Читать книгу The Metamorphoses of Kinship - Maurice Godelier - Страница 10

CHAPTER FOUR

Alliance and Residence

(Second and Third Components)

ALLIANCE: FORMS AND RULES

With the analysis of alliances and marriage we enter another zone of scientific turbulence induced by the publication, in 1949, of Claude Lévi-Strauss’ Structures élémentaires de la parenté. Taking the opposite stand from Radcliffe-Brown’s then-dominant thesis, which maintained that descent was the essence of kinship, Lévi-Strauss asserted that, on the contrary, kinship was based on alliance, since, owing to the universal taboo on incest, no descent group, no family, could perpetuate itself alone; all were compelled to contract alliances in order to reproduce themselves. And for Lévi-Strauss, alliance meant that men exchanged women from their own group, whom they were forbidden to marry.

In short, for Lévi-Strauss, if alliance takes precedence over descent, it is because kinship is basically exchange, and, more specifically, the exchange of women between and by groups of men.1

We will not go into the foundations of the incest taboo at this time – these will be dealt with later – but merely recall the explanation proposed by Lévi-Strauss in the following passage, which is more than a theory of kinship, since, according to the author, with the incest taboo we are dealing with the very origins of human society, with the decisive moment when humans extracted themselves from the state of nature and entered the state of culture:

As Tylor has shown almost a century ago, the ultimate explanation is probably that mankind has understood very early that, in order to free itself from a wild struggle for existence, it was confronted with the very simple choice of ‘either marrying-out or being killed-out’. The alternative was between biological families living in juxtaposition and endeavoring to remain closed, self-perpetuating units, over-ridden by their fears, hatreds, and ignorances, and the systematic establishment, through the incest prohibition, of links of intermarriage between them, thus succeeding to build, out of the artificial bonds of affinity, a true human society, despite, and even in contradiction with, the isolating influence of consanguinity.2

According to this approach, our distant ancestors did not originally live in society but in isolated biological families that perpetuated themselves through incest. The state of nature from which humankind had to extricate itself was that of animal-like promiscuity, of generalized incest. This had been Morgan’s vision. It was also that of Freud, in Totem and Taboo (1911), before going on to be adopted by Lévi-Strauss (1949).

EXCHANGE OF WOMEN, OR LÉVI-STRAUSS’ COUP DE FORCE

Let us take a closer look at Lévi-Strauss’ argument, which links the incest taboo, exogamy and the exchange of women by men in a single chain of cause and effect. He posits that the origin of the different marriage rules should be sought in the various forms that this exchange can take.

But this line of reasoning actually conceals a real coup de force. For, logically speaking, the incest taboo (and, I repeat, we will not discuss here whether or not it is universal), admits simultaneously of three possible forms of exchange. Either men exchange women among themselves, or women exchange men among themselves, or men and women leave their families to create new ones, and in this case it cannot be said that a brother exchanges his sister or a sister her brother for a spouse. But that is not all. The latter form is no longer a direct exchange of persons but a matter of reciprocal gifts of men and women between families.

Of course Lévi-Strauss was not unaware of the logical existence of these three possibilities, but he retained only one of them – the exchange of women by men – as the sole possibility that fitted reality. And he dismissed the other two as illusions that humans (women in particular) took pleasure in entertaining about themselves.

The female reader, who may be shocked to see womankind treated as a commodity submitted to transactions between male operators, can easily find comfort in the assurance that the rules of the game would remain unchanged should it be decided to consider the men as being exchanged by women’s groups. As a matter of fact, some very few societies, of a highly developed matrilineal type, have to a limited extent attempted to express things that way. And both sexes can be comforted from a still different (but in that case slightly more complicated) formulation of the game, whereby it would be said that consanguineous groups consisting of both men and women are engaged in exchanging together bonds of relationships [kinship].3

In point of fact, as we will see, there are societies where the women exchange men among themselves, and many more where families make each other mutual gifts of their sons and daughters. The rules are not the same in each case, even if the basic rule is always that of exchange. In all events, Lévi-Strauss posits male domination as the condition for the emergence of human kinship systems, and one that has continued to be so down to the present day. Male domination is a transhistoric, ontological fact that Lévi-Strauss links to the emergence of the human capacity for speech and symbolic thought. For, if he is to be believed, ‘the emergence of symbolic thought must have required that women, like words, should be things that were exchanged’.4

Caught up in this logic of the ontological character of male domination, one of Lévi-Strauss’ closest disciples, Françoise Héritier, would even try to show, in L’Exercice de la parenté,5 that all kinship terminologies – even those of the Crow type, usually associated with societies having matrilineal descent groups in which sisters treat their brothers like sons – bear the mark of the ‘differential value of the sexes’, which rates the man more highly than the woman, and the brother more highly than the sister.

Let me be clear. I do not deny that male domination exists; but unlike Claude Lévi-Strauss and Françoise Héritier, I do not think that it is a constituent principle of kinship. What is constituent, owing to the incest taboo, is the obligation to exchange. But exchanging women is not the universal condition for alliance and kinship. There is also the exchange of men by women, and we will give examples. As for the third possibility, it occurs every day in European and Euro-American cognatic societies, where sons and daughters leave their family to live with the one they have chosen, and no one says they have ‘exchanged’ their brothers or sisters with anyone. While male domination is a reality in numerous areas of social life in Europe, the United States and Canada, it does not play or no longer plays a role in the fact that people choose each other to form a married or cohabiting couple. Moreover, once they reach their legal majority, individuals have no need of their family’s permission to marry whomever they wish, with the exception of a small circle of blood relatives and close in-laws, who are forbidden by law. In addition to these taboos, which are the same for all citizens of the same country, there are other prohibitions specific to communities within the society, for example the prohibition for certain groups on marrying outside their own religion.

Let me say a few words about the exchange of men among women, which is an actual practice even if there are few examples. This is the rule among the Rhades of Vietnam, the Ata Tana’ai6 of the Flores Islands, the Tetum of central Timor, the Negeri Sambilan7 of Malaysia, the Nagovisi of Bougainville, the Makhuwa8 of Mozambique and a few other groups. All of these societies are matrilineal, and residence is either matrilocal or uxorilocal. Among the Rhades,9 it is the bride’s family that pays compensation to the family of the groom. In place of bridewealth, we have here groomwealth,10 made up of valuables, jugs, Chinese gongs and livestock. Among the Tetum,11 the men go from their mother’s house to that of their wife, which is built next to the ‘big house’ of the eldest woman of the lineage. It is she who keeps the lineage’s valuables, the cult objects (which are female) and the relics of the male members of the lineage. Each house is divided into two spaces: one inside, which holds the women’s quarters, the harvests, and the relics and cult objects; and the other outside, a sort of large platform in front of the house, which is also divided in two, with one side for the husbands and the other for the brothers of the women of the house. The households thus function as brother-exchange groups, the ideal being to repeat these bilateral exchanges from one generation to the next.

Among the Nagovisi,12 the women control their matrilineal land and act as stewards of its wealth. They play an important role in managing village life and, before the Europeans arrived and people converted to Christianity, women took part in the male initiations.

KINSHIP IS NOT UNIVERSALLY FOUNDED ON THE EXCHANGE OF WOMEN BY AND FOR MEN

The conclusion to be drawn from all this is that the formula: kinship is based on the exchange of women between and by men does not have the universal validity Lévi-Strauss and his disciples attributed to it. And therefore men’s domination of women is not the basis of kinship. Were this not the case, the feminist struggles of the past decades (which began in the West but have now spread far beyond) to move societies toward ever greater gender equality in the management of public affairs and private life, all these struggles that have brought our societies so far, would ultimately come up against the need to tear down kinship relations in order to achieve this equality. This is the criticism that Gayle Rubin made of Lévi-Strauss’ thesis early on in an article that was to have a considerable impact.13

Can we say then that, once rid of the abusive generalization that kinship is based on the exchange of women by and between men, once it is reduced to the proposition that kinship is based on various forms of exchange, Lévi-Strauss’ theory has an analytical basis? Roughly speaking, the answer is ‘yes’. But we must be aware that this amended theory no longer has anything to do with the first, since the exchanges that found kinship are now analyzed in a different and broader perspective. For even set on new empirical and analytical foundations, the thesis of kinship as exchange can never alone be the general theoretical basis for the anthropological analysis of kinship. It is necessary for analyzing forms of alliance and the marriage rules that express them. But it leaves aside or minimizes the importance of the forms and modes of descent reckoning. Systematically the inverse of Meyer Fortes (who saw alliance as a second and secondary aspect of kinship), Lévi-Strauss demonstrated a lack of interest throughout his work in analyzing the logics of descent, except at one later point when he rapidly explored the concept of the ‘house’. It was when studying the art of the Indians of the northwest coast of the United States and Canada, that, after having reconstructed the ties between the different types of masks and the Indians’ myths and rites,14 he tried to situate this corpus in the overall internal workings of societies whose complexity had somewhat disconcerted Boas.

Today claims to the hegemony of one thesis over the other, to the primacy of alliance over descent or vice versa, of one school over the other, are a thing of the past. Moreover, in Europe the institution of marriage is crumbling while descent stands fast. Our era is marked by a theoretical pragmatism that has nothing eclectic about it. It consists in solid knowledge of the theories, in refusing to simplify the facts under analysis to make them fit a given thesis, in knowing how to analyze the hypotheses found in highly divers and opposing theoretical approaches that have proven themselves capable of shedding light on certain facts, and in knowing how to combine them to explain other, even more complex facts without claiming to explain everything. It is in this context that we propose to examine a few forms of exchange and to scrutinize the very concept.

Let us restate the departure point of this examination. It seems well established that, owing to the incest taboo, the reproduction of kinship relations and the perpetuation of families and descent groups (where these exist) demand that individuals of both sexes look ‘outside’ for partners of the opposite sex with whom to establish the socially recognized forms of union that will ensure the reproduction of these relations and the perpetuation of these groups. This ‘outside’ can vary from very close to very distant, since the incest taboo can be limited to the closest consanguines and affines of a person’s birth family or extend beyond the boundaries of the lineage to all members of the clan, if it is truly exogamous, or even to cousins of the third degree, as in certain cognatic societies of Oceania –Tuamotu,15 the Cook Islands,16 Anuta,17 – or of Malaysia – among the Iban18 and their neighbours – and, finally, to the seventh degree of kindred (a restriction the Catholic Church attempted to impose on Western Christian societies between the eighth and the thirteenth centuries).

It should be remembered that marriage is one of the forms of socially recognized union, but it is not the only one, even if it is the most frequent. Moreover, when there are several forms of marriage in the same society, they never have the same status. Some may not imply any form of exchange, for example marriage with a captive.

Exchanges that seal a matrimonial alliance usually appear as reciprocal gifts. These gifts can take the form of either the exchange of a person for a person or the exchange of goods and services for a person. In the second case, a certain amount of wealth is collected by the groom’s family and transferred to the family of the bride (bridewealth) or assembled by the bride’s family and transferred to that of the groom (groomwealth). There can also, as among the Vezo, be two groups, one of which gives a son when the other gives them a daughter, the man and the woman being considered ‘equivalent’, and their children as mingling inseparably and indistinguishably within themselves what comes from their father and what comes from their mother.19

Finally, there are alliances without counter-gifts, as in India, which consist in giving a young virgin to the family of her future husband to which is added the obligation of a gift of wealth, a dowry. The dowry must be as large as possible to show the social status of the bride’s family and is often negotiated between the two families.

THE EXCHANGE OF A PERSON FOR A PERSON

Let us analyze the first case: the exchange of a person for a person, taking the example of direct sister exchange involving real or classificatory sisters (e.g. patrilateral parallel cousins) between two men. Such an exchange seals an alliance between the men’s and the women’s lineages.

WIFE-GIVERS ARE SUPERIOR TO WIFE-TAKERS

We have already cited an example of this case when analyzing the prevailing form of marriage among the Baruya, ginamaré, which is precisely an exchange of sisters between two men and two patrilineal lineages. Here we are in a logic of reciprocal gift and counter-gift. It is important to understand that the counter-gift of a woman in this case does not cancel the debt each of the two men or their lineages has contracted with regard to the other by receiving one of his sisters as a wife. At the close of these reciprocal exchanges, the two men and their lineages find themselves in equivalent positions: each is simultaneously the other’s creditor and their debtor. Each creates an obligation in the other as a wife-giver and is in their debt as a wife-taker. Each lineage thus finds itself in two opposing relationships with the other, but the sum of these opposite inequalities actually puts them back on an equal footing in the society (which supposes a code of values shared by all members of this society when it comes to assessing this status). This alliance sealed by a double marriage will be the starting point of a flow – between the two brothers-in-law and the two sisters-in-law and their two lineages – of goods and services that will last throughout the couples’ lifetime and be continued by their descendants. After several generations, the debts die out and marriage alliances can once more be concluded between the two lineages.

SOME DEBTS ARE NOT CANCELLED BY A COUNTER-GIFT

Why is the debt created by the gift of a woman not immediately cancelled by the counter-gift of another woman? At the close of this exchange, each woman has taken the other’s place without ceasing to belong to her original lineage. If the counter-gift does not erase the debt, it is precisely because the person ‘given’ has not been detached from the lineage that gave her. She has been ‘given’ without being truly alienated by those who gave her and who continue to have rights in her (and her descendants). Through this exchange, one woman has taken the place of another while keeping her original identity. This reciprocal permutation of persons and places is the act that produces the alliance between the members of the two lineages. These movements of persons, this production of new relations between them and between their birth (or adoptive) groups are induced by the fundamental social pressure that is the prohibition of a man marrying his sister or a woman marrying her brother. The source of sister exchange – but this is just as true in the case of brother exchange – lies in the compelling force of the incest taboo.

A few more remarks are needed to enter more deeply into the logic of these exchanges. The persons who have been ‘given’ are not alienated. They keep their original identity and are not completely detached from their original lineage. They are separated from it. They are not goods that one can alienate and which pass from seller to buyer, who does what he will with them. What is given, ceded, are the rights in the persons, rights to their domestic, sexual and economic services and to their reproductive capacities. But we should not forget that, when it comes to matrilineal societies, a woman’s children belong to her own lineage and not to her husband’s.

Furthermore, when two persons (two men or two women) are exchanged, this is an exchange of two beings whose ‘value’ is a priori deemed equivalent by their lineages and their society. Children’s socialization and education ensure that boys know how to hunt, clear forest, fight, etc., and that girls know how to cultivate gardens, raise their children, care for pigs, make clothing. And everyone is asked to conduct themselves as responsible adults who do not go about creating conflicts that are a threat to families and society. But even the best upbringing cannot guarantee that a woman will not be barren (or a man sterile), that their children will survive, and so forth.20

It needs to be stressed that, by this gift and counter-gift of women, the two lineages find themselves mutually indebted to each other and thus feel obligated for the rest of their lives to exchange services, to share the salt-bars they produce, the meat of the pigs they slaughter, to invite each other to clear a new garden in the forest and to work it together. These reciprocal gifts of goods and services, this mutual aid in the event of strife in the village, materialize the debt they contracted by each receiving a wife from the other who would enable them to perpetuate themselves (independently of what, in the child, is supposed to come from its father or mother). These gifts are therefore in no way bridewealth.

It is also important to reiterate that many societies, like the Baruya, who practise direct sister exchange for the most part, abstain from repeating the same alliances before several generations. A son may not reproduce his father’s marriage and take a wife from his mother’s lineage. Two brothers may not take wives from the same lineage, and even less marry two sisters. These bans prevent the constitution of groups made up of two lineages or two clans that reproduce together and thereby isolate themselves from the rest of the society, at least from the standpoint of the exercise of kinship. These prohibitions thus compel each lineage to multiply its alliances and ensure that each is simultaneously the starting and ending point of several alliance paths, which, despite their possible expansion, can never manage to include all the lineages of all the villages of the society. In order for an exchange system to bring all members of a society into a single network, these individuals must have been sorted from birth into distinct categories which intermarry according to rules that compel each individual, according to his or her category (moiety, section, subsection), to choose their spouse exclusively from another category, the same one every time, thus reproducing the same alliances from one generation to the next. This is the logic underlying the Australian and certain Amazonian systems, which will be examined later.

A final point: the principle of the direct exchange of women – real or classificatory sisters – is always limited by the number of ‘sisters’ available for exchange, so that, in order for all men to find a wife, there must be a prohibition or a strict limitation on certain men – the eldest, for example – exchanging all of their sisters for their own benefit and thus sentencing their younger brothers to a long bachelorhood and to delaying marriage or marrying widows. But this obstacle to direct exchange disappears when it is no longer persons but goods or wealth that are exchanged for a person. In this case, the character of the exchange and its limits are entirely different. On one side, there is a woman or a man, in other words concrete persons, and, on the other, wealth, valuables (shells, jewellery, etc.), or goods (livestock, pigs, etc.), which can be produced or procured through trade or by some other means. On the one side, there are persons; on the other, different kinds of ‘things’ that function as substitutes for persons. The equivalence between the two exchange terms takes on a new, much more abstract character than in the exchange of a person for a person. The moment persons (men or women) are exchanged for wealth, there appears a true political economy of kinship. Wealth procures women, and women procure wealth through the bridewealth they bring to their lineage and through their productive activities in their husband’s lineage.

WEALTH FOR PERSONS

To illustrate this change of logic in marriage practices and in the way societies function, we will take the example of the Melpa, a widely dispersed set of tribes whose territories lie around Mount Hagan in the heart of the New Guinea Highlands. The Melpa are famous for their system of competitive ceremonial exchanges known as moka, in which clans and tribes from an entire region used to face off on the occasion of large-scale redistributions of pigs and goldlip pearl-shells, which would be exchanged for pigs, marsupial furs and so forth with tribes further to the south, who had procured them through exchange with tribes living along the Gulf of Papua. Moka, like the Kwakiutl potlatch, consisted in giving more than the other clan could give in return, or returning more than had been given, the aim being to make the others your debtors and force them to recognize their inferiority when it came to accumulating and redistributing wealth. As one might guess, this escalation of generosity was driven by political interests. It glorified the name of the richest and most generous clans, it spread far beyond their tribal boundaries the renown of their Big Men who had managed to amass such wealth by their ability to produce it and/or to convince their kinsmen and their affines to engage their own pigs and shells in the same venture.

In these Big Men societies of New Guinea, direct sister exchange, while theoretically known, was not practised and was even explicitly forbidden by the Mendi. The general rule for concluding a marriage alliance was to exchange wealth for a woman. For example, in the Melpa society, when two lineages agreed to unite two of their children, negotiations to set the amount of the marriage payments were conducted in several stages. We will summarize this process, referring to Andrew and Marilyn Strathern’s remarkable analyses.21

In the first phase, the bridegroom’s lineage presents the bride’s lineage with a number of goods they plan to give them. In the second phase, these goods are formally given to the woman’s family, which gives other goods in exchange. The men make long speeches, while the pork from a certain number of pigs brought by the groom’s family is redistributed and eaten. The woman’s family gives the couple a number of live pigs to start the herd they will later raise and which will enable the couple to present the woman’s lineage with pigs.

FOR THE WOMAN’S HEAD AND VAGINA

In the course of these ceremonies, a number of shells are exchanged between the two groups, while others are presented to the bride’s family as a one-way gift. Several of these shells are called peng pokla, which means ‘to cut off the (girl’s) head’, in other words separate her from her lineage; but this separation is never complete. A number of pigs are also given with no expectation of reciprocity. Several of these pigs are described as kem kng, ‘for the girl’s vagina’. A particularly large pig is called mam peng kng, ‘the pig for the (girl’s) mother’s head’. Another is given to the bride’s father as a present from the groom’s family.

If we analyze these exchanges, we see that they have three components. Reciprocal exchanges of shells and pigs – approximately equivalent in quantity and quality (before the Europeans arrived, a living pig exchanged for two pearl-shells) – between the two parties designed primarily to seal their alliance. A second series of shells and pigs are given to partially detach the woman from her own lineage and from that of her mother, and to transfer rights to her sexual services and reproductive capacities to her husband’s lineage. Last of all, the woman’s family presents the couple with a number of pigs that constitute the beginnings of a herd whose eventual size depends primarily on the labour of this woman, but the product of which will be used by the husband to take part in moka and to fulfil his responsibilities in various situations (funerals, initiations, etc.) which require contributions of pork. By endowing their daughter with this ‘productive capital’, the members of her lineage establish themselves as the future couple’s primary moka partner.

This third component is not a dowry, in the sense of having been given by the bride’s lineage to ensure her material autonomy in her new family and which she could take with her in the event of divorce, as was the case with dowries in societies around the Mediterranean Basin. Lastly, it is to be noted that land never features in these endowments. Lineage land is indivisible. When land is divisible and part of a family’s land is detached and included in a daughter’s dowry when she marries, we are dealing with an altogether different logic and other marriage strategies, as shown by Jack Goody and S. J. Tambiah in their work on dowry and devolution of goods in Europe and the East22 as well as by the discussions23 sparked by their hypotheses.

Two remarks are called for here. It is clear from the Melpa example that the circulation of goods entailed in a marriage alliance can be imbedded in much broader exchange systems based on competition for prestige and renown, in short, on winning status in a political configuration. The intention behind some of these gifts is not only to repay the gift of a woman with wealth, but to make affines into moka partners. The Melpa marriage is actually fully concluded only when the groups and individuals involved become partners in the moka and enter into competition with each other even as they cooperate. Marriage alliances and kinship relations are therefore subordinated to the perpetuation of a far-flung network of competitive political-ceremonial exchanges, following a potlatch logic, and made to serve the expansion of this network, which entails many dozens of clans and thousands of individuals. In short, these relationships are of another order than kinship: they are political.

This example tells us why, with a few exceptions, marriage in potlatch societies cannot be based on the direct exchange of women. Such a system would risk short-circuiting the competition between groups in the exchanges of wealth that give access to titles, ranks and functions – whose number is limited – as well as to power and renown.24

More important still, the equivalence between the terms of exchange in the case of wealth for a person has nothing in common with the equivalence postulated when exchanging a person for another person. Even though societies may strive to limit the amount of wealth given for a woman (or a man) and to set an average exchange rate, there is no objective criterion that can justify giving six big pearl-shells and three pigs for a woman’s ‘vagina’ rather than four shells and two pigs. The nature and the quantity of the ‘things’ given are primarily indicative of the rank and status of the groups contracting a marriage alliance. Today in New Guinea, marrying the daughter of a Big Man or of a regional member of the National Assembly can entail dowries and redistributions of several hundred live or slaughtered pigs (some of which are bought from industrial pig farms with money from the sale of coffee25), to which are often added a Toyota or Nissan truck and tens of thousands of kina in cash.

The inflation of dowries, observed not only in Oceania but also in Africa and Asia, is the direct consequence of the growing involvement of these societies in the local and global market economy. This inclusion results in the generalization of the use of money in traditionally non-economic social exchanges (e.g. rituals) and accentuates and multiplies the differences of wealth between individuals and between kin groups, which are still an important component of local territorial groups’ social structure. In this shift, we see the development of a genuine traffic of women, a phenomenon that did not exist in the case of sister exchange between two men and their lineages.26

A dowry payment at the time of the wedding does not mean that the debt of the wife-taking lineage is extinguished at the end of the ceremony, if there is one. Other ‘payments’ will come due, for example, each time the couple has a child. This is the case among the Wiru and the Daribi of New Guinea, societies that do not take part in regional competitive systems like the moka, but where the bride’s father and her lineage are supposed to exercise ritual control over their daughter’s fertility throughout her life. Each child she bears is seen as a new gift made to her husband’s lineage, which is added to the initial gift of its mother. Based on this (imaginary) representation of the process of reproducing life, a flow of gifts thus accompanies the birth, marriage and death of all individuals. The dowry in this case is merely the first of a series of gifts inaugurated by a marriage alliance, which will punctuate the life of the individuals that will be born from this marriage, from their birth to their death – and beyond.27

THE CASE OF INDIA, WHERE WIFE-TAKERS ARE SUPERIOR TO WIFE-GIVERS

In a great number of societies the wife-takers are superior to the wife-givers. So that the marriage may take place, the givers offer the groom’s parents, in addition to their daughter, a dowry (and not bridewealth), the amount or the nature of which the groom’s parents can accept or reject. It is often stipulated that the girl must be a virgin. This ‘value’ attached to women’s virginity is not universal, but it prevails in the Euro-Asian zone and can be found in various forms as a principle of Christianity, of Islam, of Hinduism – but also of traditional Chinese society. In this case, the woman’s body is the vessel of the family honour and status. We will take the example of Brahmin India.28

In India the superiority of takers over givers is connected with the form of marriage practised by the purest, highest caste in Indian society,29 the Brahmins. In an ideal Brahmin marriage, a father gives his virgin and richly endowed daughter to a family of superior or equivalent rank to his own. And for this gift to be a source of merit, in the present life and beyond, it must be freely given, with no expectations or reciprocity, with no counter-gift on the part of the groom and his family.

This type of marriage was already described in the Manava-dharma-sastra, a set of normative Sanskrit texts known as the ‘Laws of Manu’, written over a period of several centuries (probably between the second century bce and the second century CE, in other words, at a time when Buddhism was spreading through India and driving the Brahmins to set their traditions down in writing). This form of marriage – the gift of a virgin with a dowry – subsequently became a norm that spread to other castes as Indian society became Sanskritized,30 and many local tribal societies adopted Hinduism, little by little entering the caste system or continuing outside it but under its influence.31

The Dharma-sastra mentions seven other forms of marriage as well, and classifies them into two groups of four. The first are entitled Dharmya, in other words ‘according to dharma’, and are therefore virtuous, adding to the merits of those who practise them. The second group uses terms referring to evil spirits, enemies and rivals of the gods to designate four forms of marriage based on interest (accepting bridewealth, a marriage payment, for the gift of a girl), or on desire (choosing each other without being authorized by the fathers), or on violence (abducting the bride), or on fraud. In short, eight forms of marriage, ranging from the purest, most virtuous, most disinterested to the most immoral and impure. But even if the last four forms are opposed to the first four, all are regarded as forms of marriage. We will return later to this list and its underlying principles after having attempted to identify the principles, values and representations of traditional Hinduism that ground the tradition of the superiority of wife-takers over wife-givers.

The explanation lies in the representations of the status of son and daughter, and therefore that of husband and wife, in Hindu tradition. According to this tradition, a man is born with three debts – to the gods, to the rsi (the original sages or seers, who ‘saw’ the Veda) and to the ‘fathers’. A man therefore owes sacrifices to the gods, owes study of the sacred texts to the seers, and owes descendants, above all a son, to the fathers. Over his lifetime, little by little, through sacrifices, study and engendering a son, the man frees himself from his original debt. In addition, a man must engender a son for himself as well, for it is a son that must perform the funeral rites when his parents die and thus enable them to enter the world of the ancestors. It is a son who will then continue to make offerings to the family ancestors for the rest of his life. In India the birth of a son is therefore a source of joy and an occasion for rejoicing.

This is not the case for the birth of a girl, for she is a woman, and according to the classical texts, women have an evil nature. She is inherently evil. Her insatiable sexual appetite is a danger to men and society. Her body is a source of pollution through the menstrual blood that periodically flows from it. Furthermore, the woman is considered unfit for reflective thought – and therefore incapable of independence. She must always submit to men, whose duty it is to guard her, ‘even from herself’, to guide her but also to protect her. ‘Her father protects (her) in childhood, her husband protects (her) in youth, and her sons protect (her) in old age; a woman is never fit for sva tantra (independence).’32

A father’s duty is therefore to keep his sons and to marry off his daughters as soon as they reach puberty or even before. The Sanskrit word for marriage, vivaha, contains the idea of dissociation (vi) and the transfer of the daughter from her father’s house to that of her husband. But it is also the rule that if a certain time is allowed to elapse after a girl reaches menarche, she is then no longer under her father’s authority and can choose her own husband – but she then loses all right to a dowry. Nevertheless, she can choose to go ahead without incurring any blame, since her father and the members of her group will be regarded as the ones responsible for not having found her a husband in time. They will be the ones censured by public opinion and not the woman or her husband. For the father, this social condemnation will be accompanied by the ancestors’ wrath, since, for a Brahmin, not to marry a daughter in time is considered to be as serious as murder, the gravest crime one can commit.33 The ancestors may even be so angry that they thirst for the girl’s menstrual blood as an offering.

It is with these philosophical-religious representations of the woman as a potential and constant source of defilement in a social and cosmic order organized around the opposition (at multiple degrees) between pure and impure that we must begin our attempt to understand why, in India, wife-takers are superior to wife-givers, as well as why the gift of a pre-adolescent or an adolescent girl who has never had sexual relations with a man (nor, they add, with a woman) is regarded as a disinterested gift, and at the same time as a religious act. But a piece of the puzzle is missing. For the gift is disinterested in the sense that the wife-takers do not reciprocate with either women or bridewealth. And yet it seems that at another – immaterial, religious and cosmic – level, the wife-takers do reciprocate with a counter-gift. They take upon themselves the responsibility of dangerous fertility, source of the pollutions occasioned by this menstruating girl. They, and no longer the girl’s father, will be responsible for the ‘bad omens’ entailed in these occasions for pollution, which the bride carries with her, in her body by her very nature. And it will be up to them to transform this dangerous fertility into a source of life, through marriage and procreation.

To celebrate the marriage of a girl is therefore at the same time to transfer from one family to another – from the givers to the takers – the dangers carried in this girl’s sexual body. These dangers are infinitely greater still if she has had illicit sexual relations with men or women before her marriage. A woman thus contains within her body the possibility of ‘bearing’ her family’s rank and status, of bringing them honour or shame. Her virginity is the means and the stake. In addition to being a virgin, she must also testify to her family’s wealth, prestige and standing. This she does when she arrives with her dowry, decked in her jewellery and accompanied by the goods given to women when they marry and which are already their inheritance. These are personal goods, for in India real estate such as land and houses is traditionally reserved for the sons and is divided among them on their father’s death.

That the gift of a virgin with her dowry has never been entirely a one-way, non-reciprocated gift, and that this counter-gift takes place on a level other than that of women and material wealth, has been largely confirmed by the results of detailed ethnographic studies, such as those conducted by Gloria Goodwin Raheja on marriage rites and payments in the dominant caste of a North Indian village. In an acclaimed book entitled The Poison in the Gift,34 she showed the importance of a set of ritual acts by which the wife-takers act with the explicit intention of ensuring the well-being of the givers by taking responsibility for the inauspiciousness that accompanies the transfer of a woman into their family and which is connected with the fact that the families’ ancestors and gods attend the wedding and mingle with the guests to ensure that the rules have been respected. Among the Gujar – but this is not the case everywhere in India – the groom’s family in turn, through other rituals, transfers onto the Brahmins who perform the ceremony for the family some of the impurities and faults engendered by the act of receiving the gift of a virgin woman who has begun to menstruate (kanya dan). In the end, all the world’s impurities converge on the Brahmin, and for that he receives goods and prestations.35

Giving a virgin with dowry to a Brahmin is, according to the Dharma-sastra, the second of the four forms of dharmya marriage that are a sign of virtue and a source of merits for the girl’s father and her family. This marriage is called daiva, a word derived from deva, meaning ‘gods’, and consists, for a rich man who has embarked upon a cycle of major sacrifices, of giving one of his daughters to the Brahmin who performs these sacrifices for him. The gift of his daughter is added to the material gifts and prestations normally given a priest for fulfilling this role.

Without suggesting in the slightest that the forms of marriage practised in India today correspond to the eight forms listed and ranked by the Laws of Manu, it is still worthwhile examining these forms briefly in an attempt to glimpse a cultural and social world with which we are not automatically familiar. They are listed in descending order starting with the first, which we have described, the so-called Brahma marriage, from the name of the great god of all the worlds. The second form is Daiva. The third is named Arsa, from the term that designates the rsi, or seers, the sages of old. In this type of marriage, the father gives his virgin daughter but receives in exchange from his future son-in-law a few head of cattle for ritual use. The fourth form of marriage is called Prajapatya, from the name of the god of procreation, Prajapati. It differs from Brahma in that the gift of the girl is solicited.

Next come the four forms of marriage placed under the sign not of the gods but of the demons or spirits inimical to the gods. Asura – a term designating spirits that personify evil nature spirits – by which a man requests the gift of a woman and makes a marriage payment (sulka) to her family, and gives the woman herself goods which constitute a sort of dowry. The Gandharva marriage – from the name of the heavenly musicians who play dance music for the nymphs – is the union of a man and a woman who desire each other and marry without their families’ consent.36 Rakasa – from the name of the highly dangerous nocturnal demons (raksas) – is marriage by the forcible abduction of the girl from her family.37 And last of all, in Paisaca – from the name of the carnivorous demons that haunt cremation grounds – a man has sex with a woman he has found in an inebriated state or asleep and makes her his.

Under the sign of the gods	Under the sign of the demons
1. Brahma	1. Asura
2. Daiva	2. Gandharva
3. Arsa	3. Raksasa
4. Prajapatya	4. Paisaca

Several remarks are called for concerning these ‘demoniacal’ forms of marriage (it must not be forgotten that demons, though below the gods, share the same nature).

In Asura, a woman is exchanged for wealth. This form is encountered in Africa and in New Guinea, among the Melpa, for example. In India, it is also called by the names of local tribes, as though it were still marked by the forms of marriage exchange practised by tribal groups subsequently absorbed into the caste system. But even more interesting is the fact that this marriage has negative connotations in India because it does not appear as a gift but as the sale of a daughter, a materially interested gift, which thereby loses its gratuitous character. The father and mother are held equally responsible for ‘selling’ their daughter because they subsequently share the goods received in exchange. It is as though giving a woman in exchange for wealth was already, at the time of the Laws of Manu, assimilated to a commercial transaction in which the parties negotiated the price of a commodity. This view partakes of a world vision and a value hierarchy that associates buying and selling, and more generally all commercial activities, with defilement. It is therefore not because the parents decide for their daughter without her consent that this form of marriage is censured. It is because this marriage is not a gift but a sale, and a sale can be concluded with whoever offers wealth for a woman – whether he is from the same or, worse, from a lower caste. In short, according to this viewpoint, commercial relations have their place in the logic of caste hierarchy but never beyond. It must also be stressed that, in this view, the parents of a girl who has been abducted or forced to marry always have the right to demand material compensation after the fact. This payment is not bridewealth but atonement for a wrong.

Another interesting fact is the recognition expressed in these texts of the woman’s right to marry whom she will if her father has not found a husband for her within a certain lapse of time after her menarche (Gandharva marriage).38 In this case, she is free to marry; however, although she is not at fault she loses her right to any dowry. She no longer represents her family and no longer has a right to their support or to their assets. Last and above all, these ancient texts leave room for the love marriage. To be sure, this is considered a bad form of marriage because it does not respect paternal authority, but it is accepted as a possibility. Moreover, it is this form of marriage that for centuries has nourished a whole poetic, lyric, even ‘romantic’ literary genre. Today its importance in people’s imaginary and aspirations has not ceased to grow and it now passes for the ‘modern’, non-traditional form of marriage, shifting desire and individual wishes to the fore, rather than the authority of one’s kin, who continue to pursue their marriage strategies in view of maintaining or raising their position in the caste hierarchy.

Today, too, the extreme monetarization of the dowry, the consequence of a distortion of ancient customs, resulted in a law being passed in 1961 in India which forbids a girl’s parents to pay a dowry to the parents of the future husband on pain of as much as six months in prison and a fine. The law also stipulates that, if a brideprice is paid, it should go to the woman or be used by the newlyweds to set up house, but should not be paid to the father of the bride. This now-banned dowry or brideprice is called vara-dakshina, from vara (future husband) and daksina (gift made to the household priest for performing the life-cycle sacrifices). It actually functions as a groomwealth, as a payment demanded by the parents of the groom for allowing their son to marry. The vara-daksina is widespread and has taken on a modern form linked to the expansion of commercial and monetary relations. Instead of the young woman arriving at her husband’s house decked in her traditional finery (gold, jewellery, clothing, etc.), she now comes bearing several hundreds of thousands of rupees, the sum her future in-laws have demanded to accept her as a daughter-in-law. More than ever in India it is better to have sons than daughters to marry.

Thus in Brahmin marriages, although the givers of a virgin do not receive a woman or wealth in return, on the spiritual level they receive a sort of indirect gift from the takers, since the latter take responsibility for the pollution their daughter bears and bears away with her. But there are also societies in which wife-takers give nothing in return. An example is provided by the former marriage practices of the Sioux Indians.

In Sioux society, where the men divided their time between making war, hunting buffalo and taking part in large-scale rituals, and where all of the material goods belonged to the women, to the wives, the rule was to marry outside the band. The ideal wife was a prisoner of war, and the adoption of children or adults was just as important as descent. Their kinship terminology was Dravidian, but they had no specific term for husband or wife. The word for ‘to marry’ was ‘to abduct’, to capture (a woman). Moreover, a man was forbidden to speak to his in-laws, who were treated as, or really were, enemies. Sisters lived under the very strict supervision of their brothers, and it was the latter who made the decisions concerning their marriage and the size of the brideprice (in horses or other forms of wealth) that the suitor was to pay. Warriors gave their sisters enemy scalps, and it was the women who tortured prisoners, sometimes to death. Ideally the brother-in-law should be killed in order to capture his sister. We are at the far pole from the practices and values of the Baruya, for whom sister exchange between two men binds them for the rest of their lives and draws them into a cycle of exchanges of goods and services that ceases only with their death. Ultimately, in Baruya society, two brothers-in-law eventually become closer and more committed to each other than two brothers.

It may be the prime importance of bride abduction, of marriage founded on capture, on predation, that explains why, in a highly unusual manner, Sioux kinship terminology derives the terms for cross cousins from those designating affines, the wife’s brother and sister. The outside prevails over the inside, but it is an outside characterized by predation and death. At least this is what Emmanuel Desvaux has tried to show in an article rich in surprising perspectives, devoted to the difficulties of Sioux kinship nomenclature.39

WHEN TAKERS AND GIVERS ARE EQUAL

Finally, the last possibility: wife-givers are neither superior nor inferior to wife-takers. Takers and givers are equal. This is the case in many societies, including the Vezo of Madagascar, studied by Rita Astuti.40 In this possibility, the exchange takes the following form. The woman’s parents say to the man’s parents: ‘We give you our daughter, take her.’ And the man’s parents respond: ‘We give you our son, take him.’ The two gifts are equivalent. The Vezo have a cognatic kinship system in which a person is linked to the ancestors through men and/or through women, indifferently. There is therefore in this case no exchange of women by the men or exchange of brothers by their sisters. Nor is there any such thing as exchanging a woman or a man for wealth (bridewealth or groomwealth). There are two reciprocal gifts of individuals of different sex by two families, who cooperate to produce a third family whose members will descend from each of the two spouses inseparably. From a certain standpoint, it cannot be said that these two gifts of persons of different sex are an exchange. The gifts are two parts of a single act performed at the same time by two families in order to form a third. In addition, according to Rita Astuti, this society does not discourage sex before marriage and does little to encourage it afterwards. The woman has more rights over the children than the man.

A summary of the various cases of marriage and the status of the partners is shown in the following table:

Wife-givers are superior to wife-takers (anisogamy)

Wife-takers are superior to wife-givers (anisogamy)

Wife-givers and wife-takers are equal to each other (isogamy)

It is clear that the obligation to pay a bride- or a groomprice to seal an alliance places young people of marriageable age in a situation of personal dependence on their parents, their elders, their family or their lineage. Generally speaking, these young people have not yet managed to build up their own means of payment (livestock, shells, gongs, pottery vessels, etc.). They rely on other people to provide these so that they may give them in turn. Yet the fact that elders control the wealth and the means of social reproduction does not mean that they behave like a ‘class’ which dominates the younger people and exploits their labour in exchange for the payment that will enable them to find a spouse. This thesis was defended in the 1970s by a number of Marxist anthropologists, citing Claude Meillassoux’s work.41 They saw control by the elders as ‘the beginning of the transformation of the society with the genesis of hierarchical social classes’, the transformation of older–younger relations into patron–client relations.42

No one can deny the general fact that younger generations depend on their elders for, according to the context, transmission of land or status, succession to functions, and marriage payments. This dependence also entails a relationship of authority between older and younger generations, and unequal responsibilities. But it does not necessarily mean domination or exploitation. In New Guinea, for example, the young people have duties to the other members of the lineage, but that is not all; they also share rights with the older members – the right later to use lineage lands, the right to the bridewealth they need to get married, the right to be avenged by their lineage in the event of homicide or to count on their armed solidarity in the event of aggression or vengeance.

In societies where the merits someone has acquired or the wealth they have produced, individually or by their capacity to enlist kinsmen, affines or friends, give them authority in their group, authority and prestige do not automatically go to the older men, and even less to the oldest or the ‘elders’, but to the Big Men (and to certain women, who achieve the status of Big Women through other means). The Big Man is an older man, but not all older men are Big Men, and even less the oldest. This is apparently not the case in the African societies described by numerous specialists, from Meyer Fortes to Meillassoux. But the example of New Guinea shows at least that the fact that the older men control the lineage lands does not automatically result in domination and clientelism. Other conditions – which largely remain to be described – must be present for this to occur.

Another point is worth mentioning: we saw, among the Daribi for instance, that gifts to the wife’s lineage continued throughout life because the woman’s father is believed to have spiritual and ritual control over his daughter’s fertility. The children she bears are therefore further gifts, as it were, from the woman’s lineage to that of her husband. This is a cultural dimension based on an imaginary representation of the source of women’s fecundity. This representation has serious social consequences, because it makes for a particular configuration of the exchanges between the two groups of affines and between individuals in virtue of the positions they occupy within their group and in these exchanges (husband/wife, father/daughter, father-in-law/son-in-law, etc.). It is important to take note of the social character and the imaginary dimension of the wealth exchanged or given as a one-way gift to seal these alliances and produce new social relationships. In New Guinea, the pig is not valued because it is the Highlanders’ principal source of protein, and to give dead or live pigs is not merely a way of redistributing a certain quantity of pork for consumption or of making a gift of sows that will have piglets. The same is true of cattle in Nuer society. But it also applies to ‘inanimate’ objects which, in addition to pigs, feature in exchanges – polished and decorated shells worked by the men and women in Melpa society, or decorated mats which circulate by the dozens in Polynesian exchanges, the most valuable, the most sacred of which were used to envelop statues of their gods or the bodies of their dead.

OBJECTS AS SUBSTITUTES FOR PERSONS

All of these valuables act as substitutes for persons living or dead. They are given, for example, by a murderer’s lineage to that of the victim to compensate this death and to save the murder’s life. They must therefore be produced (or acquired) and then transferred into other hands so as to establish not only marriage alliances but also political alliances, and alliances with gods and ancestors. By means of these objects and their transfer, individuals and groups contract relations with others, and these relationships form part of their identity. The exchanged objects are loaded with both the meaning and the strength of these relationships, vested with cultural meanings and social importance. They are thus mental and social representations materialized in animate or inanimate beings. I purposely say ‘beings’ rather than things or objects, for these ‘things’, which are substitutes for persons, are perceived as containing powers for acting on persons and therefore as being in a certain manner persons themselves. That is why, like human or supernatural people, some of these objects (shells, mats, etc. – which are of no use in daily life when it is merely a question of subsistence, but which are necessary for producing a social existence) acquire a name, an identity, a history and powers of their own.

As an example of the complexity of the imaginary and symbolic meanings taken on by some shells, which explain their use in the reproduction of kinship and political relations but also their status as both wealth and symbols of power, we will summarize some findings that Jeffrey Clark presented in his exemplary study of the symbolism of pearl-shells among the Wiru of New Guinea,43 who exchange them for ‘the body’ of a wife, for ‘the skin’ of the children she bears, and so forth.

Here briefly is how the Wiru load a pearl-shell with meaning. The naturally yellow pearl-shell is rubbed with ochre powder and its lower lip is outlined in white sap, which quickly turns black. Several notches are cut into the upper lip, and the whole shell is laid and exposed on a bark support. All of these operations are the outcome of meticulous work which transforms an object that entered Wiru society as a commodity bartered for a pig (or bought with money) into an object not only loaded with a new meaning but also one that has become more beautiful in their eyes – and by their hand. What is this new meaning? For the Wiru, yellow is a female colour, associated with a yellowish substance that they say is found in the womb and becomes a key component of the foetus when it is conceived. Ochre is associated with virility and wealth, and for this reason the sacred stones associated with the fertility of the land, human health, etc., are smeared with ochre powder. The white of the sap is associated with semen, and the black, like the ochre, is associated with virility. The cuts on the female lip of the object represent the incision of the glans of a penis. In short, these androgynous objects harbour the attributes of the men’s masculinity and the women’s femininity – in this case essentially their reproductive capacities. It is in this sense that the apparently inanimate objects given in exchanges are not only substitutes for persons but are themselves objects endowed with meaning and power.44

The same holds for the Nuer who, according to Evans-Pritchard, define all social relations in terms of cattle. When they are initiated, young men are given the animals that they will tend to throughout life, a life they will spend in a society where the production or maintenance of practically all social relations demand the transfer of cattle in various ways.45 It is understandable that in these conditions an identification is created between men and their cattle such that, when they give away their animals, they are giving part of themselves. By alienating part of their herd to get a wife, men detach part of themselves from their own lineage and expect that their wives will detach from their own bodies and lineage the children they will bear. But the husband’s lineage will not appropriate these children unless he has paid his affines a brideprice. Otherwise they will belong to the man who pays the price in the husband’s stead. This man will then become the children’s social father, while the husband is no longer the father but simply the genitor.46

WHAT IS THE BRIDEPRICE?

Making a marriage payment, paying a brideprice is therefore not ‘buying’ a woman. Except in very rare cases, such as ancient China, the woman is not completely detached from her birth group and absorbed into the group, clan, lineage or family that has alienated a portion of its wealth to obtain her. Women continue to possess rights in their birth group and have obligations to the members of this group. In addition, a wife acquires rights in her husband’s group, together with duties. This is precisely the case in China, where the daughter-in-law’s status changes once she has given birth to a boy who will carry on her husband’s line and continue to venerate their ancestors after his father’s death. When a woman dies, it is not unusual for her name to be included among her husband’s ancestor tablets. In short, we are far from the image suggested by Lévi-Strauss’ famous formula ‘kinship is based on the exchange of women by and for men’. Moreover, the groups that exchange partners are not composed exclusively of men but of men and women who descend from common ancestors and in which women also have a voice.47

In matrilineal Minangkabau society, where the husband is called by an expression that means ‘borrowed man’, the matrimonial prestations give the husband and wife rights in each other and, through them, extend these rights to their groups.

In sum, all these various prestations (bridewealth, groomwealth, dowry, conjugal prestations) establish two kinds of ties at once: conjugality between the spouses and affinity between the groups that become allied through them. In the next generations, the affines are transformed, depending on the type of kinship system, either into consanguines (as in the Western kinship system) or into consanguines and affines (as in the Dravidian system, where the mother’s brother and the father’s sister are affines for Ego and not consanguines).

To conclude this point, let us say that the exchanges that seal an alliance are always fundamentally exchanges that transfer from one group to the other rights in the persons who are detached from one group and attached to the other. But these detachments and attachments are almost never complete.

Jack Goody, following Leach, drew up a relatively exhaustive list of the various categories or rights that can be transferred by these exchanges and which constitute reciprocal or non-reciprocal obligations between the spouses (conjugal rights) and between the allied groups.

• Rights to henceforth officialized sexual relations.

• Rights to domestic services.

• Rights to the men’s and/or the women’s procreative capacities and rights for the allied groups on the children born of the marriages.

• Rights to cooperation in the production and exchange processes if the family, lineage, etc., are production and exchange units; and rights to a share of the things produced by these units or exchanged between them.

• Rights to mutual assistance and to solidarity in the event of political and social strife.

• Rights to mutual assistance in the performance of rituals and other ceremonies concerning ancestors, spirits and gods.

Of course one party’s rights are at the same time the other parties’ duties. This is why it is indispensable to distinguish among these rights those that are reciprocal and non-reciprocal, are identical or different, equivalent or non-equivalent, complementary or opposing. For example, the husband’s right to demand that his wife join him in living next to or with his father is an obligation, a duty for the wife. In other societies it is the husband who will go to live with his wife or with her parents, either during the first years of marriage or definitively.

Such transfers of rights in persons are often the starting point for transfers of material and immaterial goods48 to the individuals that are born to these alliances or who will be integrated through adoption. Lands, names and titles will be inherited by all descendants, male and female, of a union, or by only certain men or women. Standards, values and knowledge will be transmitted, social functions will be transferred. Filiation and descent return in force to unite or divide people according to sex and age, and combine with alliance to ensure the transfer from one generation to the next of a share of the conditions of their physical, material and social conditions of existence. For instance, in a society with a matrilineal descent rule, descent groups give other groups the right to their daughters’ sexual services but keep the children born of their marriages for themselves. In ancient Roman law, girls were exherited en bloc, and the father would choose a single son to inherit his potestas, or authority, and carry on the ancestral cult after his death. For as long as he lived, the father would exercise his patria potestas, the absolute right, recognized by the city of Rome, over the members of his family, including his married sons and even his married daughters, if they were still ‘under his hand’. And yet the authority that prevailed in the everyday domestic life of this ‘family’ was not founded, as Yan Thomas showed, on this unlimited potestas.49

Let us take another look at the notion of patria potestas, the absolute right of the pater familias over his descendants. To be a pater familias in ancient Rome, a man could not be under the control of another, he had to be sui juris, under the control of no other, to have been chosen by his father to inherit his potestas and have received it instantaneously on his father’s death. The pater familias was thus an autonomous subject, whereas all other members of the family were alieni juris, under the control of another. The pater familias thus had control over his wife, his children, his slaves and his goods. It was the fact of inheriting the patria potestas, the father’s power, that made a man a pater familias. Such an heir could be unmarried or be married without children. He was still a ‘pater’ familias. This power lasted all his life, for his sons, even when they married and had children, remained under the control of the pater familias until his death. Daughters too, but in ancient Roman law girls were given in marriage and then passed under ‘the hand’, the manus, of their husband, becoming as a daughter. A wife thus became the sister, as it were, of her sons and daughters. And if her husband was under the control of her father-in-law or an older ascendant, then she came under this man’s control also. But alongside marriage with manus, usually reserved for patricians, there was a marriage without manus, practised for the most part by plebeians. In this case the married woman remained under her father’s patria potestas.

These laws evolved considerably up to the end of the Republic and over the duration of the Roman Empire. Women became able to transmit their possessions to their children, who had priority over the woman’s own brothers and sisters, her agnates. But until a very late date, a Roman woman could not testify in court on behalf of someone other than herself, and she was never able to undertake the defence of someone else, or to represent anyone other than herself. The last function remained a male public prerogative. As Yan Thomas stresses, in the political arena as in intersubjective civil relations, Roman women were always prohibited from providing a service that went beyond the narrow sphere of personal interests. Until the very end, the city-state remained a ‘men’s club’, as Pierre Vidal-Naquet termed it. Never did the Roman woman receive authorization to take on the general nature of an ‘office’, a male task par excellence.

This raises several questions. Are there any matrimonial alliances that are contracted without marriage, without a more or less ceremonial act? Indeed, this is the case in many hunter-gatherer societies and among certain agriculturalists. The man and the woman begin by living together, and then their status gradually changes over time. It cannot be said that they go from being not married to being married, since marriage does not exist. What is important is that this union becomes publicly known and no one opposes it, no one finds anything to say against it on either the man’s or the woman’s side or in their respective communities. This is also the case with the millions of couples who live together without being married in the West and who declare their children with the state representatives near their place of residence. Their children automatically become citizens of the country where they were born, members of a nation whose boundaries are broader than those of their family and their local community. The fact of becoming the parents of legally recognized children gives these unmarried parents the rights and duties that the state confers on all relatives in the direct line, whether or not they are married.

If there are alliances without marriage, are there also alliances that are not marked by the transfer of goods and/or services between the families involved? In Western societies, where young people who are of age can marry without their family’s consent, many unions are contracted without an exchange of prestations between the families or between the spouses, except for a few reciprocal gifts. In many nomadic hunter-gatherer societies, which do not amass material possessions – among the Bushmen for instance50 – the young man moves in with his wife’s people until the birth of their second or third child. During this time, he shares the spoils of the hunt with his in-laws and renders them many services. Then, he may, if he wishes, return to his native band where he has kept his rights, taking his wife and children with him. Among the Purum agriculturalists of Manipur Province in India, the husband lives with his in-laws for three years and works off his marriage payment. Then he goes home to his own people.51

Lastly, do these alliances, sanctioned by marriage or not, always produce conjugal families? Among the matrilineal Ashanti, we saw that the husband visits his wife in the daytime, but spends his nights with his mother, his sisters and their children. His own children live with their mother. The couple’s residence is therefore matrilocal for the woman and duolocal for the man.

The Na are a Tibeto-Birman-speaking group, living in the Himalayan foothills of Yunnan and Sichuan provinces in southern China, who are matrilineal and matrilocal. Their residence unit is made up of groups of sisters living with their children of both sexes and with their brothers, the children’s maternal uncles. They do not have marriage. The men leave their sisters at night to visit the women in the neighbouring houses who have accepted them as temporary lovers. Even in the case of a long-lasting liaison, each man or woman can have other amorous relations at the same time and can separate when he or she chooses. Conjugal families are very rare, and if their number has grown in the past decades, it is due to pressure from the Communist rulers, hostile to the ‘furtive visits’ and concerned with imposing monogamy, a mark of civilization and of course of the superiority of socialism. Even in this extreme case, though, where there is no marriage or official direct exchange between families, an indirect exchange occurs and an adelphic family is formed in which it is the women who provide the children with their identity. The man is like the rain, a shower that awakens a seed-child in the woman’s womb, where it then develops. In Na society, then, there is no marriage, and therefore there are no husbands and no fathers. But the family exists, an adelphic family where the incest taboo between brother and sister is primordial and where any sexual allusion within the walls of the home is forbidden. The mothers and the maternal uncles exercise their authority jointly over the children engendered by the women of the house.52 The residence pattern is the same for men and for women: matrilocal.

RESIDENCE MODES

A great diversity of family types is engendered by the conjunction of different descent rules and different residence patterns. In matrilineal societies, residence can be matrilocal (Na, Rhades, Tetum), uxorilocal (Hopi), duolocal (Ashanti and Senufo), avunculocal (the family lives with the wife’s mother’s brother: Trobrianders), virilocal (the wife lives with her husband’s family). Patrilineal societies usually have residence patterns that are patrilocal (the family moves in with the husband’s father: Melpa, Baruya, Tallensi) or virilocal (the family settles on the husband’s land: Wolof,53 Tamil54 or Reunion Island). In Dobu,55 in southeastern New Guinea, a couple alternates residence according to years, sometimes patrilocal, sometimes uxorilocal. Cognatic societies often combine these principles, since one can choose to live with one’s maternal or paternal kin. In Samoa, on the other hand, women leave their village for that of their husband (virilocal residence). Finally, in Western Europe, Japan, the United States, but also among the Inuit, residence is neolocal, the couples choosing their place of residence without reference to their parents. It is easy to see that the various forms of residence have different effects on the children’s socialization, as they live closer to their paternal or their maternal kin, or find themselves surrounded by everyone, seeing their father in the daytime, their maternal uncle at night, or their maternal uncle in the daytime but never at night, etc.

RESIDENCE PATTERNS Matrilocal Uxorilocal Duolocal Ambilocal Avunculocal Patrivirilocal Virilocal Neolocal Natolocal

POLYGAMY AND POLYANDRY

Lastly we will mention some other principles that also help determine different family and group structures: polygyny and polyandry (the possibility for a man to have several wives – in Islam he is allowed four, plus concubines – or for a woman to have several husbands). Polygamy is widespread in Africa, among the Muslim populations of Asia, and in Melanesia. However, it is lessening with the Christianization of these populations, which imposes monogamy and restricts (the Orthodox Church) or forbids (the Roman Catholic Church) divorce. Polyandry remains limited to certain regions of the Himalayas, India, Amazonia and Oceania. A very rare and perhaps unique case in North America were the Shoshone, who practised both polygyny and polyandry.56 Polyandry can be adelphic (Tibet57) or not (Guayaki58). In the first instance, a woman marries a group of brothers, and the children are attributed successively to each of the brothers, beginning with the oldest, or are all regarded as descendants of the oldest brother. The main reason for adelphic marriages is to avoid dispersing family assets. In the second case, a woman has several husbands who are unrelated to each other, and the children are attributed to each man in succession.

Let us mention also the importance of the age of the persons getting married. Among the Siberian Chukchee59 a young woman can ‘marry’ a three-year-old boy, whom she will raise along with the children she conceives with her ‘authorized lovers’. The young Arapesh girl60 is betrothed at a very early age, around six or seven, and will go to live with her future husband’s family, where she will be brought up by her in-laws. In Australian Aboriginal society, the age difference may be as much as fifteen years or more, and here, too, the husband may raise his wife as his daughter, as it were. Finally, large differences are introduced in the internal functioning of families, in their members’ behaviours and in the exercise of kinship in general, depending on whether or not divorce is allowed to end the marriage that brought about the family (divorce or separation, since for co-habiting couples divorce is meaningless because there was no marriage to begin with). Whether or not individuals are allowed to remarry after divorce and on what conditions also has an influence. The question arises, too, of the remarriage of a widow or a widower. In the Christian West, because marriage is a sacrament for Catholics and the spouses are supposed to become ‘one flesh’ after their marriage, the bond cannot be dissolved and divorce is forbidden. Divorce was also banned among the Incas,61 and in present-day India it is almost unknown, even if it is permitted by law. The woman who asks for a divorce finds it hard to remarry. And if a man divorces, he runs the risk of having to return his wife’s dowry. Divorce is forbidden in Baruya culture. A man can repudiate his wife, but in this case he gives her to a brother or a parallel cousin, who takes her as a second or third wife.

DIVORCE

Divorce exists in numerous societies and is sometimes practised so intensively that an individual marries and divorces several times, which entails the appearance and disappearance of a succession of more or less reconstituted families. Generally speaking, the fate of the children after their parents’ divorce or separation is decided by custom. Among the Touareg, custom dictates that the sons go with their father and the daughters with their mother. In matrilineal societies, since children belong to the mother’s and not to the father’s lineage, divorce is much more frequent than in patrilineal societies. This is the case with the Trobriand Islanders, the Hopi and so forth.

BACHELORHOOD

A last word on bachelorhood and the status of bachelors in most societies. In many societies, for example the Baruya, it is unthinkable and forbidden not to marry. All individuals, unless they are gravely handicapped, must marry. Among the Inca62 all men having reached twenty-five years of age and all women having reached fourteen were supposed to be married or betrothed. The imperial administration took systematic population censuses and forced those who delayed to marry, sometimes even appointing a spouse. Nevertheless, remaining unmarried is valued in many societies when it is associated with the exercise of an important social function – religious or other – that demands partial renunciation of sexuality and the responsibilities of founding a family. This is the case among the Duna of New Guinea and other groups,63 where the masters of the male initiations remain unmarried for life but secretly marry a spirit-woman, who controls the fertility of the land, the abundance of the game in the forest and who is supposed to be without a vagina. The man is thus a bachelor in his village but a married man in the forest, where he spends most of his time, safe from the pollution and dangers entailed in sexual relations with women. In medieval Western Christendom, after the split between the Roman and Orthodox Catholic Churches, the Roman Church imposed celibacy on priests and monks, who found themselves married with the Church, which was represented as the mystic bride of the crucified Christ. Nuns became the brides of Christ, as attested by the ring on their finger. Like the Duna master of initiations, they were at once virgins among humans and married to a god. This is also the case with the traditional Indian ‘renouncers’.64

As for those bachelors who had no good reason to forego marriage, their status was usually frowned upon in Oceania and Africa. Given the sexual division of labour, an unmarried man had to depend on women to survive – his sisters, his mother or other men’s wives, for example those of his brothers or his uncles. Not having a wife, he could be tempted to take too much interest in the wives of others. And above all, the choice to remain a bachelor is usually regarded as a refusal to do what one is supposed to do to ensure the continued existence of a lineage, of a family, namely to marry and have children. These criticisms were usually even stronger in the case of a woman who refused to marry.

It is important to recall that in the Christian West, the Church not only promoted the celibacy of priests, proscribed divorce and made it difficult for widows to remarry, it also forbade adoption, which reappeared in the various European legal systems only at the end of the nineteenth century.

The adoption of children from outside (and not children who were orphaned or abandoned by their parents) gives children who have no genealogical ties to their adoptive parents the status of descendants. The status of parent in this case is a purely social relationship, devoid of any biological basis, which rests, as Maine said, on a ‘legal fiction’. In forbidding adoption, the Church thus helped promote a model that reduced kinship to essentially genealogical, that is to say biological, ties, even if these ‘carnal’ ties were made sacred by the sacraments of marriage and baptism. The paradox is that, at the very moment when the Church forbade adoption and its social fictions, it promoted another type of entirely imaginary parenthood, a spiritual one, which grew out of the institution at the end of the sixth century of the baptism of children at birth, which eventually replaced adult baptism. The sixth century also saw the institution of godparents.65 The child taken into the Church through baptism is accompanied by a spiritual father and mother, is sponsored by a godfather and a godmother. Theoretically a child’s godparents are even more responsible than the parents for their godchild’s spiritual upbringing. They are supposed to be particularly vigilant that the baptized child acquires the three Christian virtues of chastity, charity (caritas, which means more than the contemporary ‘charity’ and encompasses love of God and one’s neighbour) and uprightness. Godfather, godmother, godson, goddaughter: the Church that had banned adoption and its fictions replaced it with the fiction of the rebirth of children in the Church and chose for this imaginary filiation the vocabulary of genealogical filiation. Simultaneously the marriage prohibitions linked with the incest taboo applying to real kin were extended – with a variety of modalities – to spiritual parents and spiritual children. Initially prohibited to the seventh degree and then to the fourth degree of consanguinity, extended to close affines, husband’s brother, wife’s sister, etc., incest in the Christian West would threaten a second domain, that of the imaginary descent ties between Christians and their God.

Alliance, marriage, simple socially recognized union, with or without exchange, always raises the problem of who one can marry. We already know one can marry persons and into groups that do not come under the incest taboo as it is defined by a given society, or under other taboos which further extend this field: one must not repeat one’s father’s or one’s brothers’ marriage, etc. To these taboos within the field of kinship are added others that originate elsewhere: one must not marry outside one’s religion, class or rank. Which means that it is preferable or mandatory to marry someone of one’s own caste, rank and religion.

There are also a great many kinship systems where positive rules are added to the prohibitions and indicate whom it is prescribed or preferable to marry, which often entails the exchange of persons or goods. These systems are at the far pole from the Western European and Euro-American systems (which are cognatic with an Eskimo-type terminology), where, with the exception of a small group of consanguines and close affines, no one is prescribed and no one is forbidden – in terms of kinship, though not in terms of wealth, rank, name, etc. We are indebted to Claude Lévi-Strauss for having been the first to try to classify kinship systems in terms of the presence or absence of a positive marriage rule contained within the system as part of its structure. His analysis led to distinguishing three classes of systems:66

(1) Systems that make a positive statement about the class and terminological categories in which Ego can and must find a spouse, where exchange is forbidden between parallel kin but allowed and even prescribed between cross kin. For Lévi-Strauss, these systems, which he called ‘elementary structures of kinship’, are based on two types of exchange, depending on whether the wife-givers are or are not takers. In the first case, we are dealing with restricted exchange, in the second, with a generalized exchange, between kin groups.

(2) Systems that multiply marriage taboos and have no positive rule for choosing a spouse. Ego cannot take a wife in his father’s, his mother’s, his father’s mother’s or his mother’s mother’s clan, lineage or line, nor marry a certain number of cognatic kinswomen. However repetition of alliances with these same groups is not only allowed, it is sought, after a certain number of generations, as soon as distant consanguines can once more become affines. As a typical example of this category, Lévi-Strauss cites the Crow and the Omaha systems, later studied in more detail by Françoise Héritier.67 The findings of the latter’s analysis led her to object to the general thesis Lévi-Strauss used to characterize alliance.68 Héritier showed that, if it is true that two brothers or two sisters cannot marry in the same direction, a brother and a sister can do so, which means that in every other generation an exchange of ‘real or classificatory sisters is possible without violating any rules’. She thereby showed that elementary forms of restricted exchange are present in Crow–Omaha systems, even if they are concealed beneath numerous tacit prohibitions. Radcliffe-Brown’s sacrosanct rule – the genealogical equivalence of siblings – therefore does not apply in this case. Cross-sex siblings can do what parallel-sex siblings cannot; a brother and a sister do what two brothers or two sisters cannot do: marry in the same direction, replicate an alliance.

Alongside Crow-Omaha systems, which many anthropologists, such as Viveiros de Castro, continue with some reason not to regard as genuine, specific ‘systems’, Lévi-Strauss placed in the category of semi-complex systems and without analyzing them, the Iroquois and Hawaiian systems. And the matter has been left up in the air ever since. We will return to this issue later when we analyze the difference between Iroquois and Dravidian systems. The Baruya, as we will recall, have an Iroquois system, practice direct sister exchange, forbid the repetition of marriages before several generations, and diversify their alliances in each generation so that two brothers do not marry in the same direction or into their mother’s or their father’s lineages. According to this description, the Baruya should be placed in societies with semi-complex structures. But in their case, it is their ‘regime’ of ‘marriage’ that would be ‘semi-complex’ and not their kinship system or terminology.69

(3) Systems that fall into ‘complex structures’ of kinship, where marriage prohibitions concern ‘kin positions defined by their degree of proximity to Ego’. The formula corresponds specifically to cognatic systems with kindred, like the Western European systems and those of the Canadian Inuit. In this case, beyond the more or less narrow circle of prohibited kin, other criteria that have nothing to do with kinship intervene in the choice of a spouse and therefore in marriage strategies, if these exist. With complex systems, kinship, according to Lévi-Strauss, ‘leaves the determination of the spouse to other, economic or psychological, mechanisms’. We have seen that this was true of the Melpa, who sought to use marriage to make an affine into a moka partner or to make a moka partner into an affine.

Once again, combining the word ‘complex’ with the word ‘structure’ is not the best solution. When it comes to structure, contemporary Western systems, the Canadian Inuit system,70 or that of the New Guinea Garia are not complex, or much less, in any event, than the Australian or Iroquois systems – like those of the New Guinea Yafar or the Ngawbe in Costa Rica. What is complex is the variety of criteria other than kinship that determine the spouse and eventually the marriage strategies these various criteria can inspire in certain social strata or classes. In any event, today we see more clearly that kinship does not suffice to organize any society. For example, whether one considers the Australian Aborigines, the Baruya or the South Indian Pramalai Kallar,71 one cannot get married if one has not been initiated. In Europe, until recently, a young man did not marry before he had done his ‘military service’ or reached his majority. Kinship is always subordinated to other social relations, placed in the service of other goals than that of reproducing kinship.

THE ELEMENTARY STRUCTURES OF KINSHIP

Let us come back to this classification of kinship systems in order to illustrate a few examples and point out some of the problems they raise.