8 MARRIAGE AND ALLIANCE They are our enemies, we marry them. — Nuer proverb Seen from a male point of view, women are a scarce resource. No matter how male-dominated a society is, men need women to ensure its survival. In matrilineal systems, the men's sisters do this; in patrilineal societies, their wives do it; and in cognatic or bilateral societies, sisters and wives each do part of the job. A man can have a nearly unlimited number of children - in theory, he can beget several children every day - while a woman's capacity is limited to one child per year under optimal conditions, and moreover in many societies many children die before they grow up. From the perspective of human reproduction, one may thus state that sperm is cheap while eggs are expensive. This fact may be a partial explanation of the widespread tendency to the effect that men try to control the sexuality of women, as well as the tendency for men to regard the women of the kin group as a resource they do not want to give away without receiving other women in return. There may be several reasons why men in most societies want many children. They often need the labour power of the children for their fields or herds; and children can also form the basis of political support or be seen as an old age insurance policy. There are also biological explanations for the male 'drive to reproduce'. In many societies, polygyny (where a man has several wives) has been widespread. Polyandry (where a woman has several husbands) is much rarer. In fact, in the Ethnographic Atlas (Murdock 1967), a large database with comparable statistics on 863 societies, polyandry occurs only four times. Now, regarding the marriage institution as such, its rationale is evidently, at least partly, its ability to produce and socialise children. Comparatively speaking, romantic love is rarely seen as an important precondition for a good marriage. Rather, marriage is frequently arranged by kin groups, not by the individuals concerned; if the parties happen to like each other, this may be seen as a kind of bonus. Whether or not persons choose their spouses, marriage is very commonly perceived as a relationship between groups, not primarily between individuals. The ideology prevalent in 'Western' societies to the effect that marriage should be built on pure love, which may even transcend class boundaries, is peculiar if seen in a comparative perspective. Among the Maasai, for 108 Marriage and Alliance 109 example, the famous cattle nomads of East Africa, it is seen as a distinctive disadvantage if the romantic love between the spouses is too powerful. In this society, marriage is chiefly seen as a business relationship, the purpose being to raise children and make the herd grow. If the spouses fall in love, the result may be jealousy and passionate outbursts with adverse effects on business. Many Maasai women regard marriage as a necessary evil (Talle 1988). On the other hand, it is not true, as some believe, that high divorce rates exist only in modern societies. Divorce occurs in most societies in the world, and some 'traditional' peoples have higher divorce rates than the inhabitants of Hamburg. DOWRY AND BRIDEWEALTH In European and some Asian societies, the dowry has traditionally been an important institution (it is sometimes described as an 'Indo-European institution'). It means that the bride brings gifts from her family into the marriage, often household utensils, linen and other things for the home. The institution can be seen as a compensation to the man's family for undertaking to support the woman economically. A dowry can also be an advance on inheritance. In some societies, the payment of dowry entails a considerable economic burden. The costs associated with daughters getting married are a main cause of the high rates of female infanticide in India. Bridewealth (sometimes spoken of as 'bride-price') is more common than dowry in many societies, particularly in Africa. Here the groom's kin is obliged to transfer resources to the bride's kin in return for his rights to her labour and reproductive powers. The payment of bridewealth establishes the rights of the man in the woman and her children. If the bridewealth is not paid, the marriage may be void, and disagreement over bridewealth payments is traditionally a common cause of feuds among many peoples. In societies where bridewealth is common and the agnatic kin group is strong, the levirate may occur. This means that a widow marries a brother of the deceased (the levir), and in this way the patrilineage retains control of the woman and her children after the husband's death. The sororate, where a widower marries a sister of the deceased, is not a simple inversion of the levirate: in most cases it means that the woman's kin group commits itself to replacing the dead woman with a living one. Payment of bridewealth creates several kinds of moral bonds between people. First, it creates a contractual tie between lineages, being a sign of mutual trust. When the bridewealth is paid over a long period, for example through bride-service whereby the groom works for a certain period for his parents-in-law, the bonds are strengthened further. Second, the system of bridewealth strengthens solidarity within the paying group. Frequently, several relatives must contribute to the payment of the price, and often the 110 Small Places, Large Issues groom must borrow from his relatives. Such loans may create long-term debt and profound obligations on the part of the groom towards his lineage relatives. MOIETIES AND MARRIAGE Exogamous groups must by definition obtain women from outside. It is a fact that property, inheritance and political office tend to follow men in most societies, and that men often take the formal decisions regarding who is to marry whom. So even if the pattern of residence should be uxorilocal (that is, the groom moves in with the bride's family), the woman's brothers and other male relatives tend to determine her matrimonial destiny, even if they live with their wives in a different village. The simplest form of woman exchange would consist in the exchange of sisters: I give my sister to you, and you give me yours in return. In lineage societies, it is corporations rather than persons who exchange women. If a society consists of two kin groups who regularly exchange women between them, the society is divided into moieties. Frequently, moieties have a division of labour in addition to exchanging women. The moiety system of exchange is widespread among Australian peoples. In studies of these marriage systems, it has been pointed out that the outcome of a moiety system is eventually a kind of classificatory cross-cousin marriage. It happens like this: in a fairly small group, like the Kariéra of Central Australia, all members of society define themselves as relatives. They reckon patrilineal descent and are organised in two exogamous 'marriage classes'. They can marry anyone of the right gender who is not classified as a sibling. The Kariéra, like the Yanomamö and many others, have a classificatory kinship terminology, which means that they use a single term to describe many different persons, in this case everyone belonging to the same gender, generation and clan, independently of biological kinship. The Kariéra thus use the same term to describe a father, his brothers and other males of the same generation and same clan. One cannot marry persons considered as siblings, a category which includes those analytically labelled classificatory parallel cousins (father's brother's and mother's sister's children). On the other hand, father's sister's children and mother's brother's children, and everyone included in the same category, which we would call classificatory cross-cousins, are marriageable. Seen through a certain period, this kind of system takes on the form of a moiety system based on two patricians which exchange women between them. A man marries where his father married, which is into his mother's patrician. Both father's sister's children and mother's brother's children belong to this clan, since father's sister also married into that clan. A similar example, which may further illustrate the logic of exchange within a moiety system, is provided by the Yanomamö. A Yanomamö man marries a person classified as father's sister's daughter and/or mother's Marriage and Alliance 111 brother's daughter. A woman, similarly, marries a person classified as father's sister's son and/or mother's brother's son. The patrilateral parallel cousins belong to one's own group, as do the matrilateral parallel cousins, since mother's sister by definition is married to father's brother. Remember that we are talking about a classificatory kinship system and not a system which distinguishes terminologically between biological kin and others. The Yanomamö use the term suaböya about all marriageable women, who are classificatory mother's brother's daughters and/or father's sister's daughters. However, although there are only two kinds of same-generation women in Yanomamö terminology - wives and sisters - they distinguish in practice between 'close' and 'distant' cross-cousins. Many parents therefore try to marry their children into lineages with whom they want to forge alliances. Through a statistical analysis of several Yanomamö villages, Chagnon (1983) has argued that political stability is highest where the biological kinship bonds are strongest. Obviously, the members of groups which have exchanged women for several generations are related in more ways - both in terms of kinship and other obligations - than persons who have a purely classificatory kin relationship. Further, it is obviously in the interest of women to marry 'close' cross-cousins as they live in the same village as themselves. Thus the women can be close to their brothers, whom they may need for protection. The ideal model of cross-cousin marriage among the Yanomamö, as depicted in Figure 8.1, would create a very stable system where the inhabitants of the shabono were very close relatives. However, in practice the Yanomamö are often forced to develop links beyond the confines of the village, both to reduce the danger of war (see Chapter 11) and to look for wives. As a consequence, the inhabitants of the shabono are less close relatives than they would ideally be, according to Chagnon's biologically oriented model for analysis. • = A ó = A Ego Fi gure 8.1 Bilateral cross-cousin marriage among the Yanomamö The shaded persons belong to ego's patrilineage. The model is strongly mplified; in reality, a much larger number of persons would be involved. YYl Small Places, Large Issues EXCHANGE AND RANK DIFFERENCES Many peoples traditionally practise the cyclical exchange of women between more than two groups, so that, say, clan A gives women to clan B, which gives women to clan C, which gives women to clan D, which in turn gives women to clan A. Within this kind of system, a woman can only be 'paid for' with another woman. A system where three or more groups are mutually linked through some kind of cyclical exchange of wives may be on a larger scale than moiety systems, since it depends on a greater number of relationships to function. Such a system, where one distinguishes categorically between wife-givers and wife-takers, is called an asymmetrical alliance system, whereas moieties constitute a symmetrical alliance system. While the latter implies equality between the groups, an asymmetrical alliance often, but not necessarily, implies rank differences between the groups. The Kachin of upper Burma practise exogamy at the level of the patrilin-eage (Leach 1954). Their rules for wife exchange reveal a more complex and more hierarchical social organisation than that of the Yanomamö. Among the Kachin, wife-givers (mayu) have higher rank than wife-takers (dama). The Kachin, who are rice cultivators, are divided into three main categories of lineages: chiefly, aristocratic and commoner. Women move downwards within this system as every lineage is mayu to those with lower rank than themselves. The dama is obliged to pay bridewealth to its mayu, but is usually unable to pay immediately. Frequently, therefore, the groom has to work for years - sometimes for the rest of his life - for his higher-ranking parents-in-law. In this way, since wives are 'expensive', the rank differences between mayu and dama are reproduced and strengthened through time. All of the examples so far have dealt with exogamous marriage systems. Group endogamy also exists, particularly in highly stratified societies where considerable resources are transmitted through marriage. European royal families and Indian castes are thus known to be endogamous. However, we should remember that endogamy and exogamy are relative terms. All peoples are exogamous at least at the level of the nuclear family; conversely, few peoples would encourage their children to marry anybody without any discrimination. Even in societies where individual freedom of choice is stressed as a virtue, such as the United States, 'race endogamy' is common. DESCENT AND ALLIANCE THEORY A principal point in the study of marriage rules and practices concerns politics, alliances and stability. Since all groups are exogamous at some level, marriage necessarily creates alliances outside the nuclear family, the lineage or the clan. These kinds of alliances have been emphasised by many anthropologists, who have implicitly or explicitly argued against those who regard descent and lineage-based solidarity as the most fundamental facts of kinship. Marriage and Alliance New Guinea and the Anthropologists New Guinea is the second largest island in the world, with a total area of 810,000 square kilometres (the size of Great Britain is 244,046 square km). The population numbers about 3.5 million, and, since 1975, the island has been divided between the western half, Irian Jaya, which belongs to Indonesia, and the eastern half, the independent state Papua New Guinea. New Guinea has a great number of indigenous species of plants and animals, that is species which do not exist elsewhere. The landscape is dramatic and varied, containing barren swamps as well as jagged mountains and deep valleys which make large parts of the island relatively inaccessible. However, human settlement in both lowland and highland New Guinea dates back several thousand years. Most of the many hundred ethnic groups of New Guinea are traditionally hor-ticulturalists, who have settled in scattered pockets from the coast to valleys located up to 4,000 metres above sea level. Many of the peoples, especially in the highlands, keep large herds of pigs. The linguistic variation in New Guinea is exceptional. Over 700 languages are spoken, and 500 of them - the highland languages - do not seem to be related to any other language groups and are also, in most cases, mutually unintelligible. The coastal areas, where Melanesian languages are spoken, have been known to outsiders for centuries, both to Malayan and Indonesian seamen and, later, to Europeans. The highlands were virtually unknown until recently. Actually, they were generally assumed to be uninhabited until a group of natural scientists, in the early 1930s, by pure chance discovered a large people, the Enga. Crossing a hilltop just before dusk, the expedition was amazed to discover a fertile valley full of little fires and neatly cultivated gardens. Although missionary activity, the state and the monetary economy have come to influence life in the highlands, especially since the 1960s, many aspects of traditional culture and social organisation remain strong. Since the discovery of the New Guinean highland peoples by Europeans, New Guinea has been the object of intense attention by anthropologists, who immediately saw the island as an enormous resource for the young comparative science of culture and society -containing, as it did, many relatively isolated stone age peoples displaying a great cultural variation. The inflow of anthropologists has led to a certain irritation among many New Guineans, who feel that the anthropologists see them as ethnographic curiosities or even as relics from a bygone age. 114 Small Places, Large Issues Some influential, classic studies of kinship, notably Evans-Pritchard's (1940) and Meyer Fortes's (1945) studies of the Nuer and the Tallensi, respectively, focused strongly on descent-based corporations. They showed how groups with shared unilinear descent - be it factual or fictitious - were cohesive and could be mobilised politically (see Chapters 7 and 11). This corporate group, united through shared ancestry, was seen as the fundamental fact of kinship in stateless societies. Several anthropologists reacted against the elegant logical models of segmentary clans presented by this group of Africanists (Kuper 1988). In particular, this was the case among those who had done fieldwork in New Guinea, where it had been expected that the patrilineally based communities would be organised in segmentary lineages. However, it transpired that New Guinean societies included persons who did not belong to the patrilineage, and that they lacked the mechanisms for fusion and fission that had been described for the Nuer and the Tallensi (Barnes 1962). The Chimbu of highland New Guinea, for example, could just as well be described either as a cognatic system with a patrilineal basis, or as a patrilineal system with many exceptions. Thus the general validity of the models proposed by Evans-Pritchard, Fortes and others was questioned on empirical grounds - and it was concluded that they had probably exaggerated the importance of the unilineal descent groups at the cost of underestimating the importance of cognatic and affinal (in-law) ties. An interesting detail in this regard may be the fact that the anthropologists who focused on the structured, systemic aspect of kinship were associates of Radcliffe-Brown, while the critics who stressed the primacy of practice over abstract structure, notably Audrey Richards, were in many cases students of Malinowski. ELEMENTARY AND COMPLEX STRUCTURES In a justly famous study of kinship, The Elementary Structures of Kinship (1969 [1949]), Lévi-Strauss challenges descent theory in a more theoretical way. He does not regard shared descent, but rather the development of alliances between groups through the exchange of women, as the fundamental fact of kinship. Taking his cue from structural linguistics (which stressed relationship as fundamental to language) and the sociology of Marcel Mauss, where reciprocity was emphasised as a basic mode for humanity (see Chapter 12), Lévi-Strauss develops a highly original view of the institution of kinship. Indeed, he argues that the very formation of society occurs when a man gives his sister away to another man, thereby creating ties of affinity. A central element in Lévi-Strauss's perspective is the idea that all kinship systems are elaborations on four fundamental kin relationships: brother-sister, husband-wife, father-son and mother's brother-sister's son. Lévi-Strauss regarded this 'elementary structure', or 'kinship atom', inspired by similar structures from structural linguistics (see Figure 8.2), as Marriage and Alliance 115 fundamental to kinship and to human society as such. Some societies are constructed directly on the 'elementary structure', including societies based on classiflcatory cross-cousin marriage as well as societies based on asymmetrical alliances. 'Complex' systems, in Lévi-Strauss's terminology, add further relationships to the four fundamental ones as determining factors in marriage. He emphasises that elementary systems have positive rules; they do not only specify whom one cannot marry, but also whom one can marry (as among the Yanomamö). Complex systems, prevalent in modern societies and based on individual choice, have only negative rules and are therefore unable to create long-term alliances between kin groups. A ó ľA - A + Figure 8.2 The kinship atom The mother's brother is an important character in Lévi-Strauss's kinship atom. Granted the universality of the incest prohibition, and granted that men control women, the breeding of children ultimately depends on his willingness to give away his sister. Inspired by an earlier argument by Radcliffe-Brown (1952), Lévi-Strauss argues, further, that the relationship between a man and his maternal uncle is crucial. If the spouses are intimate, the wife will have a distanced relationship with her brother and vice versa. If one has a close, tender relationship with one's maternal uncle, the father will be a strict and severe person and vice versa. The 'severe uncle' usually, but not always, appears in matrilineal societies. Lévi-Strauss's argument is complex and covers much ground, both theoretically and empirically. An important point, pertinent to the earlier discussion about descent and alliances, is nevertheless that his line of thought implies that alliances between groups are more fundamental than shared descent. Affinality is thus a universal key to the understanding of the integration of society. The nuclear family, which was earlier considered to be the smallest building-block of kinship, becomes a secondary structure within this schema, since it presupposes the brother-sister relationship and affinality. PRESCRIPTIVE AND PREFERENTIAL RULES? Lévi-Strauss regarded the principle of cross-cousin marriage as a fundamental expression of reciprocity between kin groups with an H 6 Srna/] Places, Large Issues Claude Lévi-Strauss (b. 1908) is the founder of structuralism, an ethnographically informed theory about the ways in which the mind works. His most fundamental tenet is the principle of binary oppositions, the view that the mind organises the world in contrasting pairs and develops coherent systems of relationship from such a starting-point. Structuralism is chiefly influenced by two bodies of thought: French sociology, especially the work of Durkheim and Mauss, and structural linguistics from Ferdinand de Saussure, Roman Jakobson and others. Lévi-Strauss is a prolific writer who has synthesised and re-analysed enormous amounts of ethnography recorded by others, and his main works include The Elementary Structures ofKinship (1969 [1949]), The Savage Mind (1966 [1962]; see also Chapter 15), the four volumes of Mythologiques (1966-71) and the more personal, melancholic travelogue Tristes tropiques (1976 [1955]). Since he has argued that his very abstract models of thought and classification are universal, Lévi-Strauss has been subjected to severe criticism from more empirically oriented anthropologists, who have often pointed out that his general schemes do not fit their ethnography. There is nevertheless no doubt that Lévi-Strauss was the single most influential anthropologist in the period after Malinowski and Radcliffe-Brown - especially in France and other Latin countries, but also in the Anglo-Saxon world. It should finally be pointed out that his project, like Gregory Bateson's, is somewhat different from that of many anthropologists. Rather than trying to find out how societies work, or what makes people act as they do, the ultimate aim of his studies has been to reveal the principles for the functioning of the mind. He would therefore regard, say, kinship terminology not as a result of social organisation, but in the last instance as a product of the universal structures of the mind. elementary kinship system. These groups would also, according to him, have positive as well as negative marriage rules. Such elementary systems would also have unilineal descent systems and would exchange women at the level of the group. Rodney Needham, a translator and critic of Lévi-Strauss, held that the latter's model was only valid in societies with prescriptive marriage rules, even if the distinction between prescriptive and preferential systems (which Needham proposed) was not elaborated in The Elementary Structures of Kinship (Needham 1962). Lévi-Strauss rejected this view, and stated that he regarded the kinship atom as a universal elementary structure, that his theory about the exchange of women was valid for all unilineal societies and that the distinction between prescriptive and preferential systems was irrelevant. In practice, he argued, so-called prescriptive systems are prefer- Marriage and Alliance 117 ential, and in theory so-called preferential systems are prescriptive. Prescriptions thus only exist at the normative level, and in practice such rules are never followed perfectly. It thus seems necessary to distinguish between categorisations of persons one can and cannot marry (such as rules of exogamy) and cultural preferences concerning whom it is particularly beneficial to marry. To individuals, marriage practices may be perceived as prescriptive rules if their parents arrange the marriage, but at a societal level it would be misleading to use this model as a description of the overall practices. What Lévi-Strauss speaks of as prescriptive rules simply amount to the categories the members of society think through; Needham's distinction makes it possible to distinguish between these categories and the strategies actors follow to achieve specific, culturally defined aims. Even perfect knowledge of categories and rules does not enable us to predict how people actually will act, and at this point we might recall Firth's distinction between social structure and social organisation. Rules and norms are not identical with the social application of rules and norms. Most kinship phenomena can probably be interpreted from an alliance perspective as well as a descent perspective. Both alliances and descent are aspects of every kinship system, although, as Kuper (1988) has remarked, descent theorists largely concentrated on societies where agnatic lineages were particularly important in the organisation of society, whereas alliance theorists were more concerned with the study of societies where the forging of alliances between kin groups was crucial. It is nevertheless quite possible to identify important cross-cutting alliances in societies usually thought of in terms of descent groups. KINSHIP, NATURE AND CULTURE In many modern societies, it is customary to think of kinship in terms of biology. Europeans generally see themselves as more closely related to their siblings than to their cousins and more closely related to first cousins than to second cousins. Classificatory kinship seems to be more or less absent in this kind of society. However, it transpires that even this kind of society has kin terms which derive from social organisation rather than from biological kinship. Among the Yanomamö, all of the women of one's patrilineage are regarded as 'father's sisters', and all of the men in mother's patrilineage are regarded as 'mother's brothers'. In the parental generation of ego, only two kinds of men and two kinds of women exist: fathers, mother's brothers, mothers and father's sisters. Among the Kariéra and several other Australian peoples, all members of a moiety of the same generation and gender can be spoken of with the same kin term. All 'brothers' are brothers for nearly all practical purposes, even if they do not have shared biological descent. 118 Small Places, Large Issues In most European kinship terminologies, some affines are labelled 'uncles' and 'aunts', namely those who have married our parents' siblings. In many Indo-European languages, moreover, there is no terminological difference between biological and affinal uncles and aunts. The European kinship terms brother-in-law and sister-in-law may also refer to two different kinds of relatives. A brother-in-law may be the brother of ego's spouse; he may also be ego's sister's husband. Kin, in other words, do not come naturally; they must be created socially, and this is at least partly fashioned so as to facilitate tasks and to create order in an otherwise chaotic social world. Arguing against those who have emphasised the biological foundations ofkinship, Needham(1962) andSchneider (1984) have argued that kinship is an invention with no necessary connection with biological facts, and they both stress that it is the invention of anthropologists. At least, as the examples in this chapter have shown, the kinship system in a society does not follow automatically from biological kin relations. When descent is important in order to justify claims to land, it may be common to manipulate genealogies. Laura Bohannan (1952) has dealt with this in a study of the Tiv of Nigeria, an agricultural people organised in landholding segmentary patricians. In this society, the structure and origins of the lineage are frequently consciously manipulated for the benefit of the interests of the living. Anne Knudsen (1987, 1992), writing about kinship, vendettas and mafia in Corsica, shows that of the total number of cousins (male collateral kin) a person has, only a small proportion is socially activated. Only the kinsmen one has shared interests with are in practice reckoned as kinsmen. Frequently, those cousins who are genealogically the most distant ones, become the closest ones in practice. Geertz (19 8 8, p. 8) puts this openness of 'facts' to manipulation and interpretation in a more general way when he refers in passing to the North African mule, 'who talks always of his mother's brother, the horse, but never of his father, the donkey'. Despite the importance of the objections against a biologically based view ofkinship, it remains a fact that important forms ofkinship are universally framed in terms of biological descent, although other forms ofkinship - clas-sificatory, affinal, symbolic - may be more important in a variety of circumstances. SOME COMMON DENOMINATORS As we have seen, there are many different ways of resolving the problems associated with kinship, but all societies have some common denominators: all have rules regulating incest and exogamy. In all societies, alliances are forged between persons or descent groups, whether their importance is marginal or significant. All societies also seem to have developed a social organisation where mother and child live together during the first years of the child's life (a possible exception being societies with a high density of Marriage and Alliance 119 kindergartens). All societies have also developed functioning reproductive institutions, and all have rules of inheritance. Further, many societies have also developed forms of local organisation, with political, economic and other dimensions, which are based on kinship. Both religion and daily rules for conduct in such communities may be based on respect for the ancestors and ancestral spirits (see Chapter 14). Differences in power are also often related to kinship. Kinship, indeed, is often the master idiom for society and human existence. What, then, is the role ofkinship in societies which lack corporate kin groups, prescriptive marriage rules and ancestral cults? KINSHIP AND BUREAUCRACY It is doubtless correct that kin-based forms of organisation continue to be important in many societies after having gone through processes of modernisation, that is after the inhabitants have become citizens and taxpayers, wageworkers and TV audiences. In most modern states, family dynasties exist in the realm of finance (and sometimes in politics), and genealogies remain important to individual self-identity. The nuclear family is an important institution in modern societies, and in many such communities kinship is decisive for one's career opportunities, political belonging, place of residence and more. The capitalist labour market, however, is ostensibly based on formally voluntary contracts and individual achievement - not on kinship commitments and ascribed identity. It is therefore customary to regard the kin-based organisation as a contrast, and possible threat, to the bureaucratic organisation characteristic of both the labour market and the system of political administration in modern state societies. Kin-based organisation is based on loyalty to specific persons, while bureaucratic organisation ideally is based on loyalty to abstract principles, notably the law and contractual obligations. Kinsmen may be obliged to help each other out, whereas bureaucrats have committed themselves to following identical procedures and principles no matter who they are dealing with. According to a kinship ideology, it is appropriate to treat different people differently; according to a bureaucratic way of thinking, everybody is to be treated according to identical formal rules and regulations. When a person of high rank employs one of his kinsmen, others may call this practice nepotism (literally, particularism favouring nephews), that is, 'unfair' differential treatment on the basis of kinship. According to a kinship logic, however, such a differential treatment is not unjust but is rather an indication of loyalty and solidarity. The two logics, which coexist in virtually every society today, are thus difficult to reconcile in theory - they represent opposing moralities. Max Weber (1978 [1919]) was the first social theorist to write systematically about the differences between kinship-based and bureaucratic 120 Small Places, Large Issues organisation. His point of departure was the industrialisation of Europe, and he demonstrated a clear interrelationship between the Industrial Revolution, the growth of anonymous bureaucratic organisation based on formal rules and the weakening of kinship bonds. Although he was critical of some aspects of bureaucracy (he feared the inflexibility of the 'iron cage' of bureaucracy), Weber regarded this form of organisation, based on anonymous principles of equal treatment and a clear distinction between a person's professional and private statuses, as a distinctive advance over the particularistic principles that had dominated earlier. Talcott Parsons, who later developed Weber's theory further (1977), regarded modern societies as achievement-oriented and universalistic, as opposed to 'traditional' societies, which he saw as ascription-oriented and particularistic. This distinction means that a person's rank and career opportunities in a modern society depend on his or her achievements and achieved statuses, and that equal treatment for all (notably equal civil rights and equality before the law) is an important principle. In a traditional society, on the contrary, Parsons held that ascribed statuses, frequently connected with kinship, were more decisive; in other words, that it was less important what a person did than what he or she was. Dichotomies of this kind are always simplistic. First, it is definitely not true that particularistic principles are absent in modern societies. Second, anthropological research has shown that there exist many 'traditional' societies which are highly achievement-oriented, where individual achievements are more important than lineage membership. This is the case, for example, among many hunters and gatherers, as well as in highland New Guinea. Further, the very term 'traditional societies' is extremely inaccurate since it lumps together a mass of highly diverse societies - from a Quechua village in the Andes to the Chinese empire. On the other hand, dichotomies of this kind can be useful as conceptual tools, and, provided we do not confound them with descriptions of an empirical reality, they can be helpful in the process of organising facts. We should never forget, though, that they are ideal types (Weber's term); stylised, abstract models of aspects of the world, which are never encountered in their pure form 'out there'. The relationship between kin-based and bureaucratic organisation must always be explored in an empirical context. Then we will discover that the two principles very often function simultaneously; that they are not mutually exclusive in practice. A person may support both ideals of formal justice and kinship solidarity in different situations. METAPHORIC KINSHIP A lesson from the study of bureaucratic organisations is that the introduction of universalistic principles (formal rules, contracts, etc.) does not simply Marriage and Alliance j -, ^ do away with particularistic principles: the two sets of rules coexist, just as individualism has not made the family superfluous, although many of its former functions have been taken over by other institutions. Let us now consider if a kinship way of thinking may have survived in other, less obvious ways in modern state societies. Due to industrialisation and the integration of large, heterogeneous populations in nation-states, it has in many contexts become impossible to maintain clan- or lineage-based social organisation. In this kind of society, everybody is dependent on a large number of persons they are not related to, and each person is responsible for his or her life, largely without support from the kin group. The labour contract has replaced the clan land and the family trade, and social mobility is high. A marriage ideology based on individual choice has replaced the former lineage-based marriages. The monetary economy and the ideology of universal wagework has turned questions of subsistence and place of residence into individual and notcollective issues. This may lead us to believe that kinship has ceased to be important. However, it has important symbolic dimensions in addition to its social organisational potential. It is, in most known human societies, a main focus for subjective belongingness, sense of security and personal identity. In these fields, it seems clear that kinship has at least partly been replaced by metaphoric kinship ideologies such as nationalism. Nationalism presents the nation as a metaphoric kin group. Like lineage ideologies, it stresses the contrast between 'us' and 'them', and although it may be internally egalitarian and universalistic, it favours particularism in relation to other nations (see Chapter 18). The nation may also function as a de facto lineage in certain judicial contexts. If a citizen dies with no personal inheritors, the state inherits the estate. The state may also, in certain cases, assume the parental responsibility for children. A decisive difference between nationalism and actual kinship ideology is the fact that the nation encompasses a large number of people who will never meet personally; it promotes an anonymous community between people who do not know each other. If we wish to develop an ideal-typical distinction between societies of large and small scale, it may be useful to place the boundary at this point: if important aspects of one's existence depend on people one does not know, one belongs, in important respects, to a social system of large scale. KINSHIP IN ANTHROPOLOGY TODAY Whether metaphorical or not (and whether or not this difference makes a difference), kinship remains a core concern in anthropology. In his excellent overview of the anthropology of kinship, Holy (1996) reminds his readers that not all anthropologists agree about the ubiquity and universal character of kinship. However, since the days of Morgan and Maine, very many prac- 122 Small Places, Large Issues titioners of the discipline have seen it as a human universal. This view, Holy argues, rests on three assumptions: (1) That 'kinship constitutes one of the institutional domains which are conceived to be universal components or building blocks of every society' (Holy 1996, p. 151). The others, he adds, are an economic system, a political system and a system of belief. (2) The second assumption is the notion that 'kinship has to do with the reproduction of human beings and the relations between human beings that are the concomitants of reproduction' (p. 152). (3) Finally, there is the view that 'every society utilises for various social purposes the genealogical relations which it assumes to exist among people' (p. 153). Holy then goes on to show that all three assumptions are questionable: the degree and form of institutional differentiation varies from society to society; reproduction and biological relatedness carry varying meanings and social implications; and the ways and extents to which genealogical connections are traced, also vary considerably. Important variations between concepts of personhood and of relatedness may be glossed over by an over-insistence on the primacy of kinship, whether it is seen as chiefly biological or not. Be this as it may, the empirical salience of kinship in most societies -notwithstanding important variations - ensure its place as a main focus of anthropological research today, not least in studies of complex, modern societies, where its significance has probably been underestimated in social theory. The field of kinship studies is also, naturally, a main fighting ground between biological determinists and culturalists. Whatever complementarities may exist between biological or evolutionary perspectives on humanity and perspectives that posit the primacy of social constructions (and I believe these complementarities to be major), kinship has proved resilient to attempts at integrating these views. Few themes in anthropology provoke more heated debates than questions related to the biological versus the socially constructed in kinship. KINSHIP AND GENDER To round off these two chapters about kinship and marriage, it seems appropriate to linger briefly on the relationship between kinship and gender. During the heyday of'kinshipology', up to the 1960s, anthropologists were, with a few notable exceptions, not particularly interested in gender as a differentiating principle. When reading the classic studies of Boas, Kroeber, Malinowski, Radcliffe-Brown, Evans-Pritchard and Fortes today, the absence of analyses of gender and the social and cultural production of gender differences is striking. In studies of kinship, a male perspective is often taken for granted. Certainly, women have a place in these studies; they sometimes appear as wives, mothers and sisters, but rarely as independently acting persons. They appear as resources which society (that is, men) controls; they are exchanged between groups, are married, accused of witchcraft and so Marriage and Alliance 123 on. Additionally, classic anthropological studies of kinship have rarely explored how particular kinship systems create particular kinds of gender relations - what sort of ideology justifies men's power over women - or even reflected on the fairly obvious fact that a kin relationship is often a gender relationship as well. Today there exists a growing literature which tries to see social life from a gender-neutral perspective or even with an explicit female bias. Since the 19 70s, many important studies on the fundamental importance of gender as an organising principle in society and culture have been published, and some of these studies are discussed in the next chapter. However, surprisingly little of this literature links up with the study of kinship (see Collier and Yanagisako 1987; Howell and Melhuus 1993; Carsten 1997). For if Levi-Strauss is right in that the sister-brother relationship is fundamental in the social production of kinship, it is surely not without interest that this kin relationship is also a gender relationship. The following two chapters deal with various criteria, starting with gender, that are used to classify people into mutually exclusive categories, which more often than not entail differences in power. SUGGESTIONS FOR FURTHER READING Adam Kuper: The Invention of Primitive Society: Transformations of an Illusion, Chapters 10-11. London: Routledge 1988. Edmund Leach: Léví-Strauss, Chapter 6. Glasgow: Fontana 1970. Rodney Needham, ed.: Rethinking Kinship and Marriage. London: Tavistock 1971.