Kinship is one of the universals in human society and therefore plays an important role in both the regulation of behavior and the formation of social groups. Kinship systems depend on the social recognition and cultural implementation of relationships derived from descent and marriage and normally involve a set of kinship terms and an associatedn set of behavioral patterns and attitudes which, together, make up a systematic whole. All societies distinguish various categories of relation-ship by descent or consanguinity, and most societies distinguish relationships by marriage or affinity as well. Although dictionary definitions differentiate these relationships, it is convenient to extend the term “kinship” to cover both kinds. The resulting network of social relations may constitute almost the whole social structure in some of the simpler societies or be a relatively small part of a highly complex structure, as in modern industrial societies. In either case, however, the system of kin-ship and marriage plays an important role in maintaining group cohesion and solidarity and in orienting the individual members to the social maze. The use of the term “system” implies that there is a complex relation of interdependence between the component parts: the social categories and the associated rights and duties.

Kinship systems are found to vary in different societies with respect to a number of characteristics: (1) the extent to which genealogical and affinal relationships are recognized for social purposes; (2) the ways in which relatives so recognized are classified or grouped in social categories; (3) the particular customs by which the behavior of these relatives is regulated in daily life; (4) the various rights and obligations which are mediated through kinship; and (5) the linguistic forms which are used to denote the various categories of kin. Often the domain of kinship is clearly marked off, but there are frequently metaphorical and other extensions which result in related systems or subsystems. [SeeKinship, articles onDescent Groups and Pseudo-kinship.]

The near universality of the nuclear or elementary family and its role in mating and reproduction have made it a focus for studies of kinship. Here are found the “primary” relationships of parent and child, husband and wife, and brothers and sisters, and it is possible to construct a network of genealogical relationships encompassing the whole society by extension from this nucleus. But the processes of mating and reproduction are regulated in all human societies by incest rules and social convention, and although the resulting domestic family group is often based on physiological parenthood, it is the social recognition of parenthood that provides a child with a legitimate position in society. Thus, it is often convenient to distinguish the pater,or social father, from the genitor, or physical father, and sometimes it may even be necessary to distinguish the culturally assumed genitor from the actual biological father. Moreover, in some African societies women may play the role of “social fathers,” marrying and “begetting” children with the aid of a biological father.

The family of orientation into which a child is born is often part of a larger extended family which includes many additional relatives. When an individual marries, he and his spouse may establish a new family of procreation or may join a larger family structure. Normally he acquires a new set of relatives by marriage, but in those cases where marriage is specified in terms of a particular category of relatives, his affinal relations may also be his consanguineal ones. In a few instances, as among the eighteenth-century Nayar of southern India, the family of husband, wife, and children did not exist as a social unit, and kinship was correspondingly modified. These examples indicate that the kinship system may or may not coincide with the genealogical network; in every case, the degree of relationship is a matter for empirical investigation.

The kinship categories found in various societies often cut across the distinctions that seem logical in Euro-American societies. In the latter, lineal relatives are set off from the collateral uncles and aunts, and the relatives through the father and the mother are treated in parallel fashion. But in many societies throughout the world the terms for father, mother, brother, sister, and so on may be widely extended instead of being restricted to the immediate family group. In some cases the extension is by generation, the term for father being extended to his brothers and male cousins as far as genea-logical relatives are remembered, and analogously for other relatives. In other cases the extensions may be vertical, in terms of unilineal descent groups, so that all the members of a particular lineage or clan may be classed as “fathers” and “father’s sisters,” or “mothers” and “mother’s brothers,” regardless of generation or even of genealogical connection. The resulting kinship systems often have a wide range, sometimes encompassing the entire social group, in contrast to the narrow range of many Western systems. The particular patterns of grouping kinsmen show considerable variety, and each must be understood in its own terms before it can be compared with systems based on other principles of grouping.

The patterns of behavior that prevail between relatives define their relationships and as such are an integral part of’ the kinship system. In almost all societies the family is responsible for the care and support of children during their period of dependency and for their education and training for adult life. These tasks involve both love and affection and authority and discipline. The potential conflicts and ambivalences are often resolved by the allocation of authority to one parent or the other or to some relative outside the immediate family. The relationships established in the family group are affected by generation and relative age and by similarities or differences of sex. Those members of the parental generation who are in a position of authority are entitled to obedience and respect; others may share an intimacy without subordination. Friendship and support are expected of brothers and sisters, although often there are restrictions on behavior between a brother and a sister after puberty. With relatives outside the family group there is frequently a greater variety of behavior patterns, some seemingly based on the model of relationships within the family but others representing obligatory joking or teasing, on the one hand, or extreme respect or avoidance, on the other.

During the long period of socialization within the family or domestic group, the child gradually learns the proper attitudes and behavior patterns toward his various relatives. These patterns are present in the society in terms of cultural ideals and as behavioral norms, and their observance is reinforced in a variety of ways. There is considerable evidence that in most societies children learn the essentials of kinship rather early. At marriage an individual normally acquires a whole new set of affinal relatives to whom he must make varying adjustments, depending on the patterns of residence and interaction. Marriage is frequently an alliance between two groups of kin and may be mediated by exchanges of property as well as of spouses. The individual’s relation to his spouse’s relatives is often an intensification of the attitudes of respect or familiarity he has toward his own parents and siblings. Thus he may avoid his mother-in-law for a period and may be required to joke roughly with his brothers-in-law and sisters-in-law. Some societies prescribe marriage with a particular category of kin, usually “cross-cousins” (children of a brother and a sister), so that one’s new affines are also consanguineal relatives and the new behavioral adjustments are more easily handled. Such societies intensify the bonds between existing relatives at the expense of securing a new set of relatives by means of marriage. In these societies the opposition between “consanguinity” and “affinity” is often present, despite the formal absence of distinct affinal terms.

In most societies the rights and obligations of members are channeled, in part at least, through the kinship system. Thus, the right to membership in a descent group may depend on the proper marriage of the parents, in which the procreative rights in the wife have been formally transferred to the husband and his lineage. Similarly, rights to the utilization of land or other kinds of property may sometimes be secured only through member-ship in “corporate” descent groups which are both integral parts of the kinship system and units in the larger social structure. Succession to various offices or status positions usually depends on kin-ship, even though the offices are controlled by descent groups or associations. Even the rights to residence in one locality or another may be specified in kinship terms. Rights normally imply obligations or duties and are concerned with the larger society and its continuation, even though phrased in kinship terms. Many center on marriage and the resulting family and involve domestic service, labor, sexuality, procreation, and support, among other things. Where rights and duties are codified in legal or jural terms they are more easily seen, but they are an integral part of kinship behavior.

In all societies, kinship is marked by a set of relationship terms that define the universe of kin and that may be extended metaphorically to nonkin and even to various aspects of the world of nature. Kinship terms have been the center of much interest on the part of both anthropologists and linguists, and considerable progress has been made in their classification and analysis. In most societies, kinship terms are utilized in daily life, both in reference and in direct address, and often their use is required by custom. The terminological system frequently represents a distinctive subset of the lexicon, and the linguist can provide greater understanding of it by componential analysis, formal analysis, and historical reconstruction of earlier forms. There is a basic logic to kinship terminology, in that particular terms do not imply a status position so much as a relationship: the use of a particular term implies its reciprocal. Thus, if you call a man “father,” he responds with “son.” On the other hand, parallel terms in different societies may or may not have the same significance or meaning. Social anthropologists have been more concerned with the set of behavioral patterns between relatives and have tended to consider the terms used as linguistic tags representing or symbolizing the particular expected behaviors and attitudes between pairs or groups of kin. But the two systems are not always in a one-to-one relationship, and it is more profitable scientifically to consider them in a relationship of dynamic interdependence and to examine the discrepancies as possible evidence for social and cultural change.

The kinship system, in turn, has various and complex relations with the other social institutions that together make up the total social system, or social structure. Because kinship enters into economic, political, legal, and ritual relationships in various societies, there is sometimes a tendency to ignore or underestimate its significance. The function of kinship terminology in interaction is a symbolic one. Whenit is used it defines for the participants the general mode of behavior to be followed in particular social situations. The universality andenduring character of kinship suggest its importance in binding men and women together in society and providing a foundation for the building of more specific social structures.

Historical development

The scientific study of kinship systems is only a century old, but in that brief period it has engendered more controversy and a greater variety of theoretical formulations than have most aspects of human society. The early studies concentrated on the terminological systems, for the most part, and utilized them as evidence for historical relationships or as survivals of assumed earlier stages of society based on promiscuity and group marriage. The reactions against such “conjectural” history led to a denial of the sociological significance of kinship terms and to an attempted explanation in terms ofpsychological principles. This, in turn, resulted in a renewed attempt to understand kin-ship in terms of the behavioral system and with reference to the ongoing society. More recently, some progress has been made in studying changes in kinship systems over time. One recurring difficulty has been the limited number of societies for which there is adequate information on kinship systems, but this situation is improving rather rapidly. A further difficulty has been that the preliminary classifications have been based on limited criteria, and there has been a tendency to study kinship piecemeal and to search for simplified formulations in terms of causal relationships rather than to treat the complex whole.

The foundations for the study of kinship were laid by L. H. Morgan in his Systems of Consanguinity and Affinity of the Human Family(1871). [See the biography ofMORGAN, LEWIS HENRY.] In this work, the result of more than a decade of concentration on kinship, Morganassembled data on the terminological systems he was able to collect or secure for nearly every major area of the world. He grouped the terminologies into two great classes, the “descriptive” systems, which he ascribed to the Aryan, Semitic, and Uralic linguistic families, and the “classificatory” systems, which he thought were characteristic of the American Indians, the Polynesians, and many of the peoples of Asia. The “classificatory” systems merged lineal with collateral relatives in varying degrees, in contrast to the Euro-American systems, which isolated lineal relatives in theterminology. As W. H. R. Rivers noted later (1914, p. 4), no discoveryin the whole range of science can more certainly be credited to one man than the discovery of the classificatory system of relationship to Morgan.

Morgan’s early interest in kinship systems was a historical one. Discovering in 1858 that the Ojibwa Indians had a pattern of grouping relatives that was almost identical with that of the Iroquois, who spoke a quite different language, he came to the view that kinship patterns were highly stable and set out to collect kinship terminologies in order to demonstrate that the American Indians were of common descent and had originally come from Asia. When he found an almost identical system among the Tamils of India, he felt he had proved his historical hypothesis. But in the meantime, the discovery of the Hawaiian (Malayan) pattern of terminology, which was classificatory to an even greater degree, led him to explain it as a result of assumed earlier forms of marriage for which there was no existing evidence. The resultin evolutionary development of social institutions and cultural stages presented in Ancient Society (1877) aroused extended controversieswhich long obscured Morgan’s important contributions to the study of kinship.

In America, the criticisms by Kroeber in his “Classificatory Systems of Relationship” (1909) were the most influential and far-reaching [seeKroeber]. He found the distinction between “descriptive” and “classificatory” misleading and suggested that kinship terminology be analyzed, instead, in terms of some eight “psychological principles” based on the difference of generations, the distinction of lineal and collateral relationships, the difference of age within a generation, the sex of therelative, the sex of the speaker, the sex of the connective relative, the distinction of blood and affinal relationship, and the condition of life of the connecting relative. Kroeber came to the conclusion that terms of relationship reflect psychology rather than sociology and are determined primarily by language—hence they could be utilized for sociological inferences only with great caution.

In England J. F. McLennan had dismissed kin-ship terms as a mere set of mutual salutations, in the process of defending his own evolutionary formulations for society. But Rivers, who had become interested inkinship during the Torres Strait expedition of 1898–1900, where he had developed the “genealogical method” for collecting accurate data on various aspects of social organization, re-turned to Morgan’s basic ideas as a result of his studies of Melanesian society. In Kinship and Social Organisation (1914) he proposed that kinship terminology is rigorously determined by social conditions and particularly by forms of marriage and hence can be utilized to reconstruct the recent history of social institutions. These hypotheses and the accompanying illustrations have become one starting point for the modern study of kinship systems. [SeeMclennanandRivers.]

Lowie has been the most influential American ethnologist concerned with the study of kinship. Accepting Rivers’ position that kinship terminology is related to social usages, but influenced by Kroeberas well, he sought to test the hypotheses that had been proposed against the available ethno-graphic information. His own comparative studies of the Plateau Shoshoneans and the Hopi Indians led him to the conclusion that the kinship system of the latter is functionally connected with their clan system. [SeeLowie.] He summed up his general position in a statement tha is still valid:

Relationship terms are studied by the anthropologist not merely as so many words inviting philological analysis and comparison, but as correlates of social custom. Broadly speaking, the use of a specific kinship designation, e.g., for the maternal as distinguished from the paternal uncle, indicates that the former receives differential treatment at the hands of his nephews and nieces. Further, if a term of this sort embraces a number of individuals, the probability is that the speaker is linked to all of them by the same set of mutual duties and claims, though their intensity may vary with the closeness of the relationship. ([1929] 1959, vol. 19, p. 84)

In England Malinowski and Radcliffe-Brown have been most influential figures in the development of kinship studies. Malinowski, as a result of his study The Family Among the Australian Aborigines (1913) and his extended field research in the Trobriand Islands, emphasized the importance of the family as the “initial situation” for the development of kinship, from which attitudes and terminology could be widely extended. He also called attention to the significance of “sociological fatherhood” in a matrilineal society that did not recognize the genetic role; but he was more concerned with the function of kinship and other social institutions in fulfilling individual needs. [SeeMalinowski.] Radcliffe-Brown, an early student of Rivers, is the central figure in the modern study of kinshipsystems. He was the first to develop the conception of the kinship system as composed of both terminology and patterns of social behavior and to see kinship as an integral part of the larger social structure. As a functionalist he was concerned with the significance of institutions in maintaining the social system, but he went further and attempted to discover basic structural principles that were relevant to a variety of different terminological groupings and social usages. [SeeRadcliffe-Brown.]

By the end of the 1920s the preliminary classification of kinship terminologies was well underway. Morgan’s twofold classification was remodeled by Rivers, and Gifford (1922) had utilized Kroeber’s categories for the classification of Californian Indian terminologies. Spier’s classification (1925) of North American Indian terminologies into eight empirical types, based on the patterns of grouping for cross-cousins, was particularly influential. Lowie (1929) proposed a world-wide classification into four major types, based on the treatment of relatives in the parental generation. These were soon followed by Radcliffe-Brown’s classification(1931) of Australian social systems into two main types, in each of which kinship, preferential marriage, and clan groupings were systematically interrelated.

During the following decade a number of field studies were carried out by students of Malinowski and Radcliffe-Brown in which kinship received more adequate treatment. Firth’s studies of the Tikopia (1936), Warner’s on the Murngin (1930), Evans-Pritchard’son the Nuer (1951), Fortes’ on the Tallensi (1949), Tax’s on the Fox (1937), Hallowell’s on the Ojibwa (1937), Eggan’s on the Plains and Pueblo Indian groups ([1937b] 1962, pp. 35–95; 1950), and Spoehr’s studies of the southeastern Indian tribes (1941; 1942; 1944) are among those researches that have contributed to the development and modification of the structural-functional approach.

The clearest statement of this approach is found in Radcliffe-Brown’s Introduction toAfrican Systems of Kinship and Marriage (1950), in which he was concerned with the general comparative and theoretical study of kinship organization as an arrangement which enables persons to cooperate with one another in an orderly social life. In this discussion he compared and contrasted the cognatic system of the early Teutonic peoples with the agnatic lineage systems of ancient Rome and many modern African tribes and indicated the relevance of the principles of “the unity of the sibling group” and “the unity and solidarity of the lineage” for various aspects of social life. Here he was particularly concerned with the significance of unilineal descent in bringing about corporate kin groups that continue beyond the life of individual members and may control resources, exact vengeance, regu-late marriage, and engage in ritual. He saw marriage as essentially a rearrangement of social structure and discussed in detail the significance of marriage in various African societies. For a world-wide classification of kinship systems heproposed four types: father-right, mother-right, cognatic systems, and double lineage systems, each of which has a number of varieties.

Current development of kinship studies

In the modern period there have been a number of new directions in the study of kinship which involve both method and theory. Some of these include cross-cultural comparisons involving statistical and correlational techniques; others involve linguistic analyses building on Kroeber’s earlier categories or utilizing formal analyses; still others utilize models of various types, some derived from linguistics and others from mathematics. These studies have stimulated a great amount of new research and promise to broaden our knowledge of kinship phenomena in various directions.

Of particular significance is Murdock’s cross-cultural study of family and kinship organization in about 250 societies throughout the world, presented inSocial Structure (1949). Utilizing the postulational method and statistical analysis he found that kinship terminologies are primarily determined by such sociological factor as descent and residence, with marriage rules of lesser importance. He then established six types of kinship terminology, based in part on Spier’s earlier classification, and combined these with rules of descent and residence to give 11 major types of social organization. A proposed order of social change, beginning with changes in the residence pattern, was then tested against the evidence from linguistic reconstructions and other data and was found highly reliable. “It seems clear,” Murdock wrote, “that the elements of social organization, in their permutations and combinations, conform to natural laws of their own with an exactitude scarcely less striking than that which characterizes the permutations and combinations of atoms in chemistry or genes in biology” (1949, p. 183).

Murdock’s study represents a notable advance in the application of social science methodologies to the study of social organization, but there has also been considerable criticism of the sampling involved, the statistical techniques used, and the data selected for analysis. He responded with the more adequate “World Ethnographic Sample” (1957) and with a revised classification (1959, pp. 135–140) of five major types of social organization, based primarily on descent and residence patterns. He also modified his assumptions about the primary role of residence in bringing about social change.

The contributions of Lévi-Strauss to the study of kinship systems are of a different character, and in Les structures elementaires de la parente (1949) andStructural Anthropology (1958) he presents some highly original views on the nature of social structure in general and kinship in particular. “Social structure,” for Levi-Strauss, is in itself concerned not with the empirical reality of social relations but with models which give rise to them, and he discusses the relevance of mechanical models (those on the same scale as the phenomena) and statistical models (where the elements ofthe model are on a different scale) for various problems, particularly those of communication. With regard to kinship he views the terminology and the system of attitudes as representing quite different orders of reality: “The modalities of behavior between relatives express to some extent the terminological classification, and they provide at the same time a means of overcoming difficulties and contradictions resulting from this classification” ([1958] 1963, p. 310), a dialectic which is responsible for change in both systems. Levi-Strauss also proposes a some-different unit for kinship studies fromthe elementary family, which is favored by Malinowski, Radcliffe-Brown, Murdock, and others. He believes “the relationship between ‘brothers-in-law’ is the axis around which kinship structure is built” ([1958] 1963, p. 46) and thus adds the wife’s brother to the family unit. All kinship structures are constructed on this “kinship atom,” primarily by the organization of a series of oppositions between attitudes of familiarity and reserve. The resulting kinship system “does not consist in the objective ties of descent or consanguinity between individuals. It exists only in human consciousness; it is an arbitrary system of representations, not the spontaneous development of a real situation”(ibid., p. 50).

The principle of reciprocity, as manifested in various forms of exchange in social life, is central to Levi-Strauss’s view of social institutions. Kinship in human society is established and perpetuated through specific forms of marriage, and marriage as a form of exchange involves the circulation of women. He is, therefore, particularly concerned with what he calls “elementary structures,” or those characterized by preferential marriage with a particular category of kin, usually a “cross-cousin.” In this respect, Levi-Strauss has attempted the analysis of matrilateral cross-cousin marriage systems, which are found in Australia and in southeastern and eastern Asia as well as in a few other regions, and he sees the resulting dual structure of “wife-giving” and “wife-receiving” groups reflected in many other aspects of society and culture.

This complex and original contribution (summarized in English in de Josselin de Jong 1952) has stimulated a number of important studies and engendered considerable controversy. Homans and Schneider, in Marriage, Authority, and Final Causes (1955), essay an alternate explanation based on Radcliffe-Brown’s theory of sentiments. Needham attacked this strongly in Structure and Sentiment ( 1962) and went on to make a number of reformulations of what he calls “prescriptive” marriage systems. Leach, in Rethinking Anthropology (1961), shows the considerable influence of Levi-Strauss, as does Dumont, whose Hierarchy and Marriage Alliance in SouthIndian Kinship (1957) emphasizes the importance of treating certain categories of relatives as affinal rather than consanguineal. [SeeMarriage, articles onComparative AnalysiandMarriage Alliance.]

In recent years a number of anthropologists and linguists have returned to Kroeber’s analysis (1909) of kinship terminologies and have developed a more sophisticated approach, called “componential analysis.” The general framework for componential analysis derives from linguistic theory, and the kin-ship vocabulary is regarded as constituting a paradigm which can be analyzed in the same manner as other paradigmatic sets in a language. Currently Lounsbury and others are attempting to construct theories using a limited numbe of ordered roles similar to those of “generative grammar.” The resulting “formal account” specifies (1) a set of primitive elements and (2) a set of rules for operating on these to generate a model which represents the empirical data (see Lounsbury1964).

Goodenough (1956) treats Kroeber’s categories as essentially social components, but Lounsbury’s (1956) and Buchler’s (1964) analyses are based upon strict genealogical reckoning and operate in terms of the primary relations in the nuclear family and their extensions to more distant relatives. Lounsbury assumes that “the primary function of kinship terminologies is to delineate therelation of ego to the members of his personal kindred in such a way as to express some socially and legally important aspect of each of these relationships” (1964, p. 382). Friedrich, in “Semantic Structure and Social Structure: An Instance from Russian,” is concerned with seeing their interrelationships: “The semantic network symbolizes and is generated by the social network. Covariation between both net-works is significant because it can lead to yet more general inferences about native concepts” (1964, p. 132). And H. C. Conklin, in “Ethnogenealogical Method,” illustrates the steps which may be taken from ethnographic descriptionto final analysis:

The sequence I have followed has led us from specific-to-general-to-abstract-to-correlational substatements of Hanunoo ethnography. We have moved from individuals occupying established genealogical positions in a well-recognized kin net, to the examination of types of kin classes, to the analysis and articulation of the defining features, or significata, which underlie the whole category system; and finally to a brief consideration of one set of significant nonterminological correlates of the more highly structured parts of this system. (1964, p. 50)

In this procedure he finds that the natives own “model” of their system is an important part of the data.

The utilization of mathematical models for the elucidation of kinship structures has had a long history, beginning with Galton (1889) and continuing with Weil’s appendix (1949) to Lévi-Strauss’s monograph and, most recently, H. C. White’s application of matrix algebra (1963). White is particularly concerned with prescriptive marriage systems, such as the Kariera, Arunta, Murngin, and Purum, but it remains to be demonstrated whether the logical manipulation of kinship categories and marriage rules adds greatly to ourunderstanding of kinship systems. [SeeComponential Analysis.]

Comparative studies

The comparative study of kinship systems as wholes, and in relationship to ecological and historical factors as well as to other aspects of social structure, has had a more limited development. The initial model for such studies was Radcliffe-Brown’s Social Organization of Australian Tribes (1931), which has been carried further by African Systems of Kinship and Marriage (Radcliffe-Brown & Forde 1950). For North America, Eggan (1955; 1966) has been concerned with the classification and interpretation of kinship systems in a number of regions, utilizing the method of controlled comparison and attempting to study changes over time, and P. Kirchhoff has provided a preliminary survey of the kinship systems in South America (1931; 1932).

On the basis of a detailed analysis of the Cheyenne and Arapaho kinship systems, Eggan ([1937k] 1962) proposed a preliminary classification of the kinship systems of the Plains region of North America into two major types: (1) a “generational” type and (2) a “lineage” type. The tribes of the High Plains, who wereorganized in terms of bi-lateral bands composing a camp circle and lived as seminomadic hunters, were—with one exception —also organized in terms of a wide-ranging “classificatory” kinship system, in which generation and sex were emphasized. The tribes of the Prairie Plains to the east, on the other hand, were organized in terms of unilineal descent and lived in permanent villages, from which they went on periodic buffalo hunts; but they depended on horti-culture for their basic subsistence. Their kinship systems were also “classificatory,” in that lineal and collateral relatives were merged in the terminology, but they utilized the lineageprinciple to provide a wide extension to the system. There were two sub-types: (a) the “Omaha” system, associated with patrilineal descent, and (b) the “Crow” system, associated with matrilineal descent.

A comparison of these two major types indicated that each represented an adjustment to the eco-logical and social conditions of their respective regions and that the “generational” systems of the High Plains were based on the relationship of brothers, which was functionally of great importance in Plains life; whereas the Prairie Plains tribes utilized the lineage principle to provide greater stability and continuity over time. By examining the historical backgrounds of the High Plains tribes it became evident that tribes coming into the Plains with different social systems ended up with similar systems. The Crow Indians are a test case. They split off from the Hidatsa several hundred years ago and gave up their village-dwelling, agricultural life in North Dakota for the seminomadic life of the High Plains. Their social system and kinship organization are intermediate, partaking of both types. It seems probable that the conditions of life in the High Plains favored a more amorphous and mobile type of social organization, which could vary to meet changing ecological and social conditions. As Prairie Plains peoples moved out onto the High Plains to take advantage of the greater efficiency of the horse in hunting buffalo, they modified their kinship systems in the direction of a “generational” type.

North and east of the Great Lakes, the Algonkian-speaking peoples have been shown by Hallowell (1937) to have had kinship systems basedon cross-cousin marriage, and he has proposed that the con-temporary variants are intelligible as a result of modifications resulting from acculturative processes and local conditions. Eggan (1955) has extended this hypothesis to northern Algonkian groups moving into the Plains region and to the Dakota groups. The central Algonkian tribes have been shown by Callender (1962) to have shifted from an earlier kinship system based on cross-cousin marriage to a lineage-based system of the Omaha type, as these tribes moved southward into the Prairie Plains and expanded in population with the adoption of horti-culture.

The Omaha and Crow subtypes of kinship systems are not limited to North America but are seen there in their most typical form. The Omaha systems are generally associated with patrilineal line-ages or clans, and often with a dual division of the society, while the Crow systems are associated with matrilineal lineages or clans. In either casethe essential feature is that the lineage or clan is treated as a unit for kinship purposes, an individual considering all his kinsmen through the mother or father as of the same kind. This utilization of de-scent groups for kinship extensions results in both a wide range and a continuity to the social system.

In the southeastern region of North America, where all the major tribes were organized in terms of a Crow-type kinship system, preliminary studies of the Choctaw terminology collected for Morgan suggestedto Eggan (1937a) that changes due to acculturation has been underwayin all of the south-eastern tribes up to 1860 and that the degree of change in kinship terminology was related to the degree and type of acculturative pressures. A field study of the modern descendants by Spoehr (1941; 1942; 1944; 1947) not only confirmed these hypotheses but provided demonstrations of the processes by which the Crow type systems shifted to a generational pattern over the period of a century of acculturation.

In the southwestern region the Western Pueblos have all been foundto have a simple specialized type of social structure based on matrilineal line-ages and clans and a Crow type of kinship system. The Eastern Pueblos, however, though participating fully in the general Pueblo culture patterns, have a quite different social structure, based ona dual organization and a bilateral “nonclassificatory” kin-ship system which emphasizes generation and relative age (Eggan 1950). A number of hypotheses have been advanced to account for this major difference in a single culture type. Here, Dozier’s (1954) study of the Hopi-Tewa, a group of Eastern Tewa who migrated to the Hopi region around A.D. 1700 and who have rearranged their kinship system to conform to the Hopi model, analyzes an important instance of acculturation between Indian groups. The southern Athabaskan groups, made up of the Navajo and the various Apache tribes, show a furtherseries of changes in kinship, not only from their northern relatives in Canada but from one another as well.

These brief summaries can only suggest the kinds of comparative regional studies of kinship systems which have been developed on the basis of structural-functional assumptions, with the added controls of ethnohistory, linguistic reconstruction, and ecological factors. Along with them have been such studies as Bruner’s (1955–1956) on the actual processes of change in Mandan-Hidatsa kinship terminology under contemporary reservation conditions, where certain of the factors affecting choice of Indian or Euro-American kin-ship models have been clarified.

A different type of comparative study is exemplified by Schneider and Cough’s Matrilineal Kinship (1961), which grew out of a cooperative Social Science Research Council summer seminar organized by Schneider, with the additional participation of Colson, Aberle, Fathauer, Basehart, and Sahlins. The distinctive features of matrilineal descent groups are first stated in theoretical terms and in contrast to their patrilineal counterparts. They are then examined against nine matrilineal systems which are presented in detail, and more broadly, against a larger number of cases available from Murdock’s “World Ethnographic Sample” (1957). Here, with descent held as a constant, the variations in group structure, residence, kinship, and marriage are examined in terms of structural theory, cultural ecology, and evolutionary development. Of particular significance for kinship theory is the discussion of the strength of the brother-sister relationship in matrilineal societies as against the husband-wife bond, and its manifold effects on the kinship structure, especially with regard to the tensions between a man and the matrilineal descent group over control of his wife and children, which had earlier been analyzed by Richards (1950). The considerable variety of types of kinship structures associated with matrilineal descent (and the cor-responding variety associated with patrilineal and cognatic descent) indicate problems for future research.

Future developments

Kinship theory is set in the broader framework of social and cultural anthropology, as is indicated in the article on culture, where Singer discusses the structural versus the cultural analysis of kinshipsystems in terms of the controversies between Kroeber and Radcliffe-Brown. [SeeCULTURE.] In the perspective of modern kinship studies the position of Radcliffe-Brown has been the more productive, and Kroeber has partly modified his original position : “As part of language, kin term systems reflect unconscious logic and conceptualpatterning as well as social institutions” ([1901–1951] 1952, p. 172). The current interest of linguists in the componential analysis of kin term systems has clarified certain aspects of terminology but at the expense of rejecting the advances made by treating kinship as a social system. One promising move in this direction would be to include social components, such as locality and lineage grouping, along with those of generation, relative age, sex, and so on (see Leach 1961; Friedrich 1964). Levi-Strauss has been impressed with th relevance of structural linguistics to the study of kinship:

Like phonemes, kinship terms are elements of meaning: like phonemes they acquire meaning only if they are integrated systems. “Kinship systems,” like “phonemic systems,” are built by the mind on the level of unconscious thought. Finally, the recurrence of kin-ship patterns, marriage rules, similar prescribed attitudes between certain types of relatives, and so forth, in scattered regions of the globe and fundamentally different societies, leads us to believe that, in the case of kinship as well as linguistics, the observable phenomena result from the action of laws which are general but implicit. ([1958] 1963, p. 34)

But granted that the principles of duality and of reciprocity may be basic, their relevance to many social systems has not yet been demonstrated. For Radcliffe-Brown the recurrence of particular features in the diversity of kinship systems throughout the world was evidenceof a limited number of general structural principles, such as the equivalence of siblings and lineage solidarity, which were combined in varying ways.

But if Kroeber’s reformulation is to be more than a compromise, it will be important to relate systematically the linguistic, cognitive, and sociological aspects of kinship in order to develop a more comprehensive theory. Thus, the formal rules by which Lounsbury generates Crow- and Omaha-type terminological systems are related to Radcliffe-Brown’s sociological principles, as Lounsbury notes (1964, p. 357). And the principle of duality may find more general expression in the relations of ego to alter and in the attitudes of respect and familiarity than in the more specific matrimonial arrangements and dual organizations of society.

The emphasis on descent systems with reference to kinship has been balanced in recent years by a greater concern with bilateral or cognatic systems, but the precise relationships between these two types are not yet clear. Both lineage-based and bi-lateral kinship systems are faced with similar problems but solve them in somewhat different ways. The historical changes noted above suggest that greater efficiency in adaptation to particular ecological situations may be an important factor. Whether there are broad evolutionary changes is not yet clear. The early formulations of Morgan have been discredited but no large-scale evolutionary sequence with regard to kinship systems has been developed to take their place. L. A. White (1939, pp. 569–570) has proposed a more limited development to account for the Iroquois-Dakota kinship terminology in relation to the Omaha and Crow types: “When the clan system is young and weak the kinship system will be of the Dakota-Iroquois type, regardless of the sex in which descent is reckoned. As the clan system develops, however, and comes to exert its influence more and more upon the social life of the tribe, the Dakota-Iroquois terminology will be transformed into the Crow type in a matrilineal society and into the Omaha type in a patrilineal society.” Murdock’s study (1949) lends considerable statistical support to this view. It is clear, however, that the Dakota type of kinship system can develop in association with cross-cousin marriage without the presence of any clan organization (Eggan 1955). Where we find the classic Omaha and Crow kinship systems they are generally associated with well-developed “corporate” patrilineal and matrilineal groups, respectively, but not all societies with well-developed corporate lineage groups have Omaha or Crow kin-ship systems. For North America there is some evidence for a cyclical oscillation between kinship systems based on a generational principle of organization and those based on a lineage principle. These are thetwo major axes for the classification of kin, and Murdock has provided a theoretical formulation of change in social structures to be tested against the empirical evidence. Here, studies of the type made by Krader (1963) with regard to the Turkic and Mongol kinship systems and by Friedrich (1963) with regard to the historical development of the Russian kinship system will be particularly important.

Studies of Euro-American kinship systems have so far been concerned primarily with terminological patterns and their historical development. However, Schneider and Firth are engaged in a large-scale comparative study of kinship in Chicago and London, respectively, which should both yield new and important results and bring our knowledge of kin-ship systems in contemporary industrial society up to the level of those of nonliterate groups.

Our knowledge of affinal kinship—the relationships established through the marriage tie—has been seriously neglected in most studies of kinship systems. Some societies, such as the Ifugao inthe northern Philippines, reduce the significance of the affinal tiealmost to the vanishing point, in contrast to consanguineal relationships; others build much of their social structure on the relationshipsbetween spouses. The contributions of Levi-Strauss, Dumont, and Needham to our understanding of affinity in cases of preferential or prescriptive cross-cousin marriage, where affinal terminology is disguisedin the consanguineal system, have been mentioned. But there is as ye no comprehensive classification of affinal kinship terminology, nor any general theory of the nature of affinity, although Aginsky (1935) long ago called attention to the importance of the problem. Friedrich(see Goodenough 1964, pp. 131–166) has provided a beginning with his detailed analysis of the Russian affinal system.

We can look forward to a continued flow of empirical data on kinship and to a continuing dialogue between studies utilizing analytic variables and statistical methods, on the one hand, and intensive studies of a more limited range but concerned with kinship systems as wholes and in their ecological and historical contexts, on the other. Outof this dialectic should come more adequate concepts and classifications, as well as a greater under-standing of the phenomena of kinship and the processes relevant to its development. At a more general level, such studies also furnish a body of data to clarify the relations between culture, as a set of ideas and symbols, and social structure,as a system of social interaction. Kinship organizes social relationsin terms of cultural patterns.

Fred Eggan


Aginsky, B. 1935 The Mechanics of Kinship. American Anthropologist New Series 37:450–457.

Beattie, J. H. M. 1964 Kinship and Social Anthropology. Man 64:101–104.

British Association for theadvancement ofscience(1874) 1954 Notes and Queries on Anthropology. 6th ed., rev. London: Routledge.

Bruner, Edward M. 1955–1956 Two Processes of Change in Mandan-Hidatsa Kinship Terminology. American Anthropologist New Series 57:840–850; 58: 554–556.

Bruner, Edward M. 1956 Primary Group Experience and the Processes of Acculturation. American Anthropologist New Series 58:605–623.

Buckler, I. R. 1964 Measuring the Development of Kinship Terminologies: Scalogram and Transformational Accounts of Crow-type Systems. American Anthropologist New Series 66:765–788.

Callender, Charles 1962 Social Organization of the Central Algonkian Indians. Milwaukee Public Museum Publications in Anthropology, No. 7. Milwaukee, Wis.: The Museum.

Conklin, Harold C. 1964 Ethnogenealogical Method. Pages 25–55 in Ward H. Goodenough (editor), Explorations in Cultural Anthropology: Essays in Honor of George Peter Murdock. New York: McGraw-Hill.

Dozier, Edward P. 1954 The Hopi-Tewa of Arizona. Berkeley: Univ. of California Press.

Dumont, Louis 1957 Hierarchy and Marriage Alliance in South Indian Kinship. London: Royal Anthropological Institute of Great Britain and Ireland.

EGGAN, FRED 1937a Historical Changes in the Choctaw Kinship System. American Anthropologist New Series 39:34–52.

Eggan, Fred (1937b) 1962 The Cheyenne and Arapaho Kinship System. Pages 35–95 in Social Anthropology of North American Tribes. 2d ed., enl. Univ. of Chicago Press.

Eggan, Fred 1950 Social Organization of the WesternPueblos. Univ. of Chicago Press.

Eggan, Fred (1955) 1962 Social Anthropology: Methods and Results. Pages 485–551 in Social Anthropology of North American Tribes. 2d ed., enl. Univ. of Chicago Press.

Eggan, Fred 1960 Lewis H. Morgan in Kinship Perspective. Pages 179–201 in Gertrude E. Dole and Robert L. Carneiro (editors), Essays in the Science of Culture, in Honor of Leslie A. White.… New York: Crowell.

Eggan, Fred 1966 The American Indian: Perspectives for the Studyof Social Change. Chicago: Aldine.

Evans-Pritchard, E. E. 1951 Kinship and Marriage Among the Nuer. Oxford Univ. Press.

Firth, Raymond W. 1930 Marriage and the Classifica-tory System of Relationship. Journal of the Royal Anthropological Institute of Great Britain and Ireland60:235–268.

Firth, Raymond W. (1936) 1957 We, the Tikopia: A Sociological Study of Kinship in Primitive Polynesia.2d ed. London: Allen & Unwin. → A paperback edition was published in 1963 by Beacon.

Fortes, Meyer 1949 The Web of Kinship Among the Tallensi: The Second Part of an Analysis of the Social Structure of a Trans-Volta Tribe. Oxford Univ. Press.

Freeman, J. D. 1961 On the Concept of theKindred. Journal of the Royal Anthropological Institute of Great Britain and Ireland 91:192–220.

Friedrich, Paul 1963 An Evolutionary Sketch of Rus-sian Kinship. American Ethnological Society, Proceedings [1962]: 1–26.

Friedrich, Paul 1964 Semantic Structure and Social Structure: An Instance From Russian. Pages 131–166 in Ward H. Goodenough (editor), Explorations in Cultural Anthropology: Essays in Honor of George Peter Murdoch. New York: McGraw-Hill.

Galton, F. 1889 Note on Australian Marriage Systems. Journal of the Royal Anthropological Institute of Great Britain and Ireland 19:70–72.

Gifford, Edward W. 1922 California Kinship Terminologies. University of California Publications in American Archaeology and Ethnology, Vol. 18. Berkeley: Univ. of California Press.

Goodenough, Ward H. 1956 Componential Analysis and the Study of Meaning. Language 32:195–216.

Goodenough, Ward H. (editor) 1964 Explorations in Cultural Anthropology: Essays in Honor of George Peter Murdoch. New York: McGraw-Hill.

Greenberg, Joseph H. 1949 The Logical Analysis of Kinship. Philosophy of Science 16:58–64.

Hallowell, A. Irving 1937 Cross-cousin Marriage in the Lake Winnipeg Area. Pages 95–110 in Philadelphia Anthropological Society, Twenty-fifth Anniversary Studies. Edited by Daniel S. Davidson. Publications of the Philadelphia Anthropological Society, Vol. 1. Philadelphia: Univ. of Pennsylvania Press.

Hassrick, Royal B. 1944 The Teton Dakota Kinship System. American Anthropologist New Series 46:338–347.

Hockett, Charles F. 1964 The Proto Central Algonquian Kinship System. Pages 239–258 in Ward H. Goodenough, Explorations in Cultural Anthropology: Essays in Honor of George Peter Murdoch. New York: McGraw-Hill.

Homans, George C.; and Schneider, David M. 1955 Marriage, Authority, and Final Causes: A Study of Unilateral Cross-cousin Marriage. Glencoe, 111.: Free Press.

Josselin DE Jong, Jan P. B. DE 1952 Levi-Strauss’s Theory on Kinship and Marriage. Mededelingen van het Rijkmuseum voor Volkenkunde, No. 10. Leiden (Netherlands): Brill.

Kirchhoff, Paul 1931 Die Verwandtschaftsorganisation der Urwaldstamme Siidamerikas. Zeitschrift fur Ethnologic 63:85–193.

Kirchhoff, Paul 1932 Verwandtschaftsbezeichnungen und Verwandtenheirat. Zeitschrift fur Ethnologic 64: 41–71.

Krader, Lawrence 1963 SociaZ Organization of the Mongol–T urkic Pastoral Nomads. Indiana University Publications, Uralic and Altaic Series, Vol. 20. The Hague: Mouton.

Kroeber, Alfred L. (1901–1951) 1952 The Nature of Culture. Univ. of Chicago Press.

Kroeber, Alfred L. 1909 Classificatory Systems of Relationship. Journal of the Royal Anthropological Institute of Great Britain and Ireland 39:77–84.

Kroeber, Alfred L. 1917 Zuni Kin and Clan. American Museum of Natural History, Anthropological Papers, Vol. 18, part 2. New York:The Museum.

Leach, Edmund R. 1961 Rethinking Anthropology. London School of Economics and Political Science, Monographs on Social Anthropology, No. 22. London: Athlone.

Levi-Strauss, Claude 1949 Les structures elementaires de la parente. Paris: Presses Universitaires de France. Levi-Strauss, Claude (1958) 1963 Structural Anthropology. New York: Basic Books. → First published in French.

Lounsbury, Floyd G. 1956 A Semantic Analysis of the Pawnee Kinship Usage. Language 32:158–194.

Lounsbury, Floyd G. 1964 A Formal Account of the Crow- and Omaha-type Kinship Terminologies. Pages 351–393 in Ward H. Goodenough (editor), Explorations in Cultural Anthropology: Essays in Honor of George Peter Murdoch. New York: McGraw-Hill.

Lowie, Robert H. (1916) 1960 Historical and Sociological Interpretation of Kinship Terminologies. Pages 65–74 in Robert H. Lowie, Selected Papers in Anthropology. Berkeley: Univ. of California Press.

Lowie, Robert H. 1917 Culture and Ethnology. New York: Boni & Liveright.

Lowie, Robert H. (1929) 1959 Relationship Terms. Volume 19, pages 84–90 in Encyclopaedia Britannica. 14th ed. Chicago: Benton.

Lowie, Robert H. (1934) 1960 The Omaha and Crow Kinship Terminologies. Pages 100–110 in Robert H. Lowie, Selected Papers in Anthropology. Berkeley: Univ. of California Press.

Lowie, Robert H.; and Eggan, Fred 1963 Kinship Terminology. Volume 13, pages 407–409 in Encyclopaedia Britannica. Chicago: Benton.

Malinowski, Bronislaw (1913) 1963 The Family Among the Australian Aborigines: A Sociological Study. New York: Schocken.

Malinowski, Bronislaw (1913–1941) 1963 Sex, Culture and Myth. London: Hart-Davis.

Malinowski, Bronislaw 1930 Parenthood: The Basis of Social Structure. Pages 113–168 in Victor F. Calverton and Samuel D. Schmalhausen (editors), The New Generation: The Intimate Problems of Modern Parents and Children. New York: Macaulay.

Matthews, G. H. 1959 Proto-Siouan Kinship Terminology. American Anthropologist New Series 61:252–278.

Morgan, Lewis H. 1871 Systems of Consanguinity and Affinity of the Human Family. Smithsonian Contributions to Knowledge, Vol. 17, Publication No. 218. Washington: Smithsonian Institution.

Morgan, Lewis H. (1877) 1964 Ancient Society. Cam-bridge, Mass.: Harvard Univ. Press.

Murdock, George P. 1949 Social Structure. New York: Macmillan. → A paperback edition was published in 1965 by the Free Press.

Murdock, George P. 1957 World Ethnographic Sample. American Anthropologist New Series 59:664–687.

Murdock, George P. 1959 Evolution in Social Organization. Pages 126–143 in Anthropological Society of Washington, Evolution and Anthropology: A Centennial Appraisal. Washington: The Society.

Murdock, George P. (editor) 1960 Social Structure in Southeast Asia. Viking Fund Publications in Anthropology, No. 29. Chicago: Quadrangle Books.

Needham, Rodney 1962 Structure and Sentiment: A Test Case in Social Anthropology. Univ. of Chicago Press.

Parsons, Talcott 1943 The Kinship System of the Contemporary United States. American Anthropologist New Series 45:22–38.

Pehrson, Robert N. 1954 Bilateral Kin Groupings as a Structural Type: A Preliminary Statement. Journal of East Asiatic Studies 3:199–202.

Radcliffe-Brown, A. R. (1931) 1948 The Social Organization of Australian Tribes. Glencoe, 111.: Free Press.

Radcliffe-Brown, A. R. 1950 Introduction. Pages 1–85 in A. R. Radcliffe-Brown and Daryll Forde (editors), African Systems of Kinship and Marriage. Oxford Univ. Press.

Radcliffe-Brown, A. R. 1952 Structure and Function in Primitive Society: Essays and Addresses. London: Cohen & West; Glencoe, 111.: Free Press.

Radcliffe-Brown, A. R.; and Forde, Daryll (editors) 1950 African Systems of Kinship and Marriage. Ox-ford Univ. Press.

Richards, Audrey I. 1950 Some Types of Family Structure Amongst the Central Bantu. Pages 207–251 in A. R. Radcliffe-Brown and Daryll Forde (editors), African Systems of Kinship and Marriage. Oxford Univ. Press.

Rivers, William H. R. 1914 Kinship and Social Organisation. London School of Economics and Political Science, Studies, No. 36. London: Constable.

Schapera, Isaac (editor) 1963 Studies in Kinship and Marriage, Dedicated to Brenda Z. Seligman ... . Royal Anthropological Institute of Great Britain and Ireland, Occasional Papers, No. 16. London: The Institute.

Schmitt, Karl; and Schmitt, Iva 1952 Wichita Kin-ship: Past and Present. Norman: Univ. of Oklahoma Book Exchange.

Schneider, David M. 1964 The Nature of Kinship. Man 64:180–181.

Schneider, David M.; and Gough, Kathleen (editors) 1961 Matrilineal Kinship. Berkeley: Univ. of California Press.

Service, Elman R. 1962 Primitive Social Organization: An Evolutionary Perspective. New York: Random House.

Social Anthropology of North American Tribes. Edited by Fred Eggan. 2d ed., enl. (1937) 1962 Univ. of Chicago Press.

Spier, Leslie 1925 The Distribution of Kinship Systems in North America. University of Washington Publications in Anthropology, Vol. 1, No. 2. Seattle: Univ. of Washington Press.

Spoehr, Alexander 1941 Camp, Clan and Kin Among the Cow Creek Seminole of Florida. Anthropological Series, Vol. 33, No. 1. Chicago: Field Museum of Natural History.

Spoehr, Alexander 1942 Kinship Systems of the Seminole. Anthropological Series, Vol. 33, No. 2. Chicago: Field Museum of Natural History.

Spoehr, Alexander 1944 The Florida Seminole Camp. Anthropological Series, Vol. 33, No. 3. Chicago: Field Museum of Natural History.

Spoehr, Alexander 1947 Changing Kinship Systems: A Study in the Acculturation of the Creeks, Cherokee and Choctaw. Anthropological Series, Vol. 33, No. 4. Chicago: Field Museum of Natural History.

Spoehr, Alexander 1950 Observations on the Study of Kinship. American Anthropologist New Series 52: 1–15.

Tax, Sol (1937) 1962 The Social Organization of the Fox Indians. Pages 243–282 in Social Anthropology of North American Tribes. 2d ed., enl. Univ. of Chicago Press.

Tax, Sol (1955) 1962 From Lafitau to Radcliffe-Brown: A Short History of the Study of Social Organization. Pages 445–481 in Social Anthropology of North American Tribes. 2d ed., enl. Univ. of Chicago Press.

Titiev, Mischa 1943 The Influence of Common Residence on the Unilateral Classification of Kindred. American Anthropologist New Series 45:511–530.

Wallace, Anthony F. C.; and Atkins, John 1960 TheMeaning of Kinship Terms. American Anthropologist New Series 62:58–80.

Warner, W. Lloyd 1930 Morphology and Functions of the Australian Murngin Type of Kinship. American Anthropologist New Series 32:207–256.

Weil, Andre (1949) 1963 On the Algebraic Study of Certain Types of Marriage Laws (Murngin’s System). Pages 151–157 in Harrison C. White, An Anatomy of Kinship: Mathematical Models for Structures of Cumulated Roles. Englewood Cliffs, N.J.: Prentice-Hall.→ First published in French.

White, Harrison C. 1963 An Anatomy of Kinship: Mathematical Models for Structures of Cumulated Roles. Englewood Cliffs, N.J.: Prentice-Hall.

White, Leslie A. 1939 A Problem in Kinship Terminology. AmericanAnthropologist New Series 41:566–573.


The analysis of descent groups is a special aspect of the study of kinship. In the majority of preindustrial societies, kinsfolk are more than a narrow category of persons linked to an individual actor by filiation and siblinghood. They constitute a series of social groups that dominate the domestic organization and the process of socialization, the use and transfer of property, the settlement of disputes, religious activities such as ancestor worship, and certain political relationships. Because these kin groups influence so many aspects of social life, their structure and recruitment are highly important variables in the organization of technologically simpler societies. It is not surprising there-fore that in the last thirty years descent groups have been intensively studied by comparative sociologists and that there has been much discussion of the meaning of the term descent.

Comparative sociologists have generally accepted Rivers’ distinction between inheritance, the trans-mission of property; succession, the transmission of office; and descent, the transmission of kin-group membership. These three aspects of the authorized transmission of property, office, and group membership from one generation to the next are not entirely parallel. Inheritance, as usually understood, refers to transmission after death; the total process of transmission between holder and heirs, which includes certain types of transfer of property at marriage (as in the dowry), is referred to as devolution. Second, whereas succession and inheritance (or devolution) do not necessarily imply transfer between kin, descent does; it signifies group membership of a special kind, handed down, like property and office, from generation to generation.

Looking at the process of intergenerational trans-mission, from the standpoint of the junior generation, an individual can receive any particular set of rights in the following ways: (1) agnatic trans-mission, deriving exclusively from paternal kin; (2) uterine transmission, deriving exclusively from maternal kin; (3) bilateral (ambilateral) inclusive transmission, deriving from paternal and maternal sources; and (4) bilateral (utrolateral) exclusive transmission, deriving from either the paternal or maternal kin, depending on extraneous factors such as residence.

In any particular society, the inheritance of property, succession to office, and recruitment to kin groups usually follow the same mode of transmission. Indeed, the importance of descent groups lies in the fact that membership usually entails a claim upon basic productive and reproductive resources, as well as channeling succession to roles and offices. But this is not always the case. More-over, some differences arise in the transmission of these various rights from the fact that office is rarely divisible and its transmission is necessarily selective, whereas all members of a society usually belong to one of the series of kin groups. Property may be partible or impartible, and certain objects, such as weapons and cooking pots, are, like office, often sex-linked; on the other hand, the membership of kin groups is, in a sense, always partible and very rarely sex-linked. Moreover, it is allocated to individuals at birth (or soon after), not simply through birth.

Unilineal descent groups

Rivers used descent to refer only to kin groups whose members are recruited unilineally (or “unilaterally,” to use his own term); that is to say, groups such as clans, which are recruited either through male (agnatic) or through female (uterine) links. However, unilineal descent groups (UDGs) of this kind are not defined by the manner of recruitment alone; all the members of a particular group are also related to one another by common unilineal descent. A military regiment like the Fanti asafo company in Ghana or an occupational group like the widespread groups of Numu blacksmiths in west Africa may be recruited exclusively by ties of paternal kinship; they do not constitute a UDG unless the members are themselves linked with one another by similar ties and see themselves as having a common ancestry.

When these links are demonstrable and can be genealogically plotted, the group is known as a lineage; when the links are not demonstrable and descent is “putative,” the group is known as a clan. Some authors have attempted to establish the Roman term gens for a patrilineal UDG and clan for the matrilineal type, but most authorities prefer patrician (or patrilineage) and matriclan (or matrilineage ).

Distribution of UDGs

UDGs are found in all types of preindustrial economy and in all regions of the world; they occur among the aboriginal hunters of Australia and among the pastoralists of central Asia; among the shifting agriculturalists in Africa and among the irrigation farmers in India. They occur most frequently in pastoralism, then in agriculture, then in hunting and gathering, and only marginally in industrial communities. Ethnographic samples (Murdock 1957; 1963; Aberle 1961) support the hypothesis of those writers (for example, Lowie 1948; Radcliffe-Brown 1950; Forde 1947) who pointed to a positive association between the presence of descent groups and the amount and type of property to be transmitted. Forde (1947) observed that only when a certain threshold of stability and density of settlement is reached do the tendencies for unilineal transmission give rise to UDGs. Like other large-scale kin groups, they tend to have less importance where membership no longer provides rights in the means of production (i.e., when economies have become commercialized or industrialized). Descent groups also lose importance when their function in offense and defense is minimized by a highly centralized political system, and they no longer serve as a focus for ongoing local ties when spatial and social mobility increases.

As with all the wider ties of kinship, the disappearance of UDGs is linked with the growth of economic and social individualism in industrial societies, where a person tends to have direct ties with the political and economic agencies, mediated by specialized associations such as trade unions and political parties rather than by multifunctional units such as kin groups. In general, the role of UDGs diminishes with the importance of governmental institutions. Patricians were important within the state systems of the Zulu, early Rome, and China; matriclans played a significant part in the kingdoms of Ashanti and the Congo. But their characteristic functions of defense and retaliation were subsumed under central administrations. For example, homicide, a matter for self-help by the kin group in stateless and weakly centralized societies such as the bedouin tribes or Anglo-Saxons, became a criminal offense, dealt with by the king’s court.

Corporate groups

The functional importance of UDGs and similar social groups is sometimes expressed by the use of the term “corporate.” Following legal usage, Sir Henry Maine described the early Roman gens as a corporation in order to indicate its character as a property-holding unit that “never died.” But the word has also been used (a) to translate Weber’s Verband (a group with a hierarchy of legitimate authority); (b) for a group that regularly meets in either plenary or representative session; and (c) for a compact, localized group (as distinct from either a dispersed, or a non-localized, group, which often has few functions). The need for differentiating the part played by kin groups in different societies is recognized in the recent study of matrilineal systems edited by Schneider and Gough. The authors refer to UDGs as “descent units,” while the term “descent group” is used for “that descent unit or portion thereof which engages as a whole in activities with respect to which decisions must be made from time to time and in which all adult male members do not have equal authority” (1961, p. 4); the descent group, essentially a decision-making, or “organized,” group, is compared with what has elsewhere been called a “corporate descent group.” Murdock (1960) makes a similar distinction between corporate, occasional, and circumscriptive kin groups. The criteria that various writers have used to define “corporate groups” appear too general and too con-fused for most analytic purposes. The term can be restricted to certain aspects of the property-holding function, or else set aside altogether.

Subdivision of UDGs

While UDGs are usually segments of the society, in the sense that they are exclusive and exhaustive parts, they are rarely monolithic. Clans are often divided into smaller units (subclans) and into genealogically based units of at least five generations. In a patrilineal society, this basic lineage usually consists of a man’s father, grandfather, son, and grandson—that is, the forebears and descendants whom he actually encounters during his lifetime; it is this five-generation unit around which most systems of kinship terminology are constructed. But lineages may have a much greater generation depth; among the Nuer of the Sudan, twelve to fourteen generations are reported, and lineages of similar spread occur among other pastoral peoples of the Middle East, such as the contemporary bedouins and the Hebrews of the first millennium B.C.

The genealogy of a lineage differs in function from a chiefly pedigree. The latter acts as a validation of rights to an exclusive office; the former serves as a calculus for the relationship of group members. Consequently, a correlation exists between depth and span—that is, between order of segmentation and number of living members. Adjustment between the two is made by “telescoping” and similar mechanisms. Those ancestors whose presence in the genealogy is inessential for the reckoning of contemporary relationships gradually disappear from memory.

Morphologically, order of segmentation equals depth of genealogy, all levels of which have some importance for social action. But certain levels act as points of reference for specific subgroups of greater functional significance, for example, in the context of ceremonial food sharing, or landholding, or payments of compensation for homicide. The terms maximal lineage and minimal lineage are used sometimes for the morphological, and sometimes for the functional, differentiation of groups; terms for intermediate units, such as major lineage, minor lineage, and nuclear lineage, refer to the functional ordering of groups.

In such a merging series of subgroups of increasing inclusiveness, two groups that stand in contraposition at one level of segmentation merge when opposed to a group of a higher order of segmentation. This process of segmentation is one of opposition and identification and occurs in any series of “nesting” groups, whatever the political system; counties and cantons conflict on many local issues but unite when national interests are involved. But these processes have a greater importance when centralized administration is absent; hence, stateless societies that have a polyseg-mental structure (i.e., an extensive merging or nesting series), and particularly those based upon a developed lineage organization, are often known as segmentary societies, although the term acephalous is less ambiguous.

Some lineage systems extend to the limits of the society itself: the Tiv of eastern Nigeria all regard themselves as descendants in the male line of an eponymous ancestor, and the genealogy of his progeny lays out the complete outline of descent group structure. Sahlins (1961) sees the function of these all-inclusive lineages as one of predatory expansion. It is clear that the deeper the genealogy, the larger the lineage and hence the greater the possibility of mobilizing support on particular issues. But this mobilization may be equally as necessary in situations of defense as of offense. Extensive lineage ties are of particular significance where the population is sparse, and where local groups are bound to be small and therefore able to provide little support in emergencies. They are often important in pastoral societies where ecological conditions demand transhumance and thus the passage of flocks through the territory of one’s neighbors; the existence of ongoing ties of descent between such groups facilitates the movements that such a way of life demands.

Systems of double UDGs. UDGs are found in 60.6 per cent of the 483 societies in the “Ethnographic Atlas” (Murdock et al. 1963). In some societies (4.6 per cent), two sets of UDGs (patrilineal and matrilineal) are found side by side. In certain cases, one set of UDGs has relatively little significance. In others, both sets of UDGs have important functions; among the Yakö of eastern Nigeria, both sets are vehicles for the transmission of property (Forde 1950) and are therefore corporate, in Maine’s sense. Some writers have confined the term “double descent system” to those cases where both sets of UDGs are seen as property-holding corporations and refer to other systems where two sets occur as unilineal systems with secondary (or complementary) UDGs (Goody 1961).

In such full-fledged double descent systems, the two sets of UDGs have more or less differentiated roles. While both sets are exogamous at some level, rights to particular kinds of property are split. By this division, the basic means of production are vested in the patricians, while other wealth (money, livestock) is linked to the matriclans. Thus, in agricultural economies with full double descent systems, land passes within the patrician and movable property within the matriclan. As women marry out, the core of the local group consists of male agnates.

Systems of matrilineal descent. The “Ethnographic Atlas” shows a considerable predominance of patrilineal over matrilineal systems of UDGs: patrilineal UDGs (40.4 per cent), matrilineal UDGs (15.7 per cent), double UDGs (4.6 per cent), UDGs absent (39.3 per cent).

From many standpoints, agnatic transmission is more straightforward than uterine transmission, since dominant sex roles and descent links coincide. For the dominant sex (i.e., for males), the links are direct, running from father to children. Daughters move away at marriage, for virilocal residence overwhelmingly prevails where patrilineal UDGs occur. In this case, offspring belong to the UDGs of the husbands.

On the other hand, in matrilineal systems the sex that counts in the reckoning of group member-ship is nevertheless the inferior one in most social situations. It is men who hold the major positions of authority in domestic and political activities; yet socially the UDG reproduces itself through its female members, who, because of the incest taboo, must take outsiders as husbands or lovers. In the UDG, there is a greater interdependence of brothers and sisters.

If brothers and sisters remain together, as they traditionally did among the Nayar of southwest India, the elementary family cannot exist as a residential unit. But more usual than this duolocal solution to the “matrilineal puzzle” (the problem of combining exogamy and local descent continuity) are two other alternatives: (1) the woman joins the man (virilocal marriage); or (2) the man joins the woman (uxorilocal marriage). In the uxorilocal solution, as among the Hopi Indians of the southwestern United States, members of the dominant sex are living with affines and hence separated from their own matriclansfolk and “estate.” The virilocal solution leaves the men with their own clansfolk, but the children are now separated from their UDG. In societies with localized matrilineal UDGs, this form of marriage must be accompanied by “child return,” i.e., change of residence at adolescence or the uxorilocal marriage of at least one male member of the group of full siblings.

The nonresidential Nayar solution is rare (4 per cent of the matrilineal systems in the 1957 sample); residence with the wife’s kin is the most common (49 per cent), followed closely by residence with the husband’s kin (avunculocally—with his maternal kin—26 per cent; patrilocally—with his father—18 per cent). There is some tendency for avunculocal residence to be associated with the more stratified societies. However, it should be added that these residence types are rarely found in “pure” form; different patterns predominate, depending upon the position in the developmental cycle of the domestic group (e.g., Ashanti), the order in the sibling group (e.g., Yao), the relative status of husband and wife, and upon other factors.

Comparative studies show that matrilineal UDGs are associated with horticulture more frequently than with other types of economy. Although they are often found in stable fishing communities, they are virtually absent from pastoral economies— “the cow is the enemy of matriliny, and the friend of patriliny” (Aberle 1961). Matrilineal UDGs tend to disappear with the development of plough and irrigation agriculture; and, like all UDGs, they are absent from industrialized societies.

Secondary recruitment. We have so far assumed that descent groups are recruited according to a single criterion, which derives not simply from the observer’s classificatory schema but also from the ideology of the group itself. But probably all societies use additional modes of entry to increase the size of the group, to provide a particular member with an heir, or to regularize an individual’s change in domicile when this becomes anomalous in terms of the organization of UDGs. This last procedure is comparable to changing one’s nationality.

Various methods other than birthright are used to assimilate an individual into a descent group (Maine’s “legal fictions”). Assimilation may occur directly by the adoption of members of the relevant sex (i.e., males in patrilineal descent groups) or by the purchase of slaves. More often the assimilation is indirect: a woman in a patrilineal descent group is used to recruit personnel by contracting a form of limited marriage whereby all, or some, of her offspring are attached to her own natal UDG rather than to her husband’s (the “appointed” daughter of China, India, and the Middle East); alternatively the daughter may simply stay at home and produce children by a lover (“institutionalized illegitimacy”); a woman may “marry” another woman, who in turn takes a cicisbeo and breeds children for the female husband (as in Dahomey); or a servant or slave girl can substitute for the infertile wife (as Bilhah did for Rachel).

However, assimilation is rarely complete, and social situations such as sacrifices to ancestors may resurrect the differential status of members recruited by secondary means. Where the secondary mode of recruitment is by complementary filiation (that is, through the parent excluded from the reckoning of descent), and where this is widely practiced, the group is, in fact, recruited bilaterally, although de jure group ideology may impose a fiction of common unilineal descent upon the members. But, like recruitment itself, ideology is also a variable. In some societies, patrilineal attachment to the group may be expressed as a preference rather than an imperative, while other modes of recruitment are openly accepted, although of lower status. Here, we come close, de jure as well as de facto, to kin groups of a cognatic kind; neither in terms of recruitment nor of ideology (and this is as true of function as of form) can a hard and fast line be drawn between descent and other kin groups.

Monosexual kin groups

In the discussion of UDGs we have assumed that recruitment of the sibling group is inclusive; in other words, in patrilineal systems both male and female children be-long to the father’s UDG. But this is not the only way in which kin groups may be recruited and organized. Logically, the sibling group may be differentiated according to sex and age. Such mono-sexual groups may be differently organized for men and women; among the Apinayé of central Brazil, the brother belongs to a group of agnatically related males, the sister to a group of women linked by uterine ties. Or men may be linked cross-sexually to their mothers, and women to their fathers, to produce the alternating or cross-sexual system of the “rope” described by Mead for the Mundugumor of northern New Guinea. But such arrangements seem always to be ancillary to other more important sets of kin groups (Maybury Lewis 1960). Monosexual groups do not stand on their own, since the continuity of all kin groups must depend upon both sexes; and, therefore, the most significant kin groups in a society demand control of both sexes to ensure their character as ongoing units.

A further possibility exists, at any rate for agnatic systems: only one sex is organized on a monosexual basis (i.e., the males), leaving the females as a free floating element. But such an arrangement appears to run counter to the importance of the sibling group (brothers and sisters) in childhood and adolescence, and particularly to the sexual and marital prohibitions placed upon full and, commonly, classificatory sisters (i.e., sisters of fellow clansmen). The closest approximation to this logical possibility is the situation reported to exist in some agnatic systems of east Africa, socie-ties in which a woman is “incorporated” in her husband’s kin group, either immediately upon marriage or after she has borne children (Southall 1959). In early Rome, according to Maine, the wife was considered in manum viri and was, in law, the daughter of her husband. The resultant unit would correspond to a segment of what Murdock calls a “clan,” a compromise as distinct from a consanguineal kin group, for which he uses Lowie’s term “sib.”

All patrilineal systems transfer some rights in the woman from the bride’s to the groom’s kin group, i.e., rights in genetricem (over her procreative powers) and usually the right of bride removal. The very fact of virilocal residence is bound to place her in a structurally ambiguous position. Rather than using a simple dichotomy between UDGs that “incorporate” spouses and those that do not, it seems better to treat the degree of transfer of rights as a continuous variable and to make some assessment of the extent of formal assimilation and alienation of spouses, and the degree of their participation in the activities of the two groups involved. The limiting case would be the complete severance of a spouse from his or her natal group, but except in slave marriages it seems doubtful if this situation occurs in practice. Certainly the evidence from Africa, Rome, and China is equivocal here. The construction of such a scale would make it easier to interpret the divorce statistics of patrilineal systems where incorporation has been linked with rare divorce and nonincorpo-ration with higher divorce rates (Fallers 1957). But unless the concept of “incorporation” is broken down, one cannot be sure that it does not already involve a consideration of the ease and frequency of divorce.

Cognatic kin groups

Unilineal descent is not the only means by which kin groups are organized and recruited. Societies lacking UDGs are frequently referred to as bilateral (or cognatic). In such societies (and in unilineal systems) we find groupings based upon ties traced unrestrictedly, through both males and females at one time, or restrictedly through one parent or the other, with the particular selection depending upon nonkinship factors.

Ties traced through both males and females give rise to fluctuating personal kindred and descending kindred. The first is based upon the network of ties that radiate out from any individual, through kin of both sexes, and then descend from these ascend-ants to collateral kin. The most inclusive of these circles coincides with a man’s relatives, i.e., the whole field of consanguineal kinship, but less inclusive groups may emerge as significant social units. Personal kindreds, such as the sib in Anglo-Saxon England, are necessarily ego-orientated; each man’s grouping differs from the next (except a full brother), and the units are not true segments of the society. There is some discussion as to whether affines should be included in the definition of a kindred as such (see Freeman 1961, who uses “kindred” for the category of cognatic kin and speaks of the sib as a “kindred-based group”); for cross-cultural analysis, the problem is simply a matter of selecting a verbal tool, but for the analysis of a single society it is a question of establishing what happens in a particular case.

Like other cognatic kin groups, personal kindreds do not consist only of persons linked to a central actor by specific genealogical connections; two individuals may regard themselves as cousins simply because their respective fathers did. The distinction between genealogical and derived kinship is the basis of the formal distinction between lineage and clan.

The descending kindred is quite a different kind of unit. Unlike the personal kindred, membership is traced from an ancestor, ancestress, or an ancestral pair. Since ties are reckoned through both males and females, the resultant groupings are overlapping in terms of membership; but the members all see themselves as belonging to the same unit (“the descendants of X”). In the first respect, they resemble kindreds, in the latter a UDG.

Goodenough (1955) refers to such groups as “unrestricted nonunilineal descent groups” (Bohannan’s “omnilineal descent group”) and points out that they emerge mainly in the context of claims on fixed resources, especially ancestral land. Such nonunilineal descent groups (NDG, or Murdock’s “cognatic descent group”) become restricted by the introduction of additional criteria of entry, such as residence, parents’ residence, or the use of land rather than claims on land. If these alternatives are exclusive (as usually with residence), then the result will be a restricted kin group that is a true segment of the society; each individual is normally allocated to one such unit and one alone, as with UDGs.

Where membership of a restricted NDG is either nonexclusive (i.e., plural membership is allowed) or reversible (i.e., a man may alternate between the kin groups of his father and his mother), then the group is said to be ambilateral (Firth 1957). It is utrolateral where membership is exclusive and irreversible, as in the case of the Iban bilek.

Where the membership of a restricted UDG is organized around a geneaology, the group (a ramage) is, from this standpoint, the morphological equivalent of the lineage.

On the question of the similarities and differences between UDGs and restricted NDGs has turned the discussion as to whether groups of the latter type are properly called descent groups. Most American scholars adopt the wider usage, but some British authorities, such as Fortes and Leach, follow Rivers in making unilineality an essential criterion. They do this because first, “bilateral” groups may involve additional nonkinship criteria of eligibility. Second, these criteria involve choice or optation. It should be remarked that it is the existence of prescribed alternatives, rather than choice, that is important here; the selection of alternatives may derive from factors outside the control of the individual involved as in the case of parental residence. Third, the membership of these bilateral groups may overlap. Last, the processes of segmentation, fission, etc., are very different where the genealogical links consist of both males and females.

Whether descent is taken to refer only to UDGs or to include other ancestor-oriented kin groups is, however, not of great importance. What matters is that groups of this kind are of basic significance in the social structure of a large number of non-industrial societies. Because of their importance in the domestic, political, economic, and religious spheres, an understanding of the way they work and a knowledge of the principles of recruitment and organization are central not only to the under-standing of particular societies but of human his-tory itself and of the problems arising from the rapid social change of recent times. Their importance is explicitly recognized by the fact that pre-industrial societies are often classified by the types of kin group present, in particular by the presence of patrilineal or matrilineal descent groups, of both (double descent), or of neither (bilateral). This typology, while requiring much refinement, has had considerable relevance for explaining differences in such fields as domestic groups (Fortes 1958), incest rules (Goody 1956), rates of divorce (Gluckman 1950), relationship between adjacent generations (Malinowski 1927; Goody 1962), and developing agricultural organizations (Hill 1963).

Jack Goody

[Directly related are the entriesFamily, article onComparative structure; Marriage, article onComparative analysis.]


Aberle, David F. 1961 Matrilineal Descent in Cross-cultural Perspective. Pages 655–727 in David M. Schneider and Kathleen Gough (editors), Matrilineal Kinship. Berkeley: Univ. of California Press.

Bohannan, Paul J. 1963 Social Anthropology. New York: Holt.

Davenport, William 1959 Nonunilinear Descent and Descent Groups. American Anthropologist New Series 61:557–572.

Durkheim, Emile (1893) 1960 The Division of Labor in Society. 2d ed. Glencoe, 111.: Free Press. → First published as De la division du travail social.

Evans-Pritchard, E. E. (1940) 1963 The Nuer: A Description of the Modes of Livelihood and Political Institutions of a Nilotic People. Oxford: Clarendon.

Fallers, L. A. 1957 Some Determinants of Marriage Stability in Busoga. Africa 27:106–123.

Firth, Raymond W. 1957 A Note on Descent Groups in Polynesia. Man 57:4–8.

Forde, Daryll 1947 The Anthropological Approach in Social Science. Advancement of Science 4:213–224.

Forde, Daryll 1950 Double Descent Among the Yakö. Pages 285–332 in A. R. Radcliffe-Brown and Daryll Forde (editors), African Systems of Kinship and Marriage. Oxford Univ. Press.

Fortes, Meyer 1945 The Dynamics of Clanship Among the Tallensi: Being the First Part of an Analysis of the Social Structure of a Trans-Volta Tribe. Published for the International African Institute. Oxford Univ. Press.

Fortes, Meyer 1953 The Structure of Unilineal Descent Groups. American Anthropologist New Series 55:17–41.

Fortes, Meyer 1958 Introduction. Pages 1–14 in Jack Goody (editor), The Developmental Cycle in Domestic Groups. Cambridge Papers in Social Anthropology, No. 1. Cambridge Univ. Press.

Freeman, J. D. 1961 On the Concept of the Kindred. Journal of the Royal Anthropological Institute of Great Britain and Ireland 91:192–220.

Gluckman, Max 1950 Kinship and Marriage Among the Lozi of Northern Rhodesia and the Zulu of Natal. Pages 166–206 in A. R. Radcliffe-Brown and Daryll Forde (editors), African Systems of Kinship and Marriage. Oxford Univ. Press.

Goodenough, Ward H. 1955 A Problem in Malayo-Polynesian Social Organization. American Anthropologist New Series 57:71–83.

Goody, Jack R. 1956 A Comparative Approach to Incest and Adultery. British Journal of Sociology 7:286–305.

Goody, Jack R. 1961 The Classification of Double Descent Systems. Current Anthropology 2:3–26.

Goody, Jack R. 1962 Death, Property and the Ancestors: A Study of the Mortuary Customs of the Lodagaa of West Africa. Stanford (Calif.) Univ. Press.

Hill, Polly 1963 The Migrant Cocoa-farmers of Southern Ghana: A Study in Rural Capitalism. Cambridge Univ. Press.

Lowie, Robert H. (1948) 1960 Social Organization. New York: Holt.

Maine, Henry J. S. (1861) 1960 Ancient Law: Its Connection With the Early History of Society, and Its Relations to Modern Ideas. Rev. ed. New York: Dutton London and Toronto: Dent. → A paperback edition was published in 1963 by Beacon.

Malinowski, Bronislaw (1927)1953 Sex and Repression in Savage Society. London: Routledge; New York: Harcourt. → A paperback edition was published in 1955 by Meridian.

Maybury-Lewis, David 1960 Parallel Descent and the Apinayé Anomaly. Southwestern Journal of Anthropology 16:191–216.

Murdock, George P. 1949 Social Structure. New York: Macmillan.

Murdock, George P. 1957 World Ethnographic Sample. American Anthropologist New Series 59:664–687.

Murdock, George P. 1960 Preface. In George P. Murdock (editor), Social Structure in Southeast Asia. Viking Fund Publications in Anthropology, No. 29. Chicago: Quadrangle Books. → Papers presented at a symposium held at the ninth Pacific Science Congress, Bangkok, Thailand, 1957.

Murdock, George P. et al. 1963 Ethnographic Atlas. Ethnology 2:109–133, 249–268, 402–405, 541–548.

Radcliffe-Brown, A. R. 1950 Introduction. Pages 1–85 in A. R. Radcliffe-Brown and Daryll Forde (editors), African Systems of Kinship and Marriage. Oxford Univ. Press.

Richards, Audrey I. 1950 Some Types of Family Structure Amongst the Central Bantu. Pages 207–251 in A. R. Radcliffe-Brown and Daryll Forde (editors), African Systems of Kinship and Marriage. Oxford Univ. Press.

Sahlins, Marshall D. 1961 The Segmentary Lineage: An Organization of Predatory Expansion. American Anthropologist New Series 63:322–345.

Schneider, David M.; and Gough, Kathleen (editors) 1961 Matrilineal Kinship. Berkeley: Univ. of California Press.

Southall, A. W. 1959 A Note on Local Descent Groups. Man 59:65–66.


Pseudo-kinship includes those relationships in which persons are described or addressed by kin terms (or terms derived from the idiom of kin) but do not stand in such a relationship by virtue of the principles, however they happen to be conceptualized, of descent or marriage. It has been suggested that any relationship which employs a kin term is kin, and in that case there would be no pretext for the notion of pseudo-kinship. Nevertheless, such a viewpoint does not evade the need to distinguish between such relationships and consanguineal or affinal kinship. Every society has rules for ascribing kin status to its members, but these do not determine a relation of pseudo-kinship, which depends always upon the individual will of, at least, the initiator.

Three types of pseudo-kinship can be distinguished. (1) There is, first of all, the figurative usage of kin terms, which may be little more than a convention of speech or which may, on the other hand, designate a status within the society or within a specific context. (2) There are also customs whereby a person is given the status of kin by attribution rather than by birth—and this is commonly called “fictive” or “artificial” kinship. (3) There are also institutions which, in some ways, resemble kinship and are named by analogy with it, yet which possess a separate nature and accord a distinct status. These are sometimes regarded, rightly, as ritualized forms of friendship.

These three categories merge into one another; a figurative usage becomes an attribution of status when it enjoins behavior appropriate to kin, while the fiction of attributed kin status is no more than an analogy once the pseudo-kin are differentiated from the genuine. Again, ritualized friendship can exist without any analogy to kinship at all, as in the bond-friendship of Tikopia (Firth 1936), and the institutionalized “best friend” is sometimes no more than a trading partner. Trading partnership, on the other hand, may involve a relationship which carries the analogy with kinship to the point of imposing a bar to the intermarriage of descendants (Gorer 1938).

Figurative usage

Kin terms are used figuratively in many societies for the purpose of stressing some particular quality associated with kin. They may be employed momentarily, as when an old man is addressed as “grandad” in order to imply intimacy and age difference and when “son” or “daughter” is used to underline the seniority of the speaker. They may be particular to a given relationship, as when children are encouraged to address friends of their parents who adopt an avuncular role as “aunt” or “uncle.” Usages similar to all of these are found in other parts of the world.

Such usages are clearly distinguished from genuine kinship even when they are required by convention, since the kin relationship to the person addressed is irrelevant. They are mere expressions of attitudes, part of the common currency of personal relations. They imply a quality of behavior rather than a status, fraternity rather than the relationship of sibling. Kin terms may also provide a conventional title, as in the case of the Catholic priesthood, lay brotherhoods, or guilds, or in connection with a role in a festival, as in the Pawnee Hako (Fletcher 1904). Here the quality of the role begs an analogy with kinship.

Fictive or artificial kinship

Whether kinship is a purely social tie or whether it has any necessary biological basis is a question which has been much discussed, and the point of view taken necessarily affects what is to be regarded as pseudo-kinship. However, it seems advisable to consider as genuine kin those to whom the custom of the society ascribes such status, whatever the criteria for ascription may be, and to regard as pseudo-kin those who achieve the role otherwise. The distinction is not always easily made in societies such as the Eskimo, which appear to ascribe kin status on the basis of residence in a given household. When a person is adopted into a kin status to which he was not born, the definition of kinship becomes even more complex. He may acquire the status permanently and thus be established as a member of a kin group. The adopted individual becomes almost equivalent to one who was born into that status, is linked to all its members through ties of kinship, and addresses them by kin terms. In Western as in Far Eastern civilization, the desire for an heir is the common motive for such adoptions.

In Japan, the mukoyōshi, “groom-foster-son” as it has been translated, is adopted as a husband for the daughter when there is no son or no suitable son to whom the patrimony can be entrusted. Such an institution is reminiscent of the custom of the French peasantry, expressed in the phrase entrer gendre (to go in as a son-in-law), which defines the position of the bridegroom who joins the family farm in order to succeed his father-in-law. In Japan, however, there are other forms of adoption: the adoptive daughter who is later married to a “groom-foster-son,” even the kaiyōshi, “buyer-adoptive-son,” who takes over the patrimony of a bankrupt family and acquires kin status in it. A tenant can also become an adoptive son, and his descend-ants form a branch of the family which adopted him (Nagai & Bennett 1953; Norbeck & Befu 1958; Befu 1962). It must be pointed out that the fiction here is not complete, in the sense that the adoptive child does not become equivalent in every way to one born into the status. Though he uses the terminology of kinship to all the members of his adoptive parents’ kin, he is distinguished from the natural legitimate progeny in that his rights of inheritance are not always equivalent to theirs, and he retains his ties to his natural kin. Moreover, the very fact that he marries the daughter of his adoptive parents demonstrates that the groom-foster-son is not equated to a natural son with regard to the prohibition of incest. The oyabunkobun complex is to be distinguished from adoption in that it sets up a whole system of artificial kin-ship, quite separate from natural kinship, which endows a structure of patronage in economic and political life with the ritual and terminology of kinship. Bennett and Ishino (1963) have termed this “simulated kinship.”

Slaves of warlike tribes in Africa and North America were often adopted into the lineage of their captor and thereby acquired kin, but they did not by the same token acquire the status of son in all its aspects. Such a custom can be viewed as a way of providing a kin affiliation for those who have none in a society where this is essential for the conduct of social relations. Fictive kinship is also established for specific purposes, as for example in Islam, where it allows a man access to women who are not kin.

On the other hand, adoption may establish a person as a member of the family equivalent to its offspring, as commonly in Western civilization. In some societies, however (e.g., M. G. Smith 1962, p. 90), upbringing from infancy defines the status of child rather than procreation; filiation has dispensed with the supposition of physical maternity. In other societies, paternity is defined by social rather than genetic criteria—a fact which has led anthropologists to distinguish between “pater” and “genitor.” In neither case could one speak of fictive kinship, since by definition the kinship is genuine. The genitor may at the same time receive recognition as kin.

Genealogical relationships frequently involve a certain number of fictions whereby the record of descent is adapted so as to conform with the reality of the groups who believe themselves to share a common ancestor. In certain cases all the members of a single small community are addressed and referred to by a kin term, and their behavior is modeled upon that appropriate for kin, regardless of any putative cognatic tie. The tie of neighborship is thus given the value of a tie of kinship. This may be maintained even across differences of caste in India, as Freed (1963) has shown.

Finally, we may describe as fictive kinship the instances where persons who are related genealogically in one way adopt the forms of address and behavior prescribed for a different relationship. This is commonly the case where their roles in a household have constrained the members to mutate their kin ties to conform to their mutual behavior.

Ritual kinship

There exist, throughout the world, institutions which establish ties analogous to kin ties. The participants recognize a bond which is likened to, though it is not confused with, kinship. These are commonly defined under the headings of blood brotherhood and ritual coparenthood, or compadrazgo, and they are all best classified as ritual kinship. To refer to them as “fictive kinship,” as many authors have done, is to invite confusion, since no fiction is involved; these institutions are conceptually distinct from and frequently contrasted with natural kinship (Tegnaeus 1952, p. 13). Too often the ethnographer has written that persons tied by ritual kinship are “as brothers,” yet a closer look at the ethnography shows that they are not “as brothers” in any real sense (cf. Beidelman 1963). In order to use the word “fictive,” we must ascertain that the pseudo-kinsman does in fact acquire, at least in a relevant context, a status similar to that of the natural kinsman. Moreover, just as there are fictive forms and usages of consanguineal kinship, so there are of ritual kinship—and we can hardly speak of fictive “fictive kinship.” In fact, the role of the ritual kinsman often resembles that of the brother-in-law rather than the brother.

Blood brotherhood. Early studies of blood brotherhood (e.g., W. R. Smith 1885, pp. 47–58) were dominated and obscured by the assumption that primitive peoples regarded kinship, just as Westerners do, as a matter of blood. Even those who perceived that it was the magical associations of blood which accounted for its use in rituals continued to class together, as blood brotherhood, types of relationship which have in common only this detail of the rite which initiates them. Thus the discussions as to whether blood brotherhood was to be explained as a form of kinship, a legal contract, or a political alliance reinforced by a magical curse, a private pact, or a peacemaking ceremony were fundamentally confused, for blood can be used in many different rites (Evans-Pritchard 1933). The intimate identification implied by the exchange of blood represents a bond between two persons which gives each a mystical hold over the fidelity of the other. This bond does not as a rule entail the acquisition of ties with his kin. Indeed, in some instances the pact thus established remains secret; this is particularly the case where it is made for the purpose of political conspiracy or entry into a secret society. Such a pact is also found in certain marriage ceremonies and in love pacts (Sousberghe 1960).

On the other hand, such pacts are frequently a means of cementing peaceful relations or economic cooperation. They are therefore made with potentially hostile strangers, such as those early explorers of Africa who are often pictured entering into blood covenants, and between tribal chiefs whose followers are thereby committed to refrain from hostilities. The pact of blood may also be used to end a feud or a personal enmity. Depending upon the recognized extension of the tie engendered in this way, it serves to guarantee amicable terms between tribes, kin groups, families, or individuals. In this sense the blood pact may be put to the same purpose as alliance through marriage, as it is commonly a means of formalizing an instrumental or a purely affective relationship between two men. It is rarely found between two women, since the status of women commonly precludes such independence of action.

Covenanted comradeship depends upon the individual will of the participants (though they may be subject to their parents’ approval in this, or they may be committed to such a relationship in infancy by the parents). Yet it is almost always irrevocable. It prescribes reciprocal gifts and mutual trust, feelings of amity and the obligation of mutual assistance. In some instances it imposes an incest prohibition between the children of the comrades and even between their descendants; in others it enjoins a preference for their union. This type of comradeship is initiated by a rite which commonly involves the invocation of divine powers who bear witness. The rite itself consists in exchanging some personal substance, which is very frequently blood but also may be saliva, semen, and the like, or in the ingestion of some sanctified substance in common, a “loving cup” of blood or wine, for example, or an exchange of sacrifice. Foster brotherhood may sometimes create a similar tie. Thus, by analogy, a man could establish a tie of sacrosanct amity among the Caucasian tribes by pressing his lips to the breast of the mother of his enemy, an act which obliged the latter to forgo vengeance. Among the Plains Indians of North America, men who had sexual relations with the same woman thereafter called each other “brother,” a custom which may not be unconnected with the levirate.

A relationship similar to blood brotherhood may be initiated by a rite which centers upon the role of sponsor to certain persons, festivals, or celebrations, the passage together through a religious rite, or the exchange of sanctified gifts. Thus, ritual kinship may be established in central India when groups of young people “hear Ram’s name” together from a teacher (guru); they are then considered to be ritual brothers and sisters (Mayer 1960, pp. 139 ff.). The tie thus established bears a resemblance to Christian coparenthood. It is a particular form of a relationship which exists throughout India between a guru and his disciples, who are regarded as spiritual children to him and spiritual brothers and sisters to each other.

“Compadrazgo,”, or coparenthood. The most fully documented form of ritual kinship in the literature of anthropology is coparenthood. It derives from the Christian notion of spiritual kinship, one of the three kinds of affinity, the other two being affinity through blood and through marriage. Spiritual affinity originated in the sacrament of baptism and grew to be separate from natural affinity in the early Middle Ages, when parents were forbidden to stand as sponsors to their own children. The dogmatic basis for this development was provided by the opposition between natural generation, through which the sin of Adam was transmitted, and spiritual regeneration, through which it was cleansed. A corresponding opposition is found in those customs in which a spiritual parent must replace the physical parent. Spiritual kinship involves ties of two sorts, both of which impose an impediment to marriage: that which is established between godparent and godchild and that between godparent and natural parent, the co-parents, who address each other and are referred to as compadres (Spanish), compari (Italian), koumbari (Greek), kum (Serbian and Russian). The pretexts for establishing such ties vary. The Roman Catholic religion now recognizes spiritual affinity only as arising from the rites of baptism and confirmation, and only between the officiant and the child and between the godparent and god-child. Wider extensions of spiritual affinity are still recognized by the Eastern Catholic rites and by the Eastern Orthodox churches. The Council of Trent (1545–1563) greatly restricted the range of this regard to the persons between whom spiritual affinity was recognized (two spiritual parents of the same child, the children of compadres) and to the number of godparents. Spiritual affinity between coparents has lapsed in twentieth-century dogma in the Roman church, though not in the Eastern churches.

The social functions of this institution are to be distinguished from its spiritual function. In many parts of the world, popular custom ignores dogma and institutes the social bond of compadrazgo through rites which give rise to no spiritual affinity in the eyes of the church, though in the eyes of custom they may imply an incest prohibition. Thus in Spain the marriage godparents have no liturgical significance but have considerable social significance (except in Catalonia); conversely, the god-parents of confirmation required by the church frequently have no social functions. The baptismal godparents should act as godparents in the first communion and, according to the custom of certain regions, at the godchild’s marriage. A variety of preferential rules governs the selection of the godparents; sometimes they prescribe a member of the family of one or other parent. Sometimes the landowner is viewed as the appropriate godfather for the tenant’s child, and sometimes simply a friend becomes godparent (Foster 1953; Pitt-Rivers 1958). The significant differences are that in Italy and the northern half of Spain the tie between compadres is generally regarded as less important than that between godparent and godchild and that in Italy the godparent of confirmation is given more importance than the baptismal godparent (Anderson 1957).

On the whole the quality of the relationship of compadre is similar, although in Italy the spiritual rather than the material aspect is stressed; it is even regarded as preferable to choose someone with whom there will be no business entanglements. In both countries it is a relationship of trust, mutual assistance, affection, and respect— coparents traditionally use the formal third person in address. It is said that one can deny one’s compadre nothing, but the reciprocal duties are not stipulated. The duties of the godparent are more explicit, though they vary from place to place. In sum, they involve payment for the ceremony of fiesta, gifts of amulets and first articles of clothing, aid in starting out in life, and special duties in the case of godchild’s death during childhood—in fact, whenever the individual destiny of the child is at stake. The godparent also has the theoretical obligation to act as guardian in the case of the death of the parents. In exchange, the godchild owes only respect, obedience, and affection. The relations between compadres, on the other hand, are always conceived of as relations of equality, even where they cut across class barriers. Indeed their significance in such cases is precisely that they provide a tie of intimacy between persons of different classes which is not otherwise possible. Thus within the compadrazgo there are ritual relations between superior and inferior (godparent–god-child) or between equals (coparents), and there are social relations of ritualized friendship between equals and also of patronage between social classes. (Mintz & Wolf 1950). Therefore it is not surprising that the terminology of the compadrazgo should have acquired a slang usage to describe the maneuvers of intrigue, particularly in politics.

Coparenthood (kumstvo) in southeastern Yugoslavia establishes a lasting bond between two extended families (zadruge) through the customary obligation to renew the relationship, which is inherited from father to son. In this way a collective tie is maintained from one generation to the next. Ritual kinship is recognized, and the appropriate term used, between all members of the two families.

The bond of compadrazgo takes on its widest extension and its fullest significance in Latin America, as Tylor (1861, p. 251) was the first to note. The tie between compadres is generally of greater importance than that between godparent and god-child. It is sometimes recognized not only between the compadres, as hitherto defined, but also between each one of them and the ascendants of the other (certain Indians regard it as a relationship involving all the members of each family). In addition, the occasions on which compadrazgo is formed are many more than in Europe. Thus, baptism may provide the infant with up to three or even more pairs of godparents (padrinos de pila, de ceremonia, de arras, de vela, or de evangelios), all of whom become compadres of the parents. Fresh godparents are chosen for other occasions, particularly the rites de passage of the child. Compadrazgo may be formed on such pretexts as the first hair or nail cutting of a child, the ceremony when it is first carried astride the hip of an adult, a daughter’s ear piercing (this is found in Italy also), the dedication of a house, an altar, a religious image, a rosary, or a new truck, a healing ceremony, or simply the cosponsoring of a fiesta.

There are degrees of seriousness attached to different forms. Compadrazgo with the godparents of the font (padrinos de pila) tends to be treated as the most serious and sacred, while that which is established with the sponsor of a fiesta or the “godparent” at the blessing of a truck is no more than fictive compadrazgo which entitles the participants to call each other“compadre” as long as they feel well disposed but which is unlikely to be recognized by their ascendants. An influential man often has more compadres than he can remember. The degree to which compadrazgo implies a genuine affective tie whose sacred character is respected, rather than a perfunctory affability—and the distinction is inevitable in communities where middle-aged people are nearly all compadres of one sort or another—is sometimes expressed by the term “compadres de corazón” (compadres of the heart), which is reserved for relationships of true friendship and is usually ritualized by one of the more serious forms.

It is not easy to determine where to draw the line between genuine ritual kinship and fictive ritual kinship. Seen from the liturgical point of view, all the customary forms which do not involve spiritual affinity are fictive, but, if the authority of custom is accepted, then any form of compadrazgo established on a generally recognized pretext is genuine. Nevertheless, custom recognizes that some pretexts imply a more binding relationship than others, and it is wiser perhaps to regard the distinction here as a continuum rather than a dichotomy.

The structure of ritual kinship

All the forms of ritualized friendship, whether or not they use the terminology of kinship, derive not from birth but from the mutual feelings of individuals, guaranteed by the magical power of blood or the sacrosanctity of the rite. Even where there are preferential rules as to the choice of ritual kinsmen, it is subject always to the existence of the appropriate sentiments. This is true even in the extension of the compadrazo to the parents of compadres; the relationship is established only when they exchange the ritual embrace, and it is always there-fore possible to avoid. Ritual kinship commonly forms a series of dyadic or triadic ties, not an extended structure (cf. Foster 1963, p. 1285). Unlike natural kinship, it has no origin in descent from the past and no projection into the future where kin relations grow apart and realign themselves in the course of the domestic cycle. Even in the few cases (mainly blood brotherhood) where children take over the ritual kin ties of parents and the tie possesses a collective aspect, ritual kinship does not become part of the structure of kinship. This is illustrated by a provision concerning the collective Serbian kumstvo: a marriage between two families linked in this way brings the relationship to an end. Its restricted range and independence of other ties render it apt for the purposes of assuring or restoring peace, for when two of a man’s ritual kinsmen fight each other, he must remain friendly with both, since his ties with each are equivalent and independent. Equally, ritual kinship avoids being implicated in the internal dissensions of the kinship structure, for it involves no structural issues. It does not depend upon a network of rights and duties, but upon a reciprocal claim to favor and benevolence; it makes requests (not demands) and gifts (not payments) even where custom may define what these should be, and it is reinforced by supernatural sanctions only. The quality of the relationship differs from that of kin relations; “All that is mine is thine,” is the principle which inspires both the blood brother and the compadre, as in fact they say on occasions, but, even where rights are collectively vested in a kin group this is not the language of kinship. Indeed, where ritual kinship is superimposed upon natural kinship it endows the relationship between the two persons with a sacredness which it did not previously possess; for this reason blood brothers are sometimes said to be “closer than real brothers” and compadre frequently replaces a kin term in address, even, in parts of Latin America, brother or father.

We may ask why ritual kinship so often employs the language of consanguineal kinship, especially brotherhood. It does so by analogy with the ideal of brotherly love rather than with the reality of brotherly behavior, which opens with sibling rivalry and becomes in time subject to the dissonant demands of new families of procreation and the disintegration of the family of origin. By contrast, the tie of ritual kinship is immutable and free of ambivalence. Hence we can see that while adoptive kinship supplements the kinship structure, becoming, thanks to a fiction, what cognatic kinship is, ritual kinship complements it; it is what cognatic kinship aspires to, but cannot, be.

This statement should not blind us to the fact that ritual kinship is not always what it is thought it ought to be. It is as liable to exploitation as any form of friendship. Like friendship, it depends upon a balance of reciprocal favor. Its function is to provide through the attachment of personal feelings a basis for trust between individuals, which may or may not be put to the service of political or economic ends.

The institutions of ritual kinship do not flourish in modern urban society. In Europe, they lost their significance, save in the south, before the industrial revolution; “god-sib” became gossip, and commére came to mean the same. The south of Europe uses the words in the same figurative sense, and the institution retains its full significance today only in the rural areas. The same is true in principle of Latin America, though the influx of peasants into the cities has obscured the fact. Blood brotherhood is rapidly vanishing in Africa. Considering the varied social functions which the different forms of ritual kinship fulfill, it would be hard to enumerate all the causes of this development, but it appears that ritual kinship requires the environment of the closed community and particularistic rather than universal relations (Eisenstadt 1956). The oyabun-kobun system of Japan is exceptional in this regard in representing an adaption of a traditional form of ritual kinship to fulfill a function in modern industrial organization.

In summary, pseudo-kinship is a residual category that includes a number of very different institutions which have in common only the fact that they are likened to kinship by the people themselves. To sort out this rag bag into sociological categories the following distinctions were required: whether the relationship is universal or particularistic; momentary, temporary, restricted as to context, or continual and permanent; dictated by custom or voluntary; extended to the kin of the primary participants or purely individual; and whether it entails a role within the kinship system or one contrasting with the role of kin.

The analytical distinctions may be clear, but the facts are sometimes ambiguous, and this occurs especially when the institution is found in transition. Pseudo-kin relations change their nature, and those which were once sacred are abused and devaluated; words which once denoted ritual kin extend to senses in which the original meaning is lost.

Julian Pitt-Rivers

[See alsoAdoption; Paternalism.]


Anderson, G. 1957 II Comparaggio: The Italian God-parenthood Complex. Southwestern Journal of Anthropology 1:32–53.

Befu, Harumi (1962) 1963 Corporate Emphasis and Patterns of Descent in the Japanese Family. Pages 34–41 in Pacific Science Congress, Tenth, Honolulu, 1961, Japanese Culture: Its Development and Characteristics. Edited by Robert J. Smith and Richard K. Beardsley. Chicago: Aldine.

Beidelman, Thomas O. 1963 The Blood Covenant and the Concept of Blood in Ukagaru. Africa 33:321–342.

Bennett, John W.; and Ishino, Iwao 1963 Paternalism in the Japanese Economy. Minneapolis: Univ. of Minnesota Press.

Corblet, Jules 1881–1882 Histoire dogmatique, liturgique et archéologique du sacrement de baptême. 2 vols. Paris: Palmé.

Eisenstadt, Shmuel N. 1956 Ritualized Personal Relations. Man 56:90x2013;95.

Evans-Pritchard, E. E. 1933 Zande Blood-brotherhood. Africa 6:369–401.

Firth, Raymond W. 1936 Bond-friendship in Tikopia. Pages 259–272 in Custom Is King: Essays Presented to R. R. Marett on His Seventieth Birthday. London: Hutchinson.

Fletcher, Alice C. 1904 The Hako: A Pawnee Ceremony. U.S. Bureau of American Ethnology, Twenty-second Annual Report, 1900–1901, part 2.

Foster, George M. 1953 Cofradia and Compadrazgo in Spain and South America. Southwestern Journal of Anthropology 9:1–28.

Foster, George M. 1963 The Dyadic Contract in Tzintzuntzan: Patron-Client Relationship. American Anthropologist New Series 65:1280–1294.

Freed, Stanley A. 1963 Fictive Kinship in a North Indian Village. Ethnology 2:86–103.

Gorer, Geoffrey 1938 Himalayan Village: An Account of the Lepchas of Sikkim. London: Joseph.

Hammel, EUGENE A. 1966 Ritual Co-parenthood (Kumstvo) in Serbia. Unpublished manuscript, Univ. of California, Department of Anthropology.

Mayer, Adrian C. 1960 Caste and Kinship in Central India: A Village and Its Region. London: Routledge; Berkeley: Univ. of California Press.

Mintz, Sidney W.; and Wolf, Eric R. 1950 An Analysis of Ritual Co-parenthood. Southwestern Journal of Anthropology 6:341–368.

Nagai, Michio; and Bennett, John W. 1953 A Summary and Analysis of The Familial Structure of Japanese Society, by Takeyoshi Kawashima. Southwestern Journal of Anthropology 9:239–250.

Norbeck, Edward; and Befu, Harumi 1958 Informal Fictive Kinship in Japan. American Anthropologist New Series 60:102–117.

Pitt-Rivers, Julian 1958 Ritual Kinship in Spain. New York Academy of Sciences, Transactions 2d Series 20, no. 5:424–431.

Smith, Michael G. 1962 Kinship and Community in Carriacou. New Haven: Yale Univ. Press.

Smith, William Robertson (1885) 1903 Kinship and Marriage in Early Arabia. New ed. London: Black.

Sousberche, LÉon DE 1960 Pactes de sang et pactes d’union dans la mort chez quelques peuplades du Kwango. Académic Royale des Sciences d’Outre-mer, Brussels, Classe des Sciences Morales et Politiques, Mémoires New Series 22, no. 2.

Tegnaeus, Harry 1952 Blood-brothers. Stockholm: Statens Etnografiska Museum.

Tylor, Edward B. 1861 Anahuac: Or, Mexico and the Mexicans, Ancient and Modern. London: Longmans.