name and then the nthukî name, or mûkûrû

16Combining our few elements of information with the more substantial data gathered by Mahner (1975: 405) suggests that there may have been a marriage code based on the three clan colors, which the njûrûrû songs were intended to recall when, for example, a warrior had set his sights on a girl who was forbidden to him because of their respective colors. Recording one hundred and sixty marriages contracted between 1930 and 1960, Mahner noted a widespread practice whereby White men married Red or Black women, and Black men married White or Red women, but a Black man would never marry a Black woman, nor a White man a White woman. Red men, on the contrary, could marry Black, White or Red women. The Whites and Blacks were exogamous colors, whereas the Reds could equally well marry any of the three colors. These practices are not, or no longer, codified by explicit rules. They concern the colors which are clan categories, but there is evidence suggesting that they could relate to an old territorial partition of which one can glimpse the weave in the sections and subsections designated here and there by colors (see Appendix B). The matrimonial language of the three colors, whether of clan or territory, is not far-removed from other orientations that appear in the field of alliances.

  • 4  There is, however, another form of union against which everybody is powerless. A category of indiv (...)

17The study of prohibitions leads imperceptibly to marriage preferences. Verifying the family ties of the betrothed may not bring to light impossibilities, but prospects that are held to be favorable if the union involves one of the ndûgû situations mentioned above. Informants present what are described in the professional jargon as “preferential marriages” in various ways: a good marriage is one where the young couple, from different places, meet, like each other, and, moreover, come into the category of auspicious unions. These informants express a hierarchy of criteria, at the head of which is the origin of the betrothed (it is good for them to come from different sections), followed by personal inclination (the guarantee of a stable relationship), and lastly how it fits into the existing fabric of relations. A markedly different perspective is given by one manner of implementing the ndûgû relation. A young man, it is explained, has designs on a girl from the same clan as his mother or his grandmothers, but from a different family of course. If the elderly of his own clan, whose opinion he has sought, say that it is a good ndûgû, then the young man will be encouraged to persevere in his courtship. A ndûgû relationship is judged to be propitious when it has proved to be a success in terms of the women’s fertility, the children’s health and harmonious relations between everyone. These different points of view reveal a stronger influence on the orientation of marriages than the elderly are willing to concede a priori. It is the elderly who carry the memory of past marriages and contribute to the building of reputations. The special relationship between grandchildren and grandparents exerts a particular influence in these affairs. While the idea of confiding one’s emotional ups and downs to one’s parents is inconceivable, these matters, like many others, can be freely discussed with one’s grandparents, who are (were) responsible for the psychological and emotional development of their grandchildren. And the grandparents are best-placed to know where, for example, a grandson is likely to find a descendant of cross cousins whom it would be good to marry, especially as the ritual of the Accomplished, which rather resembles a diamond wedding anniversary celebration, appears to punctuate the passage of time and recall past agreements4.

Agreement reached!

18When negotiations lead to a marriage agreement, the fiancé, his father, and the father of the fiancée call each other muramû (pl. aramû). The subsequent steps can now be taken (fig. 21). The fiancé’s mother, or a woman from his family, now comes with a gift of bananas (considered a male product) and says that she also wants to become muramû. This step and the resulting extension of the term muramû indicate that three lines of descent are brought into relation with each other, that of the fiancée, that of the fiancé and that of the latter’s mother (B, A, C); it would not be inaccurate to translate muramû by “in-law”.

19Other gifts arrive in the same way, defining the relations between the fiancé on the one hand, and the brother and sister of the fiancée on the other: a fighting club made of metal (chuuma) which the fiancé gives to the fiancée’s brother creates the bond and the appellation bakuma; a ring (kîmata) given to the fiancée’s sister establishes the bankithoi relation. These gifts that the fiancé makes to members of his fiancée’s family who are of the same generation as himself are differentiated from the previously established muramû relation. The prefix ba- always signals a relation of equality. These individuals consider themselves to be of identical level and generational status, even if the two men are of different generational divisions. The gift of a fighting club establishes a specific bond between the fiancé and the fiancée’s brother, and recalls the warrior condition of the brothers-in-law; no matter if the fiancée’s brother is not—or no longer—a warrior. The gift of a ring to the sister, by which the bankithoi relation is established, also combines two different registers. A bankithoi is the lover of a warrior, from the moment that he has offered her a ring. That is not to say that this reminder of the generational register gives form to the idea of sororate marriage. Only the death of the fiancée during the lengthy process of the engagement could lead to the deceased being replaced by one of her sisters, but this substitution is unthinkable once the marriage has been performed. It is antinomic to the ndûgû relation because it makes impossible precisely that marriage between sisters’ children which the gift of the ring and the bankithoi relation are intended to anticipate. All in all, the gifts offered by the fiancé and the appellations thereby established combine the categories of the generational system and those of alliance; through the intermediary of metallic objects, specific relations are established between affines and siblings of the same sex or of the opposite sex and the same generational level.

20These first gifts initiate the marriage relationship, but the informants explained that they are also given as compensation for the harm suffered by the girl’s family. It is obvious that the future spouses are already lovers, and the girl’s parents consider that they have been given no choice in the matter. However, the consummation of the relationship is not a transgression, although it is far from devoid of ambiguity, since it must under no circumstances result in conception. The relationships of lovers are developed in the recognized context of generation sets and categories. When they are transformed into marriage, the first payments act as compensation for the head start that has been taken in the generation system, which the family cannot do other than endorse. The suitor who brings a bundle of khat is really addressing a father who feels himself to have been wronged, and it is understandable that he should be offered a present so that he might, possibly, agree to talk with the young man who “desires the uthoni”, a relation that refers, once again, to the generational context. The initial discrepancy for which the gifts are intended to compensate can also, and without contradicting what we have just observed, be interpreted as a consequence of the preferential marriage deriving from the muntuetû relation and, more broadly, from ndûgû relations. Every new union follows more or less in the wake of previous ones, and a father who is solicited by a suitor acts like the representative of a wife-giving line, always inclined to believe that he is still owed something or, on the contrary, of a wife-taking line, afraid that he is being approached to settle a debt. For example (fig. 21), A will either marry a female descendant of FZD or of MBD in G 0, depending on whether line A has given or taken a sister from line B in G–3.

21Retrospectively, the subsequent steps shed light on the significance of the initial transactions. The time comes for paying the bridewealth in the strict sense of the term. The first contributions have one thing in common: they do not come from livestock. The fiancé’s mother, helped by three women from the clan who live nearby, collects four sacks of yams—products and property of men’s farming—and takes them to the fiancée’s family. Then it is the turn of the fiancé’s grandparents, who give pots of honey (male products). The fiancée’s mother then brings gourds of gruels which are offered and promptly eaten by the fiancé’s family. This payment of cooked products, no longer raw, and made in the opposite direction to the flow followed since the beginning of the transaction, is the first manifestation of the fiancée’s maternal relatives. Subsequent payments return to the previously established flow, and they are collectively known as mweerera or nthanîthio. The fiancé’s mother, still helped by women from the clan, gathers and brings products of women’s farming: sacks of peas, beans, sorghum, and millet. This last product is called mweere, from which the name mweerera is derived; the closely-related verb kwereria, which means “to fill” and refers to the notion of abundance, covers the same idea. A popular etymology of the synonym nthanîthio, sometimes pronounced ntanîthio, establishes a link with gûtaanua, a term that means both “to excise” and “to circumcise” and designates initiation. The fiancé’s father ends up bringing a goat, the payment that seals the uthoni relation, after which the initiation of the fiancée is organized.

Fig. 21a — Takers and givers in the establishment of the uthoni alliance

Fig. 21a — Takers and givers in the establishment of the uthoni alliance

Fig. 21b — Takers and givers in the establishment of the uthoni alliance

Fig. 21b — Takers and givers in the establishment of the uthoni alliance

22From the beginning of the transaction, the fiancé’s mother plays an active role in establishing the uthoni relation. She comes specifically to seal a muramû relation, and then takes care of the payment of the mweerera. Her attitude can be understood in the light of the muntuetû relations of which she is the agent. She represents the wife-giving lines (H which gave to C and then C which gave to A) which enabled the existence of the fiancé, who is a sister’s child for his mother’s brother and a sister’s grandchild for the brother of his maternal grandmother; the children born of his son’s marriage will be the subject of payments in favor of his line C (see below). The fiancée’s mother who brings the gourds of gruel thereby indicates that wife-giving lines also exist on the side of the fiancée’s family (E which gave to D, and D which gave to B). In a way, these diverse payments are a reminder that every marriage transaction involves lines other than those of the two betrothed; these are the lines that gave to the lines of the two fiancés, in other words the givers to the givers and the givers to the takers. However, this representation of the different wife-giving lines is brought to an end by the payment of a goat by the father, declaring the uthoni relation to be established. This first payment of a head of livestock recalls the preeminence of the father and his line in the establishment of the relation, by means of a commodity different to those that have circulated up until then.

23The theme of wife-giving lines that emerges in the payments described above is deliberately emphasized in the following phase. Before her initiation, the fiancée, accompanied by her mother and her fiancé, undertakes a trek (kurîria) that lasts several weeks and may take her throughout the whole Tigania-Igembe country. During this trek she visits her muntuetû, that is to say at least three lines (D, E, F), for the informants specify that it is possible to add the muntuetû of her father’s half-brothers, of her paternal grandfather’s half-brothers and even the muntuetû of her father’s other wives. It is obvious that the fiancée is not going to see all of them, but at this moment in time it is important to publicize the largest possible number of muntuetû. The prestige of the fiancée’s family and the effectiveness of their social networking skills are at stake. During the tour, a visit to the mother’s sister may also be planned.

24The arrival of a fiancée at a muntuetû’s house always triggers the same scene. The fiancé plants his spear in the ground in the uncle’s kraal, a most aggressive opening ploy marking the warrior’s primacy over an individual who is as if compelled by a conqueror. The mother and daughter start to dance, facing each other, and sing, demanding the maternal uncle. The wife of the muntuetû comes out and maintains that the uncle, who is in fact hiding, is away. His niece grows bolder, and climbs onto the kiara, scolding her muntuetû and continuing to sing until he comes out and agrees to give her a goat. What is the meaning of this performance, and the payment that is made with such reluctance? Between a mother’s brother and a sister’s child there is, as we have seen, the candor of a joking relationship, not forgetting that the joking is charged with heavy ambiguity. The niece can joke with her uncle, berate him and even go so far as to climb onto his kiara, a gesture of such insolence that only a sister’s child would dare to make it: the scrapheap is the symbol of successful growth where her uncle’s remains will be buried if he reaches the last age grade, and trampling on it is tantamount to cursing him. Normally, it is the uncle who seeks to obtain goats or bulls for every child born to his sisters. By insisting, climbing onto the kiara, and making fun of her uncle’s miserliness in her songs, the niece is reminding him that, just as he demanded payment of a goat for the birth of his niece, so the uncle must now give a goat for her initiation, as a sign of his approval and blessing of the forthcoming marriage. A refusal would signify the opposite and amounts to cursing the marriage. By requiring a great deal of persuasion, the uncle restores the preeminence of the wife-giver, which the conquering gesture of the warrior had undermined. The goat that he finally relinquishes expresses an agreement that he cannot, in reality, refuse; however, the condition of the animal can be used as a means to intimate the degree of reluctance or satisfaction contained in his approval, and expresses the state of family relations.

25During this tour, the fiancé also makes the acquaintance of the muntuetû to whom he will have to make payments for all the children born of his marriage with his lady love (in this case D and B). Every husband-to-be—who already knows the wife-giving lines on his side (lines C, H, G) because, during his initiation several years earlier, he also visited his muntuetû to obtain goats and blessings by means of a spectacle similar to the one just described—must know the wife-giving lines on his fiancée’s side. These tours are also the opportunity to discover the lines in which the children of the future couple could find a suitable spouse (E and F, without forgetting G and H on the fiancé’s side). Lastly, this journey that a woman undertakes, accompanied by her mother, corresponds to an important moment in her life. She is travelling around, outside the confines of her own district, like never before; this visit to the muntuetû is the occasion to recall and maintain all the existing ties and to announce the future ties that will pass through her, thereby demonstrating her importance. The songs from their nkenye tours that the old women remember are impressive in both number and quality of artistic expression, bearing witness to the emotional highlight that this voyage represents, and the contrast with the prolonged period of seclusion that follows.

26At the end of this tour, the fiancée’s initiation is organized in parallel with that of the male sub-set then open for recruitment. When she has been excised, the woman, who is no longer a nkenye but a ngutu, returns to her parents’ home where she will be secluded for months in a hut. During this time, the rûracio, also called the ngombe (“cow” or “bovine”), is gradually brought. According to Lambert (1945, ms.: i, 38), its etymology derives from the verb kûrachia, meaning “to give in a continual exchange”; this sheds light on the content of the relation of alliance, and more generally on every relation that the Tigania-Igembe consider to be never entirely new or never really ended. The seclusion of the ngutu continues until the rûracio has been paid in full. It consists in heads of livestock that the fiancé’s father takes from his own herd, or from the bridewealth of a daughter, which can only be used by her brother. Three goats, a bull, and a heifer are paid, together with clothes made of leather (nguû, generic term for all clothing) for the fiancée’s parents. There are local variations: sometimes two heifers are brought instead of one, not three but five goats, and a sheep too; some old women remember that there was also an axe and a farming staff (armed with a metal point). Finally, to bring out, or even extract, the bride from her hut of reclusion and her parents’ home, the bridegroom’s mother must pay the ndatho, a term which, again according to Lambert (1945, ms.: i, 44), is derived from a verb meaning “to strike a blow, to pierce”: this also refers to the fact that during this final phase, under the cover of a game, a confrontation can be perceived that seems to echo the warrior’s attitude towards the maternal uncle. The ndatho is composed of bananas and yams, products that replicate the payments made at the very beginning of the negotiations and that intervene after a transaction in which the main player was the fiancé’s father. The future husband also fetches a goatskin dress adorned with cowries and pearls, which his fiancée wears when it is time to go.

27When their daughter leaves, the father gives her a cow and two goats (or maybe more), while the mother gives her a large earthenware pot filled with the things she wants to take with her, which constitute a kind of dowry. The earthenware pot will be used to make millet beer. The father also chooses for his daughter the “wife’s name” by which her husband will call her after the birth of their first child. It is constructed in such a way as to recall the name of this wife’s father. For example, when they are married, the daughters of a man called M’Therebe are all Ciotherebe. The father then declares that he “cannot feed two of them” (atiumba kûrithia maurithi yaili), that is to say the heifer received as bridewealth, and his daughter, who must henceforth live outside the paternal mûciî. The juxtaposition of the bridewealth heifer and the daughter is symbolized by the particular names given to the descendants of the rûracio heifer. This latter is called mwalle— an alteration of the term mware, “daughter”—and the subsequent ones are called by ordinal names, kaanja, nchama, ntitû; that is to say, four matrilineal generations of cows, which are a projection of the girls to be born to the daughter who has left in exchange for this heifer. The parallel drawn between the woman and the cow derives from the wider identification between the fate of humans and that of the livestock. Through the descendants of the heifer, it represents a memorization by the wife-giving side of the future relations of which the departing daughter is the agent, and anticipates the marriages that will once again be possible in the ntitû generation (see fig. 21). A mwalle heifer that turns out to be infertile must be replaced. The brother of the girl who has left to get married can use mwalle or one of its descendants for his own marriage. There is also the possibility that the mwalle heifer may devolve upon the father’s sister, if she and her brother are still alive. Lastly, the bridewealth sheep is given to the fiancée’s mother.

28In the event of divorce without children—it is said that “the daughter has returned to her father’s house”—the father must return the equivalent of the nthanîthio and rûracio. Lastly, a man who does not have the means to pay the bridewealth can work for his family-in-law: this service of the son-in-law is called kwoyuwa nî athoni (“being taken by the allies”). We should also specify that Kitigania-Kiigembe distinguishes between the uthoni relation, which seals the establishment of a marriage alliance, and the act of “taking wife” or of “being taken as wife” (to marry, in other words), which refers to the change in personal condition that occurs from the moment an uthoni relation has been established. Two verbs appear, used in a passive or active form: kûgûra, “to obtain (a wife)”; kûgûrwa, “to be taken as wife” (during the colonial period kûgura ended up meaning “to buy” in general); kûyûkwa, “to be taken”, which is the passive form of a frequently-used verb.

29The overall process presents a number of remarkable features. The different constituents of the bridewealth that are successively handed over seem to represent a sampling of the products of economic activity: products of gathering (khat) and collection (honey), products of crop farming, herding and craftwork (metal objects, pottery, leather), products of male and female activities. Products of hunting are an exception, because this activity, when encoded, involves forces that are ambiguous if not harmful. Moreover, the total cost of the transaction is not high, even without counting the sort of dowry that the bride takes with her; it represents a certain quantity of goods, but in total it is a modest sum that does not impose an excessive drain on the household’s resources. It is not the cost that might prevent a man from getting married, but rather the lengthy nature of the procedure, the diversity of payments, and the number of different steps to be taken that could cause problems, because they require the collaboration, and therefore the consent, of many relatives. Another striking feature is the manifestation of the wife-giving lines, both near and far-removed. The next section will throw more light on a mechanism that could not have been ignored and which we must now examine in greater detail.

3. The ndûgû payments and the maternal uncle

30A bond of alliance leads (led) to relations that weave a network where each individual and each family kraal lie at the intersection of a number of wife-giving and wife-taking lines. Even when the uthoni relation has been established, the bridewealth paid, and the young woman initiated and then escorted to the newlyweds’ home by the warriors who, along with the elders, build their first establishment, this does not mean that the payments have come to an end. Every newly married man must give his wife’s brother a ndûgû cow, in other words a cow that opens the ndûgû relation and which, it is specified, does not form part of the rûracio. The rûracio, indeed, has been paid by the fiancé’s father, whereas the ndûgû cow is paid by a married man to his wife’s brother. This cow, sometimes called nkûrio, marks the starting point of a relation involving the pair of siblings of opposite sex and the sister’s husband. This relation is hinted at during the preliminaries, when the fiancé gives a fighting club to his brother-in-law and a ring to his sister-in-law, and it is sealed once the couple is established.

Fig. 22 — Ndûgû payments received by the giver

Fig. 22 — Ndûgû payments received by the giver

Fig. 23 — Ndûgû payments owed by the taker

Fig. 23 — Ndûgû payments owed by the taker

31In the name of the ndûgû relation thus formed, the following payments are gradually made (fig. 22): for every child born to this marriage, a bull (for a boy) or a goat (for a daughter) is given to the maternal uncle or to the son who has inherited his credits. The same payments must be made to the maternal uncle or his heirs when the grandchildren are born: 2 gives to the line A for his children 3 and 4; the line C, in addition to their payments to B, also give to A for the grandchildren 5 and 6. If we shift our attention from one sole line to all the wife-giving lines (fig. 23: D, A, E), then besides the initial payment to A, a father must pay, by virtue of the ndûgû credits, three bulls for every son and three goats for every daughter. There can be no denying that the value of the bridewealth may be relatively low, but the ndûgû payments are very onerous and hardly an encouragement to have many children. They are part of the debts that are transmitted from father to son and of which every sister is a vector.

32After three generations, the ndûgû payments cease but the relation is not extinguished, since it is good for a marriage to be concluded between descendants of the former wife-giving and wife-taking lines. Either the former wife-giving line again gives a girl, or the current is reversed and a girl comes to get married in what had initially been the wife-giving line:

Nobeyire mwaarî ungî kana twîtûîte nao tugîre mwaarî ungî
He can come and find another girl to marry,
or we will go there to find another girl to marry.

Iragwa ndûgû igûcoka mûciî uria yaumire
It is said that the
ndûgû has returned to the family kraal from which it came.

33The ndûgû payments are associated with forms of preferential marriage, and in a way they bring about that which is foreshadowed by the descendants of the mwalle cow. We must examine this association more closely to better understand the mechanisms and implications of this bovine idiom.

34The first thing that should be noted is that while the children of sisters are ndûgû from the outset and can marry each other, the descendants of cross cousins only become ndûgû when the ndûgû payments cease, as if the duration of the payments corresponded to the time needed to transform a cross cousin whom one cannot marry into a matrilateral parallel cousin whom one can marry. Detailed analysis of the vocabulary of kinship (fig. 24) reveals this transformation. Remember that for a sister’s son, his maternal uncle and the latter’s son and grandson are muntuetû, while the daughter and granddaughter of the maternal uncle are îcia, a merging of generations or obliquity that is characteristic of Omaha terminology. However, the people bear in mind two different categories of kinship terms that function simultaneously: informants propose the first of these spontaneously, and they use the second when asked for clarification, in a process that, according to P. Menget (1977), one might describe as “alternative designation”. Thus, the son of the mother’s brother is muntuetû, but he is also mwana wa muntuetû “the child of muntuetû”, a technonym that restores the generational level. The daughter of the mother’s brother is îcia (a mother) and sometimes maama (meaning “male mother”), but she is also called ciomwanake (dimin. kaake) “youngest mother”, indicating a difference in age rather than generation, or mware wa muntuetû (“daughter of muntuetû”), using a technonym that denotes filiation and generation. The mother’s brother’s grandson is still muntuetû, but he is also called jûûjû, “grandson” or “grandfather”, which may seem surprising given the position of Ego (one generation above), but which makes sense if we relate it to his position as the mother’s brother’s grandson; this is a use of the generational code according to which a man and his grandson are the same—what comes after is the return of what came before, a way of combining the Omaha logic with the passage of time, and of restoring the appropriate generational level in relation to the maternal uncle. The principle of the identity of alternate generations, observed in a number of populations, may seem trivial, but among the Meru it assumes particular depth, because it is the only symbolic mechanism through which the rebirth of an individual takes place. The granddaughter of the maternal uncle is ndûgû-îcia, a term that combines two contradictory indications, that of a cousin whom one can marry, and a suffix -îcia (mother) which shows respect and refers to this impossibility: the name ndûgû-îcia combines the possibility and the prohibition of marriage and thereby signifies that one can almost marry, as the informants explain very simply. After this, “one is no longer called” by a particular name; it is said that “ndûgû comes back” or “starts again”, and an alliance can be formed.

(=) very marked avoidance
(-) marked avoidance
(+) marriage possible

Fig. 24 — Omaha obliquity, alternative designation, and preferential marriages


ndûgû, ndûimwe (ndûgû from the same womb)




> ciomûkûrû diminutive nkoolo
< ciomwanakî diminutive kaake




mwana wa muntuetû (child of muntuetû)


ciomwanakî, dimin. kaake (youngest mother)
mware wa muntuetû (daughter of muntuetû)


or jûûjû (grandson)

mbsd, mbdd

ndûgû îcia (ndûgû through our mother)


mwarewaaba (father’s daughter)
mware wa ndewe wa baaba (girl who has the same mother as papa)
nkoolo, kaake (youngest mother)


mwonomware (child of daughter)
mwononkoolo (child of youngest mother)
mwareciomûkûrû (daughter of the eldest mother for fzd)

fzdch, fzsch

jûûjû (grandson, granddaughter)

fzdchch, fzschch

ndûgû yetû (our ndûgû)

35An identical process of alternative designation can be observed in the case of the patrilateral cross cousins. In reference, the father’s sister is mwarewaaba (“the father’s daughter”) but people explain that this is also the contraction of mware wa ndemwe wa baaba: “the daughter who has the same (mother) as father”, ndemwe being the contraction of nda imwe, which refers to the womb from which she came, a way of restoring the appropriate generational distance by referring to the mother of Ego’s father and of Ego’s father’s sister. In address, the father’s sister is nkoolo (dimin. of ciomûkûrû) or kaake (dimin. of ciomwanakî), depending on whether she is older or younger than the father. These terms are identical to those used to designate the mother’s sisters: all these women are from the generation above Ego and they are all related to the latter; they are like older or younger mothers. The father’s sister’s children are mwonomware (“child of the daughter”), but they are also called mwononkoolo (“child of the eldest mother”); the father’s sister’s daughter is mwonomware (“child of the daughter”), but also mwareciomûkûrû (“daughter of the eldest mother”). Their children are always mwonomware, but also ndûgû yetû (“our ndûgû”)—in contrast to ndûgû îcia, which designates the children of matrilateral cross cousins—or jûûjû (the grandchildren), which restores the appropriate generational distance from the level of the father’s sister. It will soon be possible to marry these ndûgû yetû who are also jûûjû; a marriage can take place in the next generation. Although she knows that marriages can be contracted, a woman continues to mark the differences between generations (see fig. 18): the descendants of her father’s sister are described as ngûkû (great-grandchildren), whereas at the same generation level (G+2) she no longer designates the descendants of her mother’s brother, as if time was somewhat standing still on the maternal side. These slight differences in vocabulary probably stem from the fact that the male Ego has to seek a wife.

Fig. 25 — Cattle links, kin terms, and preferential marriages

Fig. 25 — Cattle links, kin terms, and preferential marriages

36All in all, the Tigania-Igembe employ the Omaha obliquity in a quite particular fashion: it undeniably exists and they make use of it. However, the use of this terminology is inseparable from the process of alternative designation, that which denies time and that which restores it, aligning the terminology with the dominant temporality of the generation-set system. It is also in accordance with a very Meru practice of combining different sorts of time: linear time, cyclical time, continuous flow, discontinuous flow, suspension, acceleration. This process is all the more effective since a phrase heard very early on, but the significance of which only became apparent when relating the Omaha terminology to the ndûgû payments, provides the key to it. Concerning the ndûgû payments, it is said in the simplest possible manner that “after three generations, we forget”, and it is of course the term nthukî-generation that is used in this expression. The counting of the generations starts from the pair of the brother and the sister-wife whose husband paid the ndûgû cow. It is therefore no accident that the husband, the wife, and her brother are expressly clasped within the generational relation. The ndûgû payments are made during three generations, and the concomitant effect of the Omaha generation shift gradually fades away until marriage is again possible. However, the obliquity of the terminology has a double effect which probably explains its recurrent use. It is consistent with the preferential marriage rule of the generational system, whereby the sons and daughters of a given set of fathers can marry each other—except for brothers and sisters, of course: the idea that for a brother, his sister, shifted down to the next generation, is his daughter, expresses this impossibility with great economy. It also establishes, at the heart of the egalitarian generational system, an asymmetrical relation where the wife-taking line is indebted to the wife-giving line. The opposite approach, where a sister is shifted up by one generation and considered like a mother, is characteristic of a Crow terminology most often associated with matrilineal filiation. It runs counter to the generation system in three ways: with regard to time, filiation and generational incest. And yet this is precisely what happens during the ritual for the establishment of a new warriors’ gaaru (see Ch. 6): a girl, the sister of a warrior, is promoted to the position of mother of the warriors, and she is married, moreover, to a blacksmith of a rather particular type (but in principal endogamous), the goldsmith, which represents a fourth transgression. What is performed on the day of the ritual indicates, by inversion, the daily norm. The beginning of a Crow obliquity can also discerned in the way that a female Ego uses the same term to designate her brother and her son, adding a different possessive suffix:

mutaanocia: my brother

mutaana okwa: my son

ataanabacia: my brothers

ataanabakwa: my sons

37The term mutaana refers to the state of initiate; by describing her brother as “my initiate” the woman is treating him rather like a son and performing a similar manipulation to that applied to a sister by her brother. However, the shift is not taken any further.

38When “ndûgû comes back” after three generations, what alliance can be concluded (fig. 25)? It could be marriage with a descendant of a matrilateral or patrilateral cross cousin. The informants specifically emphasize that both of these occurrences are possible, something that the vocabulary—also used to greet each other—reveals in the differentiation between the two sides, which are either ndûgû-îcia, i.e. ndûgû matrilateral, or ndûgû-yetû, i.e. ndûgû patrilateral. In the former case, the ex-muntuetû gives another girl, and the former wife-giving line remains wife-giving. In the latter case, the current is reversed, the side that was formerly the giver becomes the taker. The possibility of this inversion is present in the previously mentioned term baite, which we should re-examine. A maternal uncle and his nephew are muntuetû/mwonomware, but once the nephew is initiated, they may call each other baite: the prefix ba- and the reciprocity together signal a relationship of equals. More generally the term designates all lines and clans holding a giver/taker relationship with a given family or individual. Baite is also an important term from the generational register, used between two generations that have handed over power to each other, and more generally by representatives of the two generational divisions that alternate holding the power. Whereas the term muthoni, often used when hailing a man, refers to either a potential or actual relationship of alliance, baite, a term heard just as often, signals that one is addressing the other as one would a mother’s brother with whom there is no longer a relationship of subordination, which opens up possibilities of exchange. Baite even became the nickname given by the Kikuyu to Meru they encountered in Nairobi, so much was the term used as a greeting among the latter. Following a process of alternative designation that the Tigania-Igembe seem decidedly fond of, the appellations baite/baite and muntuetû/mwonomware include a basic variation between equality and inequality, and carry the possibility of inversion, by which the former superiors, the givers, can become takers, and vice-versa. Additionally, the double appellation covers the transformation of kin into potential affines, leading thus to two possibilities. Yet only one of these possibilities—marriage with a descendant of a matrilateral cross cousin—seems more familiar to informants; the other possibility, though apparently less frequent, is still valued by some.

39It seems necessary at this point to provide information on the representation of the body and of the person. Together with the concept by which, with the passing of age grades, maturation occurs ending with the head—the last organ to take shape—the body is conceived as if divided lengthwise into two equal halves, the left side being that of the mother and the right side that of the father, a division represented by the iraa blessings applied in a line from the bridge of the nose up the middle of the head. The lines of the hand are also meant to carry the mark of both mother and father; it is said that people could read them to identify maternal and paternal clans. Thus a warrior could find out whether he had, during battle, killed a kin, for which he would have to pay in blood money. The heart— which as we have already seen is the organ of speech—is located in a spot where the principal of lateralization and categories of gender are combined: the heart of a man is on the left, that of a woman on the right, a disposition exemplified at the moment of coupling when the complementarity of their bodies is demonstrated by the joining of their hearts, which at that moment face each other. Lateralisation also defines the layout of the family home, determined in relation to the exit: the father’s gaaru is on the right, the mother’s house on the left, and the kiara, also on the left but between the hedges, in a liminal space. Theories relating to disease and good health fall within this logic and refine it: any disease that affects a child or an adult is, save in particular cases, interpreted by the diviner-healer as the presence of mûgiro, a contamination the result of a deliberate or involuntary transgression, of which the culprit is to be found on the maternal side, and more generally, among the givers. The diviner evaluates the muntuetû relations of the individual and of his mûciî to identify the muntuetû who is suspected or held responsible, which means finding the person with whom relations have been poisoned for one reason or another. A delay in ndûgû payments is a reason frequently given. Within this concept, good influence, like bad influence, springs from the muntuetû. If the child walks, talks, grows, and reproduces normally, it means the muntuetû relations were good, that the muntuetû are a good side, a good match in a way. At this level a strange passivity on the paternal line is detected, because a father cannot be held directly responsible for the mûgiro affecting his children. He is its conveyor, however, in two ways: through him passes the mûgiro of the paternal grandmother’s brother, who is responsible for various particular diseases; and then there is the mûgiro from the muntuetû of a young bride, another reason to add to those restricting polygyny.

40This theory of the body and the individual permits a better understanding of how the muntuetû/mwonomware relationship and the ndûgû payments work. The muntuetû has a formidable influence over the sister’s children and the fulfillment of a family’s destiny. This web of relationships—through which currents both auspicious and malicious circulate, the result of either blessings or curses from the muntuetû—also explains the round of visits, as large as possible, undertaken by a girl or a boy to the muntuetû just before initiation. He must also give his consent to the great family ceremony organized around the sister’s children’s ear-piercing (gûtûra matû), as well as to the initiation of his sister’s sons and the marriage of her daughters. The goat he offers is a symbol of his consent and desire to support what his sister is passing on. A muntuetû relationship that is coming to an end without difficulties is proof of the benevolence and quality of the muntuetû, and that the alliance forged three generations previously was fruitful. A girl from this side can be taken once again. She comes from another territorial division, from the outside, which is good, but from a place that has proven its worth, which is reassuring. This ideal geometry also explains the rule whereby it is good, though not obligatory, for a man to take a wife from the clan where his mother comes from, or from clans where his grandmothers came from (the other meaning of the term ndûgû), but obviously not from within these women’s mûciî of origin. Relationships are good with these people and they are expected to remain so. Marrying descendants of matrilateral cross cousins is quite reminiscent of the marriage with MBD from the kinship theory, in that it traces a field of oriented relationships and stems from an ideal model defined by successful relationships between givers and takers that must be developed and maintained. Thus the muntuetû relationships are ever present: as far as we can see, there will always be a pair of siblings of opposite sex joined via the relationship with their mother, herself one part of a pair of siblings of opposite sex... a mise en abîme that is all the more striking since the children are themselves indebted to the muntuetû of a young wife, and since the gichiario of every giver are also considered givers.

Fig. 26 — The scattering of “taking” lines

Fig. 26 — The scattering of “taking” lines

Fig. 27 — The grouping effect created by the marriage of sisters’ children (6 and 8)

Fig. 27 — The grouping effect created by the marriage of sisters’ children (6 and 8)

41The comparison ends there because closure occurs only after three generations; it is not A who gives to B who gives to N...who gives to A—rather, it is A who after a set amount of time gives back to B. Other partners have joined in the exchange. The marriage can also involve a descendant of a patrilateral cross cousin, where B gives to A. With the MBD descendant, the taker remains taker and the giver remains giver; with the FZD descendant, the giver becomes taker and the taker becomes giver. The coexistence of the two formulas logically and sociologically resolves the question of the closure of the exchanges. Still, people express the FZD version less spontaneously, as if they were less aware of it, and only those with a good knowledge of kinship, thus of matrimonial strategies, are able to explain it. These latter say it is good for a girl to return to marry in the place her mother comes from. Yet again, the difference between strict marriage with the FZD and the ideal of a girl’s return to where her mother comes from should be emphasized. She follows the same path in the opposite direction, a return that is reflected in the rites of marriage in some families whereby a girl leaves her father’s mûciî walking backwards. When after three generations this result is achieved, the former giver becomes taker and vice versa—a permutation that momentarily neutralizes the tension between taker and giver and provides an opportunity to redistribute the cards.

42The importance given to the position of wife-giver also explains why marriage between children of sisters is possible, and, according to some, welcome since it involves the same muntuetû whose position and line of descent are reinforced through it: a single muntuetû organizes takers who gather rather than disperse, a form of accumulation and concentration of the qualities of muntuetû A. Figure 26 offers a diagram of all the debts owed by B, C, D and E to A due to the marriages that occurred in G 0. In fact these debts do not all carry the same weight. In G+1, B (for 1 and 2) and D (for 7 and 8) will not fail to make the ndûgû cattle payment to A. The relationship is active, and the bonds of alliance and coaffins are cherished; cooperation and mutual aid feature strongly in these bonds, all the more so since there are great chances that these men are of the same generational division as well as the same gaaru. In G+2, muntuetû A begins to distance himself, as his place is no longer of prime importance: B and D will pay what they owe to A (the muntuetû’s heir) with less urgency for the births of 3, 4, 9 and 10. They will delay in doing so, despite some children being named after these grandparents, but the muntuetû’s curses can have an effect. C and E, newly takers, will delay ad hoc payments even more for their 5, 6 and 11, 12 progeny; they will fall in line if the mugaa detects a curse coming from this side. However, given the genealogical and geographical distance, they can push back these payments and A’s heirs will lose the entitlement.

43By contrast, if the “children of sisters” marry (fig. 27), the number of lines of descent diminishes, the probability of seeing the payments increases, and the line that was initially giver is strengthened. Because of the union forged in G 0, A will profit from the ndûgû payments for children numbered 1 to 9. In G+2, a union—and a single one only because it would be a simultaneous exchange—can be made between the children of sisters C6 and D8, for instance. They have the same maternal grandmother; she sees some of her grandchildren marry among each other and thus maintain her line of A origin. This preferential marriage that occurs earlier—in G+2, whereas the other kinds of preferential unions are allowed only at G+3—accelerates the grouping of A descendants; moreover, it does not preclude a subsequent union and even reinforces the prospect of its occurrence. In G+3, A may marry C10 (a descendant of ndûgû yetû) who already carries in her the qualities of A. Their progeny in G+4 will carry even more: thus 12 gives rebirth to the grandfather’s grandfather, a recurrence that is that of the name of the Accomplished, and corresponds to the depth of the gîtiba cycle. Girl 11, for her part, gives rebirth in G+4 to the foremother in G 0, through whom everything arrived; in other words, the cionkoolo or “eldest mother”, borrowed from the kinship terminology based on generation level (see fig. 16); in this table the kiûjûûjûi who surpasses the “eldest mother” in G-5 can be interpreted as the ultimate reminder of the preeminence of the wife-giving line. Marriage between children of sisters thus strengthens the influence of the muntuetû and illustrates the advantage of being a wife-giver; it shows how to concentrate the muntuetû qualities, a bit in the manner of the so-called Arab marriage in the register of other parallel cousins. It may also carry the mark of low polygyny, in the sense that monogamy may somehow have encouraged marriage among children of sisters: while a man generally has only one wife, he may on the other hand have many sisters, whose matrimonial maneuvers open the possibility of preserving his line—as long as he is an astute and persevering maternal uncle. Thinking in terms not of lines of descent but of clans and territorial divisions, one has the surest method of organizing lasting exchanges among different sections: the less the sections involved grow in number, the better the chances of establishing links and forging durable paths.

44We can now understand why a muntuetû would see an advantage to marriages between sisters’ children. Yet the significance of the relationship does not end there. It is present from the very beginning of matrimonial transactions, when a bridegroom offers a specific gift to the brother and sister of his future bride. Moreover, spouses of two sisters call each other ûrûcia, since through these sisters they share a same mother (-cia). In fact, this term can be found at the level of the generation-set system. Men from the same generation set call each other ûrûcia: they have the same mother, the very one who ritually established their warrior gaaru, so they are like husbands of sisters to each other. Their children will be able to intermarry—another way of informing the unions within the generation-set system. Yet again, by means of union, and in this instance of coaffinity, an additional network between kinship and the generation-set system arises: the marriage of children of sisters combines the generational ideal, by which the children of one set are destined to marry among each other, and that of the preeminence of the wife giver.

4. First assessment

45The uthoni alliance and the ndûgû ties seem relatively more structuring, in that they are less ephemeral than the patrifiliation relationships forged within the mûciî. Whereas a family disappears with the death of its founder and nothing, apart from the rule of exogamy, keeps his memory alive, which slowly fades to blend into a patrilineal clan-category with blurred edges, the uthoni-ndûgû relationship establishes connections of a specific length and importance. The ties that result from the union and ndûgû payments are memorized and carefully maintained such that it is good to come together when things went well. Together with the gwato cattle contracts, the ndûgû payments form part of the relations that every father tries to forge and accumulate, and that heirs try to appropriate. They are more advantageous on the side of the wife-giver; a lack of girls prevents a man from building these alliance networks where the father, and then a son, are givers and thus beneficiaries. Moreover, cooperation and mutual aid occur more frequently with real and potential allies than with the agnates. Herding livestock is more readily entrusted to a brother-in-law or to a son-in-law than to a brother or even to a son. The ndûgû ties allow for establishing relationships with families and divisions scattered throughout the Nyambene hills, providing opportunities for bartering, clearing, and grazing, or emergency aid in the event of climate hazards. We can thus conclude that the ties of alliance and the ndûgû-muntuetû ties organize relationships more durably than do patrilineal ties.

46If we no longer think in terms of individual ties but of a more general model, an ideal as well as a practice emerges from the muntuetû relation, the ndûgû payments and the various forms of preferential marriage, in which, on the one hand, husband and wife come from different territorial sections, and on the other, it is good for regular currents to be established between same sections, regardless of whether one is giver or taker. This model partly explains the statistical realization issues we were confronted with. During an in-depth survey carried out in a territorial subsection of Athiru Gaitî, there appeared for the first time preferential marriages, particularly those where a wife came from the same clan or section as a mother or grandmother. Asked about these repetitions, the elderly began to offer explanations whose meaning became apparent only much later. To determine the relative importance of these unions, a survey was begun but was unable to be carried much further; people showed very little enthusiasm for answering questions and had clearly forgotten many things. Their amnesia might have been explained by an acculturation already begun in the 1930s and intensified by the land reform carried out as of the 1960s: among other upheavals, this one resulted in a partial reinforcement of the patrilineal ties and set off a race for unclaimed land among these societies. Simultaneously, the ndûgû, muntuetû and muthoni relations generally fell to the wayside, so much so that it is one aspect of the social crisis now affecting the Tigania-Igembe (see Part 4). We tend to think that though at one time the understanding of the precise relations seemed to be much better, a vagueness must have nevertheless always existed. Most important was that spouses came from different territorial sections: this was the first matrimonial rule stated by the elderly. The distance—regularly recreated and maintained through cattle exchanges—was sufficient to mark the preferential alliance: either a union was taken up with a giver or taker already known to one, or the way was opened for new exchanges. Indeed, we must not conclude from this development that preferential marriages were preponderant: if one was lucky enough to have a good muntuetû, it was wise to cultivate the relationship, but this was not often the case. Relations between muntuetû and mwonomware were too susceptible to poisoning to not be tempted to look for the new. However, the extension of the ndûgû category is such that an a priori new relation could always end up becoming a known relation. The varying levels of knowledge and even competence shown by the elderly in these matters of kinship suggest that some of them had a great command of these rules and strategies regarding them, while others practiced them with only a short-sighted view, a contrast that confirms what we already know about the capabilities of the mugambi that set them apart from the majority of their fellow citizens.

47In any event, we can see a society emerging here, made up of an enmeshment of relationships among givers and takers, separated geographically, and represented by the various transfers of cattle—the rûracio and the ndûgû—as well the gwato contracts, whose resemblance to these transfers is undeniable. If it had been possible to reconstitute and visualize enough ties, it would be apparent that the doubly-dispersed habitation of the Tigania-Igembe—as if spread out randomly amid the potentialities of the ecosystem of the Nyambene Hills—is in fact organized through a web of transactions already concluded or to come. In other words, the image of interconnected mûciî would have appeared, not through the patrilineal ties, but through these relationships among givers and takers. In short, with a territorialization congruent with the polysemy of the term mwiriga, as well as with the way the Tigania-Igembe liken their relationships to the circulation of cattle and to a bovine idiom, it is clear one must follow the cow (or the goat or the bull) and trace the path it takes. The extension of the term baite, and its use in salutations among those of the same territory, offer an apt description of the Tigania-Igembe: a group of givers and takers, or potential allies, in a matrimonial field coextensive with the generation-set system.

48Another matter, which does not explain the alliance though it sheds light on it, is the position of children, their belonging, and the modalities of filiation. The influence of the mother’s brother and of the more distant muntuetû can be seen in the number of payments that must be made at every birth, and in their capacity for blessing and cursing, to such a point that one wonders to whom the child actually belongs—the father or the muntuetû. At several points in his research (for example 1945, ms.: i, 1) Lambert cites the proverb Mwana nî wa ithe na muntuo (“The child belongs to the father and to the mother’s brother”). I found nothing as explicit in the Tigania-Igembe apart from such truisms as: “The child is the father’s and the mother’s,” an expression whose significance became clear only much later. Converging facts prompt the examination of this issue: the rounds of visits of sister’s children before their initiation, the modest bridewealth, the payments subsequent to every new child, the recipients of compensations for murder, the theory of health and illness, and logically, the cause of deaths. An individual who dies before having become a grandparent is deemed to have been a victim of mûgiro, for which the muntuetû is or are blamed. The corpse of the victim, deemed so dangerous that it is touched by no one, is taken to the bush by a pariah, and left there for the hyenas, likened like vultures to the mother’s brother; this corpse is subjected to a fate not unlike that of the foreskin of the circumcised. One must also try to understand what consanguinity is to the Tigania-Igembe, and what is transmitted from a sister and from a pair of siblings of opposite sex, another fundamental relationship that stands out in Meru kinship. Children of sisters may marry because their patrilineal sides are different, say the informants; their identical matrilineal sides are not an obstacle. On the contrary, it seems wise to reinforce them. Cross cousins may never marry but their grandchildren will be able to, and locals talk of the “return of ndûgû”: in other words, individuals whose patrilineal parts are no longer an obstacle while their matrilineal parts, almost identical, are nice to think and maintain. These varied facts become increasingly telling when related to a myth presented in the next chapter, and then to the generation-set system.

Fig. 28 — The family kraal and the cattle enclosure

Fig. 28 — The family kraal and the cattle enclosure

Ill. 34b — Lairangi North Tigania, 1989

Ill. 34b — Lairangi North Tigania, 1989

Photos by Anne-Marie Peatrik.


1  Following standard notation, and designating kin in relation to Ego: D = daughter; S = son; Z = sister; B = brother; F = father; M = mother; ch = child or children.

2  Lambert (1945, typescript: 272) reports that depigmentation (rwanga), leprosy (ndwanga or mutiri) and priapism were interpreted as symptoms of the transgression of a sexual taboo, or of murder, involving an agnate or a gichiario. He was also told that lovers in gichiario ran the risk of remaining stuck together like mating dogs, and that there was no way to separate them; they remained in that state until death ensued. This is a version, in an animalistic register, of the story of the lovers crucified by the side of the road.

3  There is another category of individuals in this same region, called nkio, “Those who have a tail”, but as far as I know, this particularity derives from another cause than an unwarranted mix of colors— perhaps the transgression of another taboo.

4  There is, however, another form of union against which everybody is powerless. A category of individuals, both men and women, known as girane, believed by some to have been conceived during the end-of-mourning rites of unispicious deads persons, have the particularity of causing their successive spouses to die, but they are so beautiful that they always find a new husband or wife; the girane effect disappears after the death of seven spouses. When asked about the existence of these girane, some people confess that the only time they have seen one is when passing by a woman on the road about whom it is whispered that she is girane; others affirm that whole families have this reputation, and the daughters can only get married by becoming the young wives of polygamous men, like those families, it is added, whose boys have the reputation of always getting injured on the front of their legs. The theme of the girane lies somewhere between a story to illustrate the dangers of uncontrolled desire and belief in a form of heredity—in this case negative—carried by individuals.

Table des illustrations

Titre Fig. 17 — Main terms of kinship: male Ego
Fichier image/jpeg, 37k
Titre Fig. 18 — Main terms of kinship: female Ego
Fichier image/jpeg, 49k
Titre Fig. 19 — Allies: male Ego
Légende * Cio. .: uses the name given to the spouse by her father, or even muka or mwekûrû.
Fichier image/jpeg, 19k
Titre Fig. 20 — Allies: female Ego
Légende * The prefix Ba- is followed by the name of the gift they exchanged** Mwuntû wa…: uses the gaaru name and then the nthukî name, or mûkûrû
Fichier image/jpeg, 15k
Titre Fig. 21a — Takers and givers in the establishment of the uthoni alliance
Fichier image/jpeg, 22k
Titre Fig. 21b — Takers and givers in the establishment of the uthoni alliance
Fichier image/jpeg, 57k
Titre Fig. 22 — Ndûgû payments received by the giver
Fichier image/jpeg, 13k
Titre Fig. 23 — Ndûgû payments owed by the taker
Fichier image/jpeg, 18k
Légende (=) very marked avoidance(-) marked avoidance(+) marriage possible
Fichier image/jpeg, 13k
Titre Fig. 25 — Cattle links, kin terms, and preferential marriages
Fichier image/jpeg, 21k
Titre Fig. 26 — The scattering of “taking” lines
Fichier image/jpeg, 10k
Titre Fig. 27 — The grouping effect created by the marriage of sisters’ children (6 and 8)
Fichier image/jpeg, 13k
Titre Fig. 28 — The family kraal and the cattle enclosure
Fichier image/jpeg, 59k
Titre Ill. 34b — Lairangi North Tigania, 1989
Crédits Photos by Anne-Marie Peatrik.
Fichier image/jpeg, 300k

© Africae, 2020

Licence OpenEdition Books


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search