92 Small Places, Large Issues In several of his books on epistemology, Bourdieu criticises social scientists for overestimating the importance of representations and reflexivity in their comparative studies of society and culture. This cognitive, and especially linguistic, bias, Bourdieu argues, is characteristic of our occupational specialisation and tends to lead us to ignore the fact that the social world is largely made up of institutionalised practices and not by informants' statements. Other anthropologists (such as Bloch 1991) have also pointed out that the social world is under-determined by language; in other words, that there are large areas of social life and of cognition which are not only non-linguistic, but which cannot easily be 'translated' into language. The transmission of knowledge and skills, Maurice Bloch (1991) argues, consistently with Connerton, frequently takes place without recourse to language. Many cultural skills can only be explained by showing them in practice. In other words, if an over-reliance on interviews is a methodological pitfall, an overestimation of the linguistic character of the social world is an epistemo-logical error. We have now introduced some of the most fundamental theoretical issues of social science, including anthropology. It should be noted that after the critique ofstructural-functionalism in the 1960s, anthropology has made a distinctive move in two directions: first, there has been a shift from emphasis on structure to emphasis on process. Change is now seen as an inherent quality of social systems, not as an anomaly. Second, there has been a no less significant shift from the study of function to the interpretation of meaning. As an implication, anthropology has, in the eyes of many, moved away from the social sciences in the direction of the humanities. Be this as it may, it is beyond doubt that contemporary anthropologists often are cautious of positing explanatory accounts of social processes, and concentrate instead on understanding and translation. SUGGESTIONS FOR FURTHER READING Fredrik Barth: Models of Social Organization. London: Royal Anthropological Institute, Occasional Papers, no. 23 (1966). Peter L. Berger and Thomas Luckmann: The Social Construction of Reality. Harmondsworth: Penguin 1967. Mary Douglas: How Institutions Think. London: Routledge 1987. Anthony Giddens: Central Problems in Social Theory. London: Macmillan 19 79. 7 KINSHIP AS DESCENT No society (I believe) is bloody-minded enough to ban sex from marriage, and there is an obvious convenience in combining the two; but sex without marriage one can have and one does. — Robin Fox Generations of anthropologists have been flabbergasted at the intricate kinship systems existing in many 'primitive' societies. Several famous examples of such complicated systems are to be found in the Australian aboriginal population. These peoples, traditionally hunters and gatherers, have the simplest technology in the world. They lack metals, domesticated animals and writing, and in most cases they do not even have the rudiments of agriculture. Nonetheless, many of these nomadic groups have kinship systems so complex that it may take an outsider years to comprehend them fully. They can name a large number of different kinds of relatives, they have accurate rules determining who can marry whom, and the groups are subdivided into moieties, clans and sub-clans. The study of kinship has always been a core topic in anthropology. Towards the end of the 1940s kinship was so central, especially in British social anthropology, that despairing lay people (and students) spoke ironically of the subject as 'kinshipology'. Many non-anthropologists have reacted with incomprehension at the great interest in kinship still prevalent in the profession. What is it about kinship that makes it so important? The simple answer is that in very many societies kinship is the single most important social institution. The kin group, in many cases, takes care of one's livelihood, one's career, one's marriage, one's protection and one's social identity. Chapter 4 offered a first glimpse of the importance of kinship, and indicated that there is a close interrelationship between the kinship system and other aspects of social organisation. In some cases, what anthropologists delineate as the kinship system may indeed be coexistent with social organisation, since the members of society, and most of their activities, may first and foremost be organised along kinship principles. There are many social ways of organising, and thinking about, kinship. Although it is a widespread cultural notion in 'Western' societies that kinship is related to biology and blood ties (Schneider 1984), anthropological research generally analyses it as cultural classifications of people and as 93 94 Small Places, Large Issues aspects of group formation. This chapter introduces some central features of kinship seen as social organisation, and discusses different ways of reckoning kin. In the next chapter, the focus is on marriage systems and symbolic aspects of kinship. INCEST AND EXOGAMY All known human societies prohibit sexual relations between persons who are classified as close blood kin, which includes at least the father-child, mother-child and sibling relationships. This does not of course mean that such relations do not occur, but rather that there is a norm prohibiting it. This universal rule is often spoken of as the incest taboo. There are significant cultural variations concerning who is included in this taboo; in many societies, persons others might regard as very distant cousins are included in the prohibition. Sanctions against violations of the incest taboo are not universally strong; however, marriage between close kin is always strictly prohibited. Why is the incest taboo universal? Since the time of Tylor and Freud, several explanations have been put forward. Several anthropologists have pointed out the social advantages of the rule, including the expansion of the group through the inclusion of new members and the forging of alliances across kin boundaries (see Chapter 8). A functionalist explanation of the incest taboo, common among non-specialists, is that widespread incest would lead to biological degeneration, and that functional mechanisms preventing incestuous practices are therefore called for. This kind of explanation is not satisfactory. Notably, it does not explain what it is that makes people reject incest, since they are in many cases ignorant of its possible negative effects on the genetic material. Some anthropologists have argued that people who have grown up together will scarcely feel mutual erotic attraction, while others have invoked the term 'instinct' to explain why close kin do not feel sexually attracted to each other. Lévi-Strauss has argued that men divide the women surrounding them into two mutually exclusive categories, 'wives' and 'sisters', and that only the former are seen as potential sexual partners. He holds that the exchange of women between kin groups is, when all is said and done, an effect of reciprocity, which is a fundamental structure of the human mind, be it conscious or unconscious. Some anthropologists have even argued against the use of the term 'incest' at all (Needham 1971a), since its meaning varies cross-culturally. Most would nevertheless agree that it is a useful concept, referring to sexual intercourse between persons who are locally (emically) defined as close relatives. There are societies which prescribe their members to marry their relatives, though never the very closest ones. This kind of practice is called endogamy; Kinship as Descent 95 one marries inside the group. The opposite practice, whereby one marries outside of the group, is called exogamy. The two concepts are relative: the Yanomamö are endogamous at the level of the ethnic group (they scarcely marry non-Yanomamö) but exogamous at the level of the clan. They are divided into a number of exogamous clans (named groups with a common ancestor), and are required to marry persons who do not trace their genealogy to the same ancestor as they do. Since they reckon kinship along the male line (patrilineality), father's sister's children and mother's brother's sister do not belong to one's own clan (they are cross-cousins) and are thus classified as marriageable. Among many Middle Eastern peoples, on the contrary, marriage with father's brother's child (one's parallel cousin) is widely practised. In a sense, all human groups are both endogamous and exogamous to varying degrees: one is expected to marry 'one's own kind', but not someone classified as a close relative. Who is a close relative and who is not is naturally culturally specified, although the,people classified as parents, children and siblings in Europe are virtually everywhere seen as close kin. CORPORATE GROUPS Kinship concerns much more than the reproduction of society and the transmission of cultural values and knowledge between the generations, although these aspects are certainly important. Kinship can also be important in politics and in the organising of daily affairs. In many societies, a man needs support from both consanguineal kin (blood kin) and from affines (in-laws) in order to follow a successful political career. In other societies, family members join forces in economic investments. Among the Hindus of Mauritius, for example, it is common for groups of brothers and cousins to set up a joint business. Although there may be no rule to the effect that one has to be related to run a business together, kinship can give a practical advantage. One can usually trust one's relatives, since they are tied to oneself through webs of strong normative obligations. In many societies, and especially stateless ones, the kin group usually forms the basis for political stability and for the promotion of political interests (see Chapter 11). The group is tied through mutual bonds of loyalty and can often function as a corporation in situations of war, as well as in peaceful negotiations over, say, marriage payments or trade. In this kind of society marriage does not take place between single 'autonomous' individuals, but between groups. Inside the kin group, norms specify roughly how one is to behave towards different categories of kin. These norms prevent the dissolution of the group and ensure that people carry out their duties. The entire division of labour may thus be organised on a kinship principle. Corporate kin groups tend to be unilineal, which means that new members are recruited on a genealogi- 96 Small Places, Large Issues cal principle, either becoming a member of the father's kin group (patrilineal system) or the mother's kin group (matrilineal system). Peter Jane Henry Margaret i O A^O Susan Ronald I Molly Edward Jonathan A O David Mary Figure 7.1 Kinship symbols A triangle denotes a man, a circle a woman, and the equals sign that they are married (sometimes depicted by a horizontal line connecting the two from below). A horizontal line connecting two persons from above indicates that they are siblings. A diagonal line through a symbol indicates that the person in question is dead. In other words, Ronald is married to Molly, and their children are David and Mary. Ronald's sister is Susan, and Molly's brothers are Jonathan and Edward. Ronald's mother is Jane, and his father Peter is dead. Molly's parents are Henry and Margaret. INHERITANCE AND SUCCESSION Group membership, politics, reproduction and social stability have been mentioned as important aspects of kinship. A further important dimension of the kinship institution is the judicial one: it is no coincidence that a large part of the anthropological kinship vocabulary derives from Roman law. Blood is thicker than water, it has been said in many parts of the world (though not everywhere), and the famous return of the biblical lost son was celebrated with a great feast. Fratricide and patricide are considered the most serious crimes imaginable in many of the world's societies, and parents have always meddled in the marriages of their children - both before and after their consummation. These nearly perennial problems are tied to the fact that kinship is connected with inheritance and succession. Both institutions are to do with the transmission of resources from one generation to the next. Inheritance concerns the transmission of property, while succession refers to 'the transmission of office' (Rivers 1924); transmission of specified rights and duties as ascribed statuses. Kinship as Descent 97 All societies have rules regulating who is to inherit what when someone dies, although these rules are often contested or interpreted in varying ways. There is no universal link between the kinship system and the rules of inheritance in societies. There are patrilineal systems of descent where men and women are equals in terms of inheritance, and there are systems which give priority to one of the genders, usually the male. In some societies the eldest son receives a larger part of the inheritance than his siblings (primogeniture); others follow the opposite principle and give priority to the youngest son (ultimogeniture). Whereas the corporate principle functions in an integrating way, inheritance is a source of potential disruption, since it reveals conflicts of interest among the relatives. Rules of succession are often closely linked with the principle of descent. In patrilineal systems, a son (or a younger brother) will frequently take over the commitments of the deceased; in matrilineal systems, a man commonly succeeds his mother's brother. However, it should be remembered that many forms of succession do not follow genealogical principles at all, for example in societies where chiefs and shamans are appointed or elected on the basis of personal merit. This is the case among the Yanomamö and in many other small-scale societies. WAYS OF RECKONING KIN Formally, there are six possible principles for the transmission of kin group membership and other resources from parents to children (Barnard and Good 1984, p. 70). As already noted, the same principle does not have to hold true for succession, inheritance and descent, although the three are frequently lumped together, particularly in patrilineal societies. 1. Patrilineal. Transmission of membership and/or resources takes place unilineally through the father's lineage. 2. Matrilineal. Transmission of membership and/or resources takes place unilineally through the mother's lineage. 3. Double. Some resources are transmitted through the father's lineage, others through the mother's lineage. The two lineages are kept separate. 4. Cognatic. Resources can be transmitted through kin on both mother's and father's side (bilaterally). 5. Parallel. Rare variety whereby men transmit to their sons and women to their daughters. 6. Crossing or alternating. Rare variety which represents the opposite of the previous one: men transmit to their daughters, women to their sons. This simplistic typology should not lead anyone to believe that, for instance, persons in patrilineal societies 'are not related to' their mother's relatives. Practically all kinship systems organise kin relations on both the mother's 9 8 Small Places, Large Issues and the father's side, although rights, names and group membership frequently give priority to one side. In a patrilineal society, one's commitment to the father's lineage is, in most situations, stronger than one's commitment to the mother's. In many societies, moreover, classificatory kinship terminologies complicate any view of there being a simple relationship between biological and social relatedness. These terminologies lump together persons with varying degrees of biological relatedness under the same kinship term, so that e.g. all boys of the same generation and the same clan are called 'brothers'. COGNATIC OR BILATERAL DESCENT In most of Europe and North America, kin on both sides are in principle regarded as equally important. Both mother's and father's relatives of both genders are our relatives, and we often do not distinguish systematically between the two lineages terminologically. In English, for example, we do not have separate kinship terms for father's mother and mother's mother (but such distinctions exist in the Scandinavian languages for example). Most of these societies have traditionally given the father's side a certain priority, since the father's surname has been passed on to the children. However, in recent years many women have begun to retain their maiden name after marriage, and often the children's family name therefore becomes that of the mother. As shown in the example of Bremnes in the previous chapter, it is difficult to organise stable, tightly incorporated groups within the framework of a bilateral kinship system. The kin group cannot be clearly delineated, since ego's relatives will always have relatives to whom ego is unrelated. Most societies which are constructed about incorporated kin groups therefore base them on a unilineal principle. However, in some parts of the world, particularly in the Pacific, cognatic corporate groups exist. They are constructed on an eclectic basis, drawing pragmatically on both matrilateral and patri-lateral kinship. A cognatic or bilateral way of kin reckoning creates problems in the construction of genealogies as well. For each generation one moves back in time, the number of kin is doubled. We have two parents, four grandparents, eight great-grandparents, sixteen great-great-grandparents, and so on. Genealogies thus tend to be shallow in this kind of society - most persons are unable to name ancestors more than three or four generations back and, as mentioned, corporations based on kinship are relatively rare. Instead, it could be said that class endogamy, particularly among the upper classes, often forms the basis of group solidarity and a pooling of resources reminiscent of the logic of kinship corporations. For a long time, bilateral kinship received little attention in social anthropology, despite the fact that a third of the world's kinship systems are 100 Small Places, Large Issues bilateral. This could be a result of the tendency, particularly strong in British anthropology in the post-war years, to see unilineal descent as the main mechanism of social integration in traditional societies. PATRILINEAL DESCENT The lineage in a patrilineal system includes, at least, ego's siblings and father, father's siblings and the children of the men in the group. Father's sister's children, however, do not belong to ego's group, but rather to her husband's group. The lineage is usually larger than this, and its size depends on the structural or genealogical memory of the group. If one includes, say, every descendant of a shared ancestor ten generations back, the group will naturally be much larger than if one starts reckoning at an ancestor five generations back. This kind of difference is not caused by mere variation in memory or forgetfulness, but is rather related to organisational features of society. In a society with a long genealogical memory it is necessary, or at least possible, to organise fairly large kinship-based networks and corporate groups - much larger ones than societies which stop their kin reckoning two or three generations back. An example of the latter kind of society is the San hunters and gatherers of the Kalahari, where the social group is small and genealogies are shallow. Let us consider an instance showing what patri-lineally based kin groups do. When the Dogon settled where they live now, according to the myth, each village was founded by several brothers, who were the ancestors of one or several present-day clans (Beaudoin 1984). All the descendants of an ancestor live in the same hamlet, called a ginna. The word ginna is used about the clan's land, about the family house where the clan chief (ginna band) lives and about the smaller houses where the households are based. Most of the land is administered, and distributed among the heads of household, by the ginna bana. The concept of ginna thus groups what anthropologists see as property, place of residence, social rights and duties, and politics, in a single kinship term. Only men and their children are members of the lineage. The wives/ mothers belong to other lineages which are centred elsewhere. Marriage is usually organised by the two fathers, who strengthen informal ties of friendship in this way. The Dogon are exogamous at the lineage level. Divorce does occur; as the patrilineal principle is all-encompassing, the father and his lineage are entitled to keep the children when this happens. Since kinship is formally only recognised through the father, every Dogon has kinship obligations towards a limited number of relatives, notably close male agnates (patrilateral kin); that is, one's father, brothers and father's brothers. Kinship as Descent 101 In addition to the patrilineal principle of descent, the pattern of residence is virilocal. This means that the newly married settle in the man's household or at least in his ginna. In this way, all of the most important resources controlled by the Dogon men - land rights, politics, children and relatives -are concentrated in the same geographical place. The system seems to be most beneficial to the men, who control the most important resources. Women in patrilineal and virilocal societies are outsiders and are often associated with danger; in some African societies they are prone to be accused of witchcraft. Their own agnates may be far away in a different village, and in a certain sense they live among strangers throughout their married lives. Married women thus represent a threat to the cohesion of the lineage, since they are strangers within the community. This reminds us of the potential conflict, mentioned in Chapter 4, between the household and the lineage. The man's loyalty becomes divided, and sometimes difficult situations may arise, where he may have to choose between loyalty to his lineage and to his nuclear family. Patrilineal systems are capable of concentrating all valuable resources in a single principle, namely the principle of descent through the agnatic line: succession, inheritance, property rights, place of residence, marriage partner, children and political rights. Many societies, particularly in the Middle East and North Africa, further practise parallel cousin marriage, meaning that a male ego if possible should marry his father's brother's daughter. This system, if and when it functions according to the rules, certainly creates powerful and compelling forms of integration, since each individual will be related to other members of the in-group in several different ways. In the words of Emmanuel Todd, 'the endogamous community family is probably the anthropological environment which more than any other in the history of humanity integrates the individual' (Todd 1989, p. 140). We now turn to looking at matrilineal systems, which present a less tidy picture. MATRILINEAL SYSTEMS Although most peoples in the world are classified by anthropologists as either patrilineal or cognatic, many groups, particularly in Melanesia, Africa and North America, are classified as matrilineal. It is a common misunderstanding that matrilineal kinship systems are simply inversions of patrilineal ones, where women have taken the place of men. Some may even believe that matrilineality is the same as matriarchy. This is wrong. In matrilineal societies, just as in patrilineal ones, men usually hold formal political offices and control important economic resources. This implies that matrilineal systems tend: to be more complicated than patrilineal ones. In Figure 7.2 above, the difference between a simplified patrilineal and a simplified matrilineal system is illustrated. Biologically, the kin relations are identical on , Q2 Small Places, Large Issues both sides of the diagram; in practice, however, we see that person A belongs to a kin group composed in a very different way from person B's group. The Trobriand islanders, who were first studied by Malinowski during the First World War, are one of the most thoroughly studied matrilineal peoples in the world. The inhabitants of Kiriwina (the largest of the Trobriand Islands) are all members of matrilineages, which form four matriclans altogether. The most important task of the clan, which includes several lineages, is to arrange marriages. The lineage, which collectively owns land, magical incantations and other resources, is the most important corporate group in Trobriand society (Weiner 1988). Both female and male relatives of mother, mother's mother and mother's mother's mother (etc.) belong to one's matrilineage. One's father, however, belongs to a different lineage, namely his own matrilineage. So far, the matrilineal system appears as a mirror-image of the patrilineal one. However, the political power remains with men, although descent is traced through women. Each matrilineage has a male chief. Moreover, important resources are transmitted through inheritance from men to other men, usually from mother's brother to ego. The mother's brother also acts as an authority to his sister's children, while the father is expected to be kind and gentle (Malinowski 1984 [1922]) - quite the opposite of the general situation in patrilineal societies. Further, the Trobrianders are virilocal: the newly wed couple settle with the man's family, not with the woman's. The Trobrianders are horticulturalists and the staple food is the yam (a tuber rich in carbohydrates). However, each household does not primarily cultivate yams for its own consumption. Rather, they grow yams for their matrilineal relatives. A man and his household thus grow yams for the man's sister. (Women also grow some yams for daily consumption in gardens allocated to this purpose.) The islanders also give yams to other relatives, and a man may redistribute yams given to his wife by her brothers. The purpose of this traffic in yams may be seen to be purely symbolic - as a tangible reminder of kin obligations -but it has a political aspect as well. The woman's husband controls the yams and he is obliged to repay her brother for them - either by giving him presents or through political support. A man with several wives thus has good opportunities for acquiring political power, since he receives a lot of yams not required for food, which may thus be invested. At this point we leave the Trobrianders, whose exchange system will be dealt with in some detail in Chapter 12. Here we should note that matrilineal systems in general create a sharper conflict between household and lineage than patrilineal ones. The most important authority in the socialisation of children, the mother's brother, lives outside the household; and the head of household has profound obligations towards his sister, who also lives outside the household. The fact that the Trobrianders are virilocal (which is not common among matrilineal peoples) complicates matters further. Kinship as Descent 103 'Matrifocality' A well-known example of a 'very consistent matrilineal system' (Radcliffe-Brown's expression) is the Nayar people of the Malabar coast, southern India. Among the Nayars, stable nuclear families do not exist, and the man has no rights in his children - only in his sister's children (Gough 1959). According to custom, the marriage is broken off after only a few days, and the woman is later allowed to take lovers. Her children belong to her matrilineage, and the men concentrate their efforts on socialising their sisters' children. In addition to being matrilineal, this kind of arrangement can be described as matrifocal. This is no kinship term, but rather a description of a household type. It entails, simply and literally, that 'the mother is the focal point'. The term is used about households where the father for some reason is peripheral; where the marriage bond is unstable, as in many Caribbean societies or - as an extreme Case - among the Nayars. Matrifocality is rarely relevant for inheritance and succession, group formation and kin terminology. The phenomenon may occur in patrilineal as well as matrilineal and cognatic societies. It has been pointed out that in modern societies such as the USA, matrifocality is particularly widespread among the poorest. DIFFERENCES BETWEEN THE SYSTEMS Let us now sum up the central differences between matrilineal and patrilineal systems of descent. Both kinds of systems are usually dominated politically by men. Inheritance, particularly land rights, often follows men. In patrilineal societies, such rights are transmitted from father to child (frequently from father to son); in matrilineal societies, they are transmitted from mother's brother to daughter's son. Ascription of group membership varies along the same lines. In a matrilineal society, ego will be a member of the same kin group as his or her mother, mother's mother, mother's brother, mother's brother's children, etc.; in a patrilineal society, one belongs to the same group as one's father, father's father, father's brother, father's brother's children, etc. In patrilineal systems, the wives of the men ensure the continuity of the group; in matrilineal systems, the men's sisters do it. In certain societies, ego can take over certain rights through his or her father and others through the mother, but wherever kin-based corporations exist, one of the principles is nearly always followed in this respect. In certain rare cases, furthermore, one is a member of two lineages, one patrilineal and one matrilineal. This kind of system (dual descent) should not be confused with cognatic or bilateral kin reckoning. Among the Herero of Namibia, for example (Radcliffe-Brown 1952), everybody is a member of two separate 104 Small Places, Large Issues lineages - a matrilineal one through the mother and a patrilineal one through the father. Some rights are vested in the patrilineage and others in the matrilineage. Thus the men ensure the continuity of the patrilineages (their sons and daughters become members of their patrilineage), while the women, similarly, ensure the continuity of the matrilineage. CLANS AND LINEAGES So far, we have used the terms clan and lineage without defining them. In much of the professional literature, they are used nearly as synonyms, although clans tend to be regarded as larger, less tightly incorporated groups than lineages. As a general rule, we may say that a lineage consists of persons who can indicate, by stating all the intermediate links, common descent from a shared ancestor or ancestress. A clan encompasses people who assume shared descent from an ancestor/ancestress without being able to enumerate all of these links. Among the Pathans of Swat valley, northern Pakistan, thousands of persons regard themselves as members of the patrilineal Yusufzai clan, assuming by general consent that they are agnatic descendants of the mythical Yusuf without being able to 'prove' it (Barth 1959). Thus lineages are generally historically more shallow, and as a consequence smaller, groups than clans. In many societies, including the Trobriand Islands and Swat valley, several lineages considered to be related occasionally form alliances and so appear as clans - as kin groupings at a higher systemic level. Anthropologists have developed concepts about several kinds of clan organisation, but we should keep in mind that these notions are our own and not those of the informants. One widespread form is the conical clan, which is hierarchically ordered with a centralised leadership. There is one recognised leader, usually the oldest man of the lineage, at each level, and the clan as a whole has a chief at a higher level. Conical clans are typical of relatively stable chiefdoms, which nevertheless - unlike states - are based on kin loyalty, not on loyalty to the law or to the flag. Another model of clan organisation, which has been very influential in studies of African peoples, is the segmentary clan, which has largely been studied as a political form of organisation. The Nuer of southern Sudan, who were studied by Evans-Pritchard in the 1930s (Evans-Pritchard 1940; see Chapter 11), are probably the most famous example of a segmentary clan organisation. Unlike the conical clan system, the segmentary clan is non-hierarchical; it is acephalous (literally 'headless'), meaning that it has no recognised leader but is composed of structurally equal lineages and sub-clans. All members of the clan regard each other as relatives, but they have clear notions about relative genealogical distance: some are close relatives, while others are more Kinship as Descent distant. In peaceful periods, and when grazing land is abundant, the household may function as an autonomous unit, more or less like the Fulani household. If a feud with another group develops, or if there is a drought, the lineages and sub-clans may unite at a higher level, temporarily, as corporate groups. These two types of political clan organisation are the most elegant ones at the level of models, but they are not the only ones extant. In the New Guinea highlands it is quite common for the local group to be organised as an independent, patrilineal, exogamous clan with customary rights to a territory. Such a group, which may include a few hundred or a thousand persons, does not usually consider itself as related through kinship to another group. Further, such patricians are often scattered over a large area, provided potential rules of virilocality are not strictly enforced. Besides, it seems that the principle of shared descent is less important to many Melanesian and South-Asian peoples than it is in segmentary African societies (Carsten 1997). BIOLOGY AND KINSHIP At this point, a few words need to be said about a trend inside and (especially) outside anthropology, which places a great emphasis on kinship and, unlike most social and cultural anthropologists, sees it as being primarily biological. This perspective has already been mentioned in passing several times, and it is time to give it a slightly fuller treatment. Within this tradition, generally known as sociobiology (E.O. Wilson 1975), it is assumed that the single most important driving force in human action is the drive for reproduction. Men do their utmost to spread their genetic material, and women seek men who can protect them and their offspring while the children are small and defenceless. Culture and society, including kinship systems, develop more or less as side effects of these inborn needs. (One of the most extreme statements of this position is that of Richard Dawkins, who has claimed that we organisms are simply survival machines for our genes; see Dawkins 19 76.) Solidarity between family members can thus ostensibly be explained by the fact that they have shared genes. It would therefore be rational, from a genetic point of view, for a man to die for two of his brothers or for four of his first cousins. The most influential theorist of sociobiology, Edward O. Wilson (1975, 1978), originally regarded the social sciences as the most recent branches of biology, which have not yet been fully integrated into their 'mother science'. For obvious reasons, this kind of argument had to be met with strong reactions among cultural and social anthropologists (see Ingold 1986, pp. 68-73, for a scathing, biologically informed dismissal), who usually emphasise the non-biological aspects of human existence and who saw Wilson's 'new synthesis' as an unwelcome and irrelevant intrusion into their domain. In his The Use and Abuse of Biology, Sahlins (1977) argues against 106 Small Places, Large Issues sociobiology on the basis of a cultural relativist position. He shows, among other things, that the actual kinship systems studied by anthropologists in no way support the idea that solidarity between humans is strongest where there is shared genetic material. Indeed, many kinship systems create enduring and strong commitments between people who are not what we would call 'blood relatives'. Homosexuality, which exists in most societies, also seems difficult to explain within this framework. In his most recent work, Wilson (1998) seems to have modified his views, and calls for cooperation across disciplinary boundaries rather than subsuming one (anthropology) under the other (biology). Generally, adherents of Darwinist interpretations of human life have modified their formerly strongly determinist views since - roughly - the late 1980s. Culture is no longer regarded as an epiphenom-enon, and nobody would seriously claim today that cultural phenomena are necessarily 'biologically functional' or adaptive. Since the mid-1990s, furthermore, the term sociobiology itself is hardly used by anyone but Wilson himself, and terms like 'evolutionary psychology' or simply 'Darwinian social science' have replaced it. It nevertheless remains a fact that adherents of Darwinist views of humanity tend to emphasise uniformities presumably founded in genetic dispositions (sex, violence and kinship are typical topics), which challenge the sociological and cultural relativist underpinnings of mainstream anthropology, often in provocative ways. In spite of recent rapprochements between sociocultural and biological anthropologists (Dunbar et al. 1999), there remain deep tensions between their respective accounts of kinship. One of the most controversial studies in this regard is arguably Martin Daly and Margo Wilson's work on family violence, published in Homicide (Daly and Wilson 1988) and in a string of subsequent publications. A main argument in their work, supported by statistical material from several Western societies, is that step-parents are more likely to harm children than biological parents, because they do not share a lineage of genes with them. Sociocultural anthropologists have tried to interpret their findings in other ways, arguing that the social relations in a family with a step-parent are likely to be systematically different from families where both are biological parents. The most common view among social and cultural anthropologists is expressed clearly by Holy (1996), in an introductory text on kinship. He notes that many peoples in the New Guinea highlands become relatives by virtue of sharing food, and that their relationship to their mother seems to be based on her role as someone who offers food rather than as the person who brought one into the world. After some further examples of variations in cultural ideas about kinship, he concludes: All societies have their own theories about how women become pregnant. As these theories may ascribe widely different roles to men and women in procreation, the notions concerning the relationship between the child and its father and the child and its mother may differ considerably from society to society. (Holy 1996, p. 16) Kinship as Descent i q 7 This chapter has chiefly described kinship through descent, inheritance and succession, and has indicated how corporations can be formed on the basis of notions of shared descent. In the next chapter, we move a step further and consider the importance of marriage for descent and kinship in general. SUGGESTIONS FOR FÜRTHER READING Alan Barnard and Anthony Good: Research Practices in the Study of Kinship. London: Academic Press 1984. Ladislav Holy: Anthropological Perspectives on Kinship. London: Pluto Press 1996. David M. Schneider: A Critique of the Study of Kinship. Ann Arbor: University of Michigan Press 1984.