M T KS BO.I IM* ItotM oa Matrlaoaial • • • •• aasaf tlw Atoamte Kwsaai A.X. Awedoba The Atoende or Toende* are those Kusasi who live between the Red and White Voltas. This area, once the western half of the Bawku District, is now part of the Zebilla District. Bawku still remains however the principal town to which Atoende H>«°Pl« gravitate. Those Kusasi living to the east of the White Volta are the Agolle, and to the west of the Red Volta live the Nabdams and other 'Frafra' peoples. Syne (n.d.) regards the Atoende as the ethnically 'true' Kusasi and he maintains that the Agolle were an accretion of various North-eastern peoples who had adopted the sobriquet 'Kusasi'. Agolle and Atoende have similar social institutions arid speak the same language. There are however differences both in cultural norms and language. The Upper East Region of Ghana constitutes a cultural zone, a fact which becomes obvious when Fortes' accounts of the Tallensi are compared to those of neighbouring people such as Frafra, Nabdams, Kusasis, Kasena-Nankana etc. However, one has to be careful not to assume similarities in norms and practices. The peoples themselves are conscious of ethnic differences between themselves and neighbouring peoples. In my survey of the Atoende area I heard often remarks to the effect that 'Only the Nabdams do this; we Kusasi do not1. But even within the same locality I heard repeated reference to a particular descent group, the Ataa Bisi, who were represented as holding onto different and peculiar practices and norms. Reaarks on Social Organisation and Kconoaty of Atoende Atoende people are predominantly farmers and grow cereals such as millet and guinea corn as well as legumes and vegetables. They have both compound farms located close to the homestead and bush farms. They rear livestock including cattle, donkeys, pigs, goats and sheep as well as poultry. Poultry and small stock can be found around the compound and the settlement but cattle and donkeys are less conspicuous, the former are usually consigned to Fulani herdmen who live outside the Kusasi settlement areas. The people are grouped in dispersed patri-clans but some of these are more concentrated in some parts than in others. The Gballis are associated with Tili and the northern parts, the Bims with Binaba and Biengu (now in Burkina Faso), the Kusaab with Kusanaba, the Gbingbina with Zongoirl (also called 37 'Kpombeya Nab') etc. Almost every settlement is mixed with a sprinkling of representatives of other clans and or non-Kusasi peoples such as the Bisa who migrated from southern Burkina Faso in living memory and still maintain contact with the Bisa homeland. A settlement, comprising scattered homesteads and farms and differentiated from neighbouring settlements by a stretch of bush land, has its chief or headman. He usually comes from one of the dominant descent groups in the settlement. The earthpriest is a member of one of the descent groups who first Bettled in the locality. Syme (n.d. and Rattray (1932) concur that the Kusasi did not have chiefs until the Hamprusi kings started sending their sons out to colonize parts of Agolle such as Bawku. Many of the present day chiefs of Atoende are however Kusasi people, some of whose ancestors had of their own accord solicited from the Hayiri, King of Mamprusi, the ritual and symbolic perquisites of chief ship. This happened in some cases in the present century at the time of British penetration of the area. Health in Atoende was and still is estimated in cattle and other livestock, in offspring, in many wives and in the size of farms. Most people idealise large families with many sons and daughters, preference being expressed for an equal distribution of both sexes. There is however, a growing realisation among educated people of the limitations of large families in the modern set up. Most men would marry as many wives as possible but older men expressed in personal interviews the desire to see younger dependants taking a fairer share of wives with the backing of joint family resources. I did not hear complaints that the old men were marrying all the women. Polygamy is encouraged by the broad section of Atoende people and this includes the women too. The latter justified polygamy on the grounds that co-wives make lighter the work load of the joint family falling to women. The high fungibility (to use Meillassoux's term) or circulation of women is suggested by the institution of child betrothals which is fairly common and which is also regarded as justifiable and defensible. Bridewealth transactions are perceived positively; Atoende people consider it proper and respectable to give and receive bridewe*lta* Failure to provide bridewealth when it is considered d»« it disreputable. The reputations and prestige of those affe*t*d, particularly the wife, the husband and the off- spring, arc dented when bridewealth payments are withheld. We h«ard acoftOntt of some prominent people who have suffered disgrace on that score and how sons have acted to r—e3y what their own fathers had failed to set right. Failure to pay the matrimonial goods does not lead automatically to the wife-giver's appropriation of the offspring to the marriage, a* is sometimes supposed. 38 A im of the Paper In this first part of the paper, I shall limit myself to those matrimonial goods I regard as constituting primary prestations. These may be regarded as prime prestations in Fortes' (1962b) terminology but I shall argue that though Fortes' distinction between prime and contingent prestations is a useful one the primary prestations to be described here differ in some details. I am of course aware that Fortes' Tallensi material may have influenced the setting up of the dual distinction in prestations and that Tallensi marriage practices are similiar though not identical with those of the Atoende Kusasi. I hasten to add that Hendosa (1973) has applied the Fortes' scheme to the Sisala marriage prestation, also an Upper Region people, and has expressed satisfaction with it. In my view, Mondosa appears not to emphasize certain aspects of Sisala marriage prestations which do not fit neatly into Fortes' scheme, such as their variability, refundability and negotiability. Marriage as a Prestation In the course of interviews the word da 'to buy/sell' cropped up a few times but in most Moore-Dagbani languages of the Upper East Region of Ghana the appropriate word for bridewealth prestation is sole/sol. This word applies to the transfer of those crucial matrimonial goods which serve in some ways as prime prestations. The verb sole is in no way synonymous with da from which is derived daa 'market*, da is a commercial transac- tion typical of the market place. Marriage and bridewealth are however perceived in terms of social prestation. In marriage, primary bridewealth goods are received by the wife-giver after 'he' has transferred a 'daughter' to the wife-taker. This is not an absolute transfer and alienation is partial and not total. In the character of the classic gift transactions analysed by Mauss (1954) the transactors exchange themselves though vicariously. Meillassoux (1981) argues that the bridewealth goods are like tokens though with a fiduciary value. I would argue that when the wife-giver transfers a daughter 'he' does not receive only material goods, 'he' receives also a 'son1. This view is not opposed to the emic view of marriage, in fact some informants have made this assertion. I should draw attention also to the vocabulary of affinity which also confirms this. The daughter's husband is addressed as 'son' and the son's wife as 'daughter'. Though in English the vocabulary of affinity appends 'in-law' to kinship terms as in 'brother-in-law', 'father-in-law' etc., to the English these affines mean more to a person than scae distant kin. The dividing line between affinity and kinship is not always rigorously drawn. In making the above remarks I bear in mind the three way distinction which Beattie (1964) makes between what the actors 39 do, what they My they do and what they My they ought to do. ideally Mar*!*?* effects tit* exchange of male and female persona, aai not only female persons *• vivas, but as Fortes (1949) has pointed oat affinal relations in practice subsvft* tensic *. Iron kinship relations which in his view are axiomatic subsume tensions too; one only has to cast the mind back to the Oedipus complexes, which Fortes (1959) has analysed very well. Kin groups are enriched through marriage irrespective of whether they are wife-givers or wife-takers because they share personnel of opposite sexes. There is however a difference in the degree and manner in which the spouses are shared. The woaan goes off to live with the husband since virilocal residence is the rule, but the husband remains with his lineage kin. The offspring of the marriage too, it is true, have more kinship with the father's group than with the mother's group. Their attachment to the latter group cannot be denied however and they have duties and obligations there. But in spite of virilocal residence a married daughter's status in her natal lineage does not lapse even with time and it is only on her death when her lineage permits her husband's lineage to perform her burial and final funeral rites that she becomes fully incorporated in the husband's lineage's community of the dead. A married woman will continue to be governed by the ritual prohibitions of her lineage until her death. No amount of matrimonial prestation can change that. The wife-giver does not • have the privilege of incorporating the daughter's husband in the wife-giving lineage or of compelling his residence there, but they nevertheless acquire filial rights to him. They assert this claim in various ways and on certain occasions, principally at mortuary and funerary occasions. As a 'son', they expect him to show respect to them. This may be expressed in his concern for their material well- being and their health. If they should require farm help he should provide it and they need not always ask for it. He should visit them from time to time. He should provide them with loans if they ask for them and a variety of other services. A daughter by herself lacks the means (and the physical strength) to provide these services.2 in any case her ability to provide these benefits does not exempt her husband. Be is expected to get actively involved in the provisioning of burials and funerals when a key affine dies. Be should dig the grave and remove all obstacles such as rocks; he should provide musical entertainment to glorify the deceased who is being buried. The former service is due from the deceased's kinsmen. Bis lineage are however distraught with grief and cannot provide the labour necessary, for the lineage participates in the death of its member. The burial specialists are either too old to over exert themselves or are less concerned. It falls therefore to the 40 son-in-law to dig the grave and ensure that nothing prevents the grave being dag deep enough. He is a 'son1, and he is expected by filial piety to bury the wife's parent. At the same time he is a stranger and is less constrained by grief. He does not participate to the same degree in the death of his parent-in-law. Atoende informants assert that a family begetting daughters does not feel that all is lost, for daughters are exchanged for 'sons'. La Fontaine (1962, p.98-99) writes that in Gisu terms "pay- ment of bridewealth in full can never entirely wipe out what is owned to a father-in-law [who] ... has given the means by which a household will be established and a line of sons and grandsons founded to perpetuate his son-in-law's name". This view would indeed make sense to the Atoende people. The transfer of a daughter to the wife-taker establishes an obligation which is not discharged by the counter transfer of marriage goods. To completely neutralise this obligation one must transfer another woman. The Agolle are said to do this when sisters are exchanged but the Atoende do not recognise it. The next best thing is the exchange of persons of opposite sexes - the couple who marry. She as a wife becomes a daughter-in-law while the husband becomes a son-in-law. The patrilineal mode of recruitment of personnel and residence requires that he should not be trans- ferred physically or recruited literally. The daughter's husband however becomes the non-resident 'son' who pays occasional visits to his affines and steps in at critical moments to perform those services which his affines are unable to perform. In his role as a son-in-law, a man has duties defined for him by customary usage, which often shade into the grey area where personal discretion is exercised. A daughter's husband has the opportunity to be a good son-in-law or a bad one. If he excels himself and impresses his affines favourably he might be rewarded by the 'gift' of a junior 'sister' of his wife to wed. The characteristics of gift-exchange are present in Kusasi marriage. The gift of a wife is not entirely free for it elicits reciprocities of which the formal bridewealth is only one aspect. There is also an implicit obligation to make a counter gift in the same currency and that gift is a prestation of the groom. This, he can effect by making himself available to his affines i.e. by being a 'son' to them. He will try to execute this role as best he can as long as he lives. He puts his affines 'above' him. Atoende Kusasi say 'A poor man's chief is his father-in-law'. Respect due to the affines may be expressed in the language of material prestations. For the poor man without the material means of symbolising respect, the father-in-law becomes almost a chief to whom one is a subject: literally at his beck and call. This relationship of inequality, whether blatant or latent, also denies any possible equivalance 41 between W O MB and Material prestations. we could conclude that possession of cattle per se does not enhance eligibility for marriage in a man.3 This is also confirmed by informants. Prim* Pnstations and Contingent Prestations The distinction between 'Prime Prestations' and 'Contingent Prestations1 was made by Fortes (1962). Portes saw one of the functions of bridewealth as maintaining the shift from kinship status to affinal status - more specifically the shift from non-kinship to affinal status. Prime prestations correspond to what he terms capital value in marriage i.e. rights in the bride's sexual and procreative capacities and the domestic services that go with them. The prime prestations are specific. They serve as a jural instrument and the sole means by which a lawful marriage is constituted. They are not open to bargain and when divorce occurs they are returnable. Contingent Prestations correspond to ancillary values realized in marriage; and through them the bride's family are persuaded to relinquish some of their residual rights in her. Contingent Prestations may be open to bargain and an element of barter may enter into the transaction. They might be counter-balanced by reciprocal gifts and services. The distinction is no doubt a valuable one which throws light on the meaning and significance of bridewealth, and could be applied to Atoende Kusasi notions about bridewealth with some success. For them the prime prestations would include cattle, goats, sheep and poultry. Some informants suggest that in former times, donkeys were included. The Kusaal word sole refers in Atoende dialect specifically to the transfer of these items. Cattle however occupy the centre stage in sol* transactions. Their importance finds expression in the following statement - Fn m pn sul ni-i fu pu nyebid pen neln - (If you do not trans- fer cattle as bridewealth you will not find a woman to sleep with). The emphasis placed on cattle is due to the fact that they are the most expensive by far of the bridewealth items and the single item most likely to weigh heavily in marriage plans. The Prime prestation is more or less standard and comprises four head of cattle, at least two of which must be female (cows) but one of which may be an ox. The ox, or at least one of the oxen, is sacrificed. Cows are not usually killed but are either used immediately as bridewealth goods on behalf of a member of the lineage who has contracted a marriage, or they may be reared with a view to increasing the family's herd. In addition to the cattle the bridewealth includes three goats and one sheep as well as three fowls. One of the fowls is presented to the person who has acted as marriage mediator. The cows are referred to symbolically as 'horns' and it is expected that the animals tran«ferred should not be calves, 42 ( *. ? j without horns. It is not clear if emanciated animals would be rejected, but it is conceivable that a difficult father-in-law who does not approve of the marriage night do so. Informants say that in former times a good wife-giver who had sympathy for, or was well disposed towards the wife-taker might accept a donkey and three cows plus the additional items, a donkey being of lower value than a cow. Two women informants admitted that their husbands included a donkey in the bridewealth transferred on their account. Host informants also said that an extra cow or two might be taken in addition to the standard bridewealth. The usual explanation of this was that if a man had to transfer 5 cows instead of 4 then he would be entitled to claim 5 cows on the marriage of his eldest daughter. When pressed further to say why, given the customary standard, anyone would transfer more than the fixed number of cows, a number of explanations were given. (1) A man might volunteer to give more than the minimum number of cows in order to appease a difficult father-in-law who was bent on withdrawing his daughter. (2) If the daughter were to have been betrothed in infancy to a friend or afflne of the family her people would surely oppose her marriage by elopement to another. The eloper might employ all means at his disposal to induce the wife-giver to concede, including the offer to transfer an additional cow to her family. Such an offer might not be accepted. (3) A groom and his family might on their own accord, without any pressure from the wife-giver, offer an additional cow to their affines as a token of love for a beautiful woman or in appreciation of a wife who gives total satisfaction to her husband and his family.4 An informant indicated that although this gesture might appear to communicate gratitude and satisfaction yet it might not be devoid of self- interest. A beautiful wife tempts others but the extra prestation provided on her behalf could deter anybody who might otherwise be tempted to elope wih her, for any future eloper would have to make restitution in full. (4) An informant also suggested that the extra prestation might come about through loans given to the bride's family which have remained unpaid. Cattle loaned out to her family might eventually be counted as constituting part of her bridewealth and claimed when her daughters marry. There is thus, an awareness that more could be taken from the bridegroom than is usually claimed. Variation in the bridewealth is discussed in terms of additional cattle claimed or given. There was no mention of sheep, goats or poultry given in excess of the standard minimum. This may be because the other items - goats, poultry etc are not considered to be worth bothering about, cattle being the main item and also the more prestigious, since they contribute more to the definition of wealth than the rest of the goods put 43 together. Gift of additional goats and sheep occurs in other contexts and situations external to marriage transactions. The standard bridewealth say be paid in bulk or by instalments. Atoende people insist that when it is demanded and paid, and the duration of payment, all depend on the wife-giver. If he insists on immediate payment then the bridewealth will have to be transferred immediately. Some informants suggested that in former times the groom was not required to transfer the bridewealth immediately, and that it might take several months before the demand was made. The daughter would be consulted on her visit home and if she expressed satisfaction with her married life and saw prospects of its continuance, then her parents would approach the wife-taker and demand the bride- wealth. Other informants suggest that in former times the bride- wealth would be claimed only after the bride had given birth to an issue for her husband. A reasonable wife-giver might grant a temporary reprieve or allow transfers by instalment. In the view of most informants, marriages that originated in child betrothals were more likely to have been settled by transferring bridewealth by instalment, and informants gave this as one of the benefits of child betrothal marriages. They claimed that it benefited the groom because he is given plenty of time within which to come up with the bridewealth goods. The wife-giver also benefits as he claims a portion of the bridewealth long before the daughter has attained the marriage age. Atoende informants give the impression that nearly every wife-taker transfers the bridewealth eventually, but they agree that a wife might die before her bridewealth has been transferred. Atoende people are not unanimous however about whether under such circumstances the bidewealth would still be paid. Informants agree that it would have to be paid if she bore children for her husband. Some assert that if she died childless no bridewealth would be claimed. In making this assertion infor- mants often compared their system to the Habdan practice which they said they understood to require the wife-taker to transfer the bridewealth at all cost and in all cases including these circumstances where death had intervened. Others said that the wife-giver could and might still demand bridewealth for a married daughter who died childless, wife's father or his representative could compel the wife-taker to transfer the bridewealth. Her lineage by birth could deny her husband's lineage permission to bury the corpse or to celebrate the deceased woman's funeral as a means of coercing the recalcitrant affines. Informants on the other hand accept that a dutiful husband might, on his own volition, transfer the bridewealth on behalf of the deceased; they add also that such a man would almost surely be rewarded by betrothing a girl to him or allowing him to marry a sister of the deceased. 44 I I irithholdiaf of burial asd funeral performance might appear M • carious sanction employed to get a widower to settle the bridwwealth. After all, the burial and funeral rites entail a considerable outlay of resources for the widower and his family. Self interest should dictate that he cut his losses and not bother to claim the rites but Atoende people do not see the matter in this way. An informant explained the value of the sanction by suggesting that if the deceased's funeral were left unperformed her spirit would seek vengeance on her husband and his family. It would appear that fear of superhuman retribution triggered by the withdrawal of the funeral permission compels her husband and his people to concede to the demands of the wife-giver, even if the feeling is that those demands are unjustified. As most informants indicate, bridewealth is not always demanded where a wife has died childless, yet in spite of non- payment her husband and his family would usually be permitted to bury her and organise her funeral obsequies. These would however last three days instead of the 12 days usually required. People of both sexes are usually denied full funeral obsequies if they have never had issue.5 Nevertheless, a deceased wife may still tentatively be incorporated in her husband's lineage in spite of the fact that she has never had issue and her bridewealth might never have been paid nor would ever be paid. The conclusion to be drawn from this is that prime bridewealth prestation might establish marriage and account eventually for incorporating a wife in her husband's patrilineage, but in some cases she might be incorporated without any bridewealth having been transferred. The implication here is that there are exceptions to the use of prime bridewealth payment to define marriage. Responsibility for Provision of Bridewealth The responsibility for providing the bridewealth falls to the head of the minimal lineage - usually a grandfather, a father, father's brother or a senior brother. A person can depend on the lineage to provide the marriage goods necessary for his first marriage. In very critical cases another related minimal lineage may be approached for a loan, or even a clansman who is wealthy enough might step in to rescue the beleaguered groom and his kin. Informants agree that the maternal uncle may provide nelp where the situation is critical. The cattle so provided are treated as loans and must be repaid. If the sister's son fails to do so his maternal kinsmen could 'claim' his children, or the bridewealth paid when his daughter's husband provides any in future. An individual might contribute towards his wife's bridewealth. The head of family would however make the transfer, treating what is individual property as lineage property. For subsequent 45 marri««M tlM H •»•••'• Wmmttom dlwi»i«tes. •*•*• tbmrm it ft* pressure on its m o v n w ir ft would settle bridewealth obliga- tions on • polygynist's behalf but it i« recognised that other lineage •••tr*s have an equal claim to joint resources such as livestock. Therefore care would be taken to ensure that other male members who are single do not suffer unduly. Responses obtained from MM Atoeode senior people suggest that they are for equitable 'distribution of family resources in this way but •ore intensive studies would be required to confirm this view. Divorce amd Kefuad of Prime Prestations Atoende Kusasi say that bridewealth cattle are refunded when divorce occurs. Many informants insisted that husbands do not divorce wives, only women divorce their husbands. Instant divorce occurs and either partner might be responsible for it. Some ritual acts were listed which conduce to instant divorce. A husband might divorce his wife by beating her with the bow [some say the quiver] or by breaking her hearth stones - the three mud pillars on which she cooks. A wife signals her intention to divorce her husband when she kicks him at the moment when he is about to engage in sexual relations. She could also do so by beating him with the stirring stick.6 These acts are symbolic and constitute a negation of the conjugal basis of marriage. The hearth stones symbolise her role as a provider of cooked food; the conversion of what is raw and unedible into food. The division of labour in the society oampels a man to depend on women for cooked food. To break a wife's hearth stones is to symbolise her rejection. The importance of a woman's culinary role is emphasised in the burial context when she is buried facing sunset, the common practice in the Upper East. The exegesis explains that even in her grave she watches for signs of sunset in order to start cooking the principal meal of the day. A husband and his male agnates have the duty of defending the women of the lineage, the old and the young. The principal instrument of defence is the bow and the arrows which the quiver contains. These are used against enemies whether human beings or wild animals. They are also used to cull the wild and provide food for the family. To apply the bow or the quiver as instruments of battery against a wife is to symbolise that she has lost the protection of her husband and his agnates. It is tantamount to re-establishing her former status as non-kin and potential enemy, thereby making it impossible for her to remain as a wife. The stirring stick likewise is the instrument for converting raw flour into food (thick porridge, which is the principal diet in Kusasi), and to turn it into an instrument of battery is to subvert its use. The act evokes the fear that food, an item of 46 consumption and eventually the means of reproduction, can be transformed into a poison, an item of destruction. The lineage could not accept the implication that it could be harbouring so dangerous an enemy. A wife kicking a husband intentionally at the moment of sexual intercourse symbolises not only instant refusal but total contempt for the husband and a traduction of his manhood. By so doing she denies him the means of reproducing the personnel of the lineage. The consequence of instant divorce is that she must return or be returned to the wife-giver with the warning that on her death her funeral rites would not be claimed by her former husband. Bridewealth restitution might not follow immediately but she would henceforth be nobody's wife and therefore eligible for remarriage. The divorced husband will receive his animals back when the new husband has refunded her bridewealth, and not before. In other cases divorce might follow prolonged separation. After she has returned to her parents after a quarrel, her husband should indicate his continued interest in her by paying her parents occasional visits to try to talk her into returning. When she elopes with another man she is seen as having initiated divorce, and it becomes established as such when the new husband makes complete restitution of the bridewealth transferred by her former husband. The wife-giver does not usually refund bridewealth to a former wife-taker, he acts as a clearing house, but he is not a passive factor. He must exert pressure on the second husband to make the refund immediately and equal to the initial bridewealth. When the animals are available the former wife-taker is invited to collect his cattle. The previous wife-taker may or may not receive all his animals back, fie receives a full refund of cattle only where the divorced wife had given him no issue. Where the woman has had four or more issues by her previous husband, the divorced husband claims nothing. The wife-giver therefore takes a new set of bridewealth in addition to the old set. Prise Prestations and/or Contingent Prestations The primary bridewealth prestations comprising cattle, smaller animals and poultry which we have discussed above conform in some essentials to Fortes' notion of prime prestations. Together they constitute a pledge and their transfer does indeed serve to distinguish generally between mere co-habitation of a couple and legitimate marriage. They also conform to prime prestation in that they are refundable when marriage is dissolved while at the same time differing from them since there are circumstances under which only a part refund or even none at all is possible. 47 are not goods in question The matrimonial therefore 'bridewealth' in the literal sense. But since children are the main justification for their retention after the dissolution of marriage they might also be perceived as • childwealth'. This view plausible as it may be is weakened by the fact that the matrimonial goods may be claimed before childbirth but if a woman dies without issue after the wife-giver had taken bridewealth no refund can be expected. Secondly, if a wife pre-deceased the transfer of the primary matrimonial goods her parents could still demand bridewealth from the wife-taker and are sure to do so if she is survived by at least one child. Ideally, the wife-giver should demand the transfer of matrimonial goods after the birth of the first child which strengthens the view that we are dealing with child-wealth. It must however be pointed out that even then where the matrimonial goods are transferred in instalments these are not made to coincide with births. The primary bridewealth prestations differ from the paradigm of prime prestations in a number of other respects. The former are viable in quantity but the latter are fixed in quantity and therefore conform with Meillassoux's view that bridewealth tends towards uniformity, (op. cit. p.64). The evidence provided for Atoende Kusasi is that three to five cattle might be taken and that sometimes a donkey is accepted in place of one of the cows. In this regard, the specificity of Atoende primary bridewealth prestation is an aspect of the collective representation, and to regard this prestation as fixed in quantity is to ignore the reality aspect of their social experience.^ Given that four cattle are the standard number of cows transferrable although real life experience shows that more or less may actually be given to affines, one may wonder whether any number in excess of four head of cattle is not to be regarded as contingent prestation. Where the bride-taker has voluntarily added extra cows which the bride-giver has not demanded the character of the additional goods inclines towards a contingent prestation. Informants have said the giver sometimes wished to induce the wife-giver to drop whatever objections and misgivings he might have had initially. Furthermore, the extra cows might originate outside the context of matrimonial prestations as could happen when a son-in-law lends to a needy father-in-law. The contingency argument is however counteracted by the fact that the wife-giver might also explicitly demand bridewealth in excess of the standard and that this might be the condition for the marriage to be sanctioned. An informant related how a kinsman of his once succeeded in claiming more than the standard number of cows from his Nabdaa son-in-law. In this case however the excess cattle had to be returned to the wife-taker as the head of the wife-giving 48 lineage had countermanded it. It would appear here that the norm has been invoked against the individual. Where a wife-giver wishes to take more than the ideal standard of bridewealth cattle he conveys this information symbolically by taking along with him on his visit the appropriate number of cattle ropes representing the number of cows demanded. are not Contingent prestations The extra animals taken by a wife-giver fail however to meet one characteristic of contingent prestations viz. their refundability. directly refundable although they are off-set wholly or partially by counter-prestations according to the Fortesian definition. The extra cows in question are not usually off-set by counter- prestations in the Kusasi case and they must be refunded together with the other primary matrimonial goods should divorce occur. In other words they are treated as part of the primary or prime prestation. As I have pointed out above, some informants held the view that the voluntary prestation of additional cattle is sometimes motivated by the awareness that they would be treated as prime prestations. Whatever their origin the extra cattle become a prime prestation for the givers and the receivers. They are accountable when divorce occurs. The second husband must refund the whole primary prestation including the extra cattle received. Furthermore, the giver of extra cattle is required to demand an equal amount of matrimonial goods from the husband of his eldest daughter (the girl for whose mother extra cattle had been given). There are thus good reasons why the extra cattle should be viewed as prime prestation rather than contingent prestation. The prime prestation among the Atoende Kusasi is thus not fixed but is variable and although informants may say that it is fixed and standardized actual behaviour proves otherwise. Prime prestations are not subject to bargaining, unlike contingent prestations according to Fortes (op. cit.), but Atoende Kusasi bridewealth prestations are not devoid of negotiation and even of bargaining* (1) The wife-giver and the wife-taker have to agree on whether the transfer of the primary matrimonial goods should be immediate or should be delayed for some time to allow the wife-taker a temporary reprieve. (2) It has to be determined on consultation whether the transfer should be In bulk or by instalments. This involves bargaining although the wife-giver would have the final decision. In our interviews informants repeatedly referred to the disposition of the wife-giver as a factor influential in the transference of matrimonial goods. It is because negotiation and bargaining are involved in these prestations that it becomes necessary to characterise the wife-giver as 'hard1 or 'soft1. (3) It was pointed out that some wife-givers might reject poor quality livestock. Even with the Sisala, Mendosa's (op. cit. p.27) case 49 study shows that the wife-givers might argue about any sub-quality animals their affines night try to palm off on them and might insist that something extra be thrown in to off-set their inadequacies. On occasions of formal bridewealth presta- tions wife-givers are generally noted in the Upper East for their fastidiousness. This attitude, I have argued in the case of the Kasena-Nankana, symbolises the wife-giver's reluctance to accept matrimonial goods in exchange for the hand of the daughter.8 Because the prime prestation is not deemed equivalent to the girl who is being taken as a wife, it becomes a gross insult to provide goods that are not of the best quality. The wife-giver must therefore object to this, yet what constitutes quality must needs vary from person to person and reflect expec- tations of groups and individuals. The Atoende Kusasi facts reveal that bargaining is more likely to occur with the prime prestations than with the contingent prestations, which is the opposite of what the Portesian defini- tion leads us to expect. Fortes' notion of the operation of a •two-person zero-sura game' which he says is characteristic of contingent prestations is applicable to the context of primary bridewealth prestations among the Kusasi. On those occasions the wife-giver and wife-taker meet as opposed groups, a fact which is sometimes re-enacted by dramatic displays of hostility. Since the matrimonial goods are not mere tokens but wealth in substance the two-person zero-sum game that the groups can be said to engage in on this occasion is aimed at maximization of options and value. It is no doubt in the interest of the wife-giver to receive the whole of the primary matrimonial goods as soon as possible. This is because bridewealth in the form of livestock depreciates when delayed but multiplies naturally when collected immediately and reared for several years. Delay in payment is in the wife-taker's favour, but immediate payment is in the wife-giver's favour and in the wife-taker's disfavour. This can be illustrated further. We know that bridewealth once transferred is not directly refundable by the wife-giver and should the wife turn out to be barren or to predecease the birth of children the bridewealth would have to be written off. Where divorce has occurred after the prestation the previous wife-taker's matrimonial assets are locked up and he has to wait until such time that the ex-wife has remarried and the wife-giver has managed to persuade the new husband to refund the whole of the bridewealth already transferred. The stakes it can therefore be sees are not negligible for the parties involved in the matrimonial game. There are undeniably opportunities for maximisation if the participants so wish. A less sympathetic wife-giver would threaten to withdraw his daughter unless the wife-giver is prepared to settle sooner than later. A wife-taker for his part negotiates for more time, i 50 might add however that generosity is not unknown in these matters. A ,generous father-in-law may be rewarded through contingent jprestations and a generous son-in-law through betrothal of an additional daughter. Individual participants vary in their attitudes and references in informants' statements to 'difficult' and 'sympathetic' wife-givers show that marriages are not transacted solely by egoistic or altruistic groups. Formal kin groups or lineages would make concessions to wife-takers on primary matrimonial prestations because they stand to benefit much more when affinity acquires a kinship character with marriage. The history of individual marriages provides a clue to the attitudes of affines in the negotiation of marriages. Marriages initiated by elopement are more likely to provide an arena for the exercise of maximization tendencies than those based on betrothals. The latter draw attention to what Sahlins (1972) terms 'generalised reciprocity*. They tend to suppress ulterior motives as the respective parties have reached a stage where solidarity is stressed over and above the absence of direct kinship ties which has made it possible to contemplate marriage in the first place. Clan exogamy cannot be set aside. Elopement marriages by contrast incline towards Sahlins' (op. cit.) •negative reciprocity1, the mode of getting something for nothing, so to speak. Betrothal marriage implies a prior investment of some sort. But elopement marriage in the case of the Kusasi, though not for all elopements, implies that the groom has hitherto made no substantial investment in the bride to the benefit of the wife-giver, whose immediate permission has not been acquired.9 A wife-giver might nonetheless condone in an elopement, in any case an elopement marriage is not uncustomary. Elopement however plays into the hands of the wife-giver. His attitude becomes one of 'take it or leave it*. The groom and his kin group acknowledge their position and must smoothen the edge of their act by sending a trusted emissary with tobacco to beg for pardon: ngme ana (beat the palms). Conclusion The study of Atoende primary matrimonial prestations shows that the Fortesian definition of prime and contingent prestations does not apply absolutely and in any case there are grey areas where contingent and prime prestations cannot be clearly differentiated. In Part II, I shall discuss Atoende secondary or contingent prestations. 51 1. The paper is one of three papers on the subject and is I t Md on three weeks field study in 1987 supplemented by flfdftmentary sources. Questionnaires were administered at Binaba, Vttsanaba and Zongoiri jointly by the author and Mr. Bugri Manchinaba of the Institute of African Studies who is himself a Kusaaga though from Bawku in Eastern Kusasiland. I am very grateful for his selfless devotion to the project. Dr. G.P. Hagan, my one-time tutor, also of the Institute offered invaluable assistance in drawing up the questionnaire and commented on ftn earlier version of this paper. I am very grateful for his support. The then Director of the Institute in his sympathy authorised the funding of the project at a time when funds were scarce. I am grateful for his kindness. My gratitude also goes to Mr. A.J. Osei Bonsu for typing the manuscripts. Errors and infelicities found in this paper are solely mine and have nothing to do with the above gentlemen. 2. Some women who set up as millet beer traders do accumulate wealth and can provide financial help to their kinsmen. Such wealthy women would be in a position to procure hired labour for their parents, when this becomes really necessary, or to solicit for voluntary labour from their male clients. 3. One is reminded of Tiv spheres of Exchange. Wealth was classified by Tiv into hierarchicized categories or spheres. Exchange involved intra-sphere transfers or 'translations1, which, was easier to achieve than, inter-sphere transactions, c.f Bohannan (1955). 4. Rattray (19321388) notes that the number of cows given varids from two to five according to the amount of competition among suitors for a girl's hand. 5. The curtailed funeral rites might point to a diminished ancestral status and might be an exorcism of those who have not acquired full personhood. 6. Many of these prohibitions apply to neighbouring peoples such as Tallensi, c.f. Fortes (1949x108), Prafra and KasenarNankana. Their symbolism has been explored in Fortes (op. cit.) 7. Mendosa (op. cit.) shows that the bajarikiaa or prime prestation is a cow when paid pre-morfeem but a sheep is transferred when the woman dies before the payment. 8. Cf. Awedoba (1985)» 9. Among the Kasena^Nankana the suitor must send pre-marriage gifts of guinea fowls, tobacco, kola, salt etc. and depending on tne circumstances the outlay on such prestations could amount to a small fortune. Elopement does not therefore imply absence of investment as in the Kusasi case, which if I understand right, does not involve pre-marriage prestations. 52 Awsdoba, A.K. 1965. Aspects of Ifealtii and fiocfaange nit* tx> the StaeBe-Sukam of Ghana* Unpubl. D.Phil. Thesis, Uhiversity of Oxford. Beattie, John 1964. Other QOtares. Cohen and West, London. Bohannan, P. 1955. Some Principles of Echange and Investment among the Tiv, Aaerlcan Anthropologist 57 No.l, Part 1. Comaroff, J.L. (e3,) 1980. The Meaning of Marriage Bapnents. Academic Press, London. Fortes, M. 1949. The Web of Kinship awcog the Tallensi. Oxford Uhiversity Press, London, Fortes, M. 1959. Oedipus and Job in Ifest African Religion, Cambridge Uhiversity Press, Cambridge. Fortes, M. 1962. Preface to Marriage in iteUsal Societies (ed.) Fortes, M. Cambridge University Press, London. Goody, J. 1973. Bridewealth and Dowry in Africa and Eurasia, in Bridewealth and Dowry (eds.) Goody, J. and Taombiah, S.J. Cambridge University Press, London. La Fontaine, J. 1962. Gisu Marriage and Affiral Relations, in Fortes, M. (edo). Marriage in Tribal Societies. Cambridge University Press, Cambridge. Mauss, M. 1954. The Gift (Translated: I.G. Omnison), Cohen and West, England. Meillassoux, C. 1981. Maidens, Meal and Money. Cambridge Univer- sity Press, Cambridge. Mendonsa (Mendosa sic.) E.L. 1973. Aspects of Sisala Marriage Prestation in Research Review (University of Ghana), Vol.9, no.3, pp.13-47. Pattray, R..S- 1932. the Tribes of the Ashanti HintadJand, Vol. II, Oxford, Clarendon Press, Sahlins, M. 1972. StoneAgeBococBd.es, Tarvdstock. Syme, J.K. (n,d.) Ihe Kusasi, A aicrt Eistory. 53