KINSHIP, RELIGION, & ETHNICITY

Lyle Steadman

Department of Anthropology

Arizona State University

Human Behavior and Evolution Society Meeting

Albuquerque, New Mexico

July 27, 1992

When an Armenian favors an Armenian because he is an Armenian over a non-Armenian; or a person who is Jewish favors a Jew because he is Jewish; or a Chinese person favors a Chinese because he’s Chinese, most people (although perhaps not most social and biological scientists) would consider that an example of kinship behavior—explaining it by some such phrase as “blood is thicker than water.” Ethnic favoritism may be both the main benefit of being a member of an ethnic group and, at the same time, what some people describe as the main problem in the world today. Ethnic favoritism may also be an example of one of the major puzzles in evolutionary biology: altruism that can be explained by neither kin selection nor reciprocal altruism.

Given the importance of both the positive and negative effects of such—dare we say it—kinship behavior, one would think it would be at the forefront of research today. Instead, the anthropological study of kinship has languished for years, driven into confusion partly by a failure to agree on what is meant by kinship, kinship behavior, kinship terms, etc., and partly by an anti-comparative approach emphasizing the uniqueness of each culture. At the same time, the study of ethnicity has done little to increase our understanding of ethnic competition and violence that so troubles the people of many countries of the world.

Today, I would like to examine kinship and the relationship it has to ethnicity, focusing particularly on traditions, that is, on behavior that tends to be transmitted from ancestor to descendants, and hence would seem to be subject to Darwinian natural selection. When an inheritable trait or phenotype promotes the leaving of descendants, it should tend to increase in frequency through generations.

A kinship relationship is based on birth, not on the amount of shared blood or number of shared genes. Except for a few modern paternity cases, no one uses shared blood to identify kinship relationships (even though humans in many societies use the metaphor of blood when speaking of kinship). And full brothers may share all or none of their genes, even though they will average about fifty percent. A true grandparent who in fact shares none of his genes with his grandchild (which is possible) is just as much a true grandparent as one who shares half of his genes with his grandchild. A kinship relationship is based on birth—one individual coming from another individual. A kinsman is one who is linked by one or more birth links to oneself and therefore is either an ancestor (including a parent), a descendant (including one’s child), or a co-descendant (one who shares a common ancestor). Siblings, uncles and cousins, for example, are co-descendants. So long as two individuals, living or dead, are identified as being either co-descendants, or one is the ancestor of the other, they are kinsmen, no matter how distant. If we assume that chimpanzees and humans have a common ancestor, then we assume that they are kinsmen of one another. The difference between closer kin and more distant kin is simply that [closer] kin have fewer birth links separating them.

Now, for kinship behavior. It seems that three elements must be present for behavior to be identified as kinship behavior. First, the behavior must be social, not merely interaction (mere interaction, of course, can include violence between enemies). Indeed, I would say that the behavior must be altruistic—at least one of the individuals favoring the other at his own personal expense.

The second element is even more obvious: the individual favored must be a kinsman—that is, an ancestor, descendant, or co-descendant. But these two elements alone—altruistic behavior towards a kinsman—do not appear sufficient to constitute an example of kinship behavior. For example, given that a tribe consists of individuals who share a common ancestor, any altruistic act toward a fellow tribesman would have to be seen as kinship. That is, it would not be possible to exhibit non-kinship altruism toward a fellow tribesman. Let me enlarge on this issue.

Given that all 400 or so Bantu speaking tribes in Africa today descend from one tribe some two thousand years ago, as argued by Harold Schneider, the individuals in these tribes today would all be kinsmen of one another. Hence any altruistic act between individuals of any of these tribes would have to be seen as kinship behavior. To make this problem still more obvious, because all mammals are presumably related through a common ancestor, feeding one’s dog would have to be seen as an act of kinship. Indeed, going to the extreme, given that all life on earth may have a common ancestor, any act of altruism toward any organism would have to be seen as kinship.

This is absurd, and is not what is meant by kinship behavior. The use of the term kinship does not imply every act of altruism. Surely, the question is not whether every act of altruism that happens to be between kinsmen is kinship behavior but whether the behavior is aimed at helping an individual because he is a kinsman. That is, we must distinguish the behavior that is aimed at helping an individual because he is a kinsman from altruism directed toward an individual who just happens to be a kinsman. We feed our dog not because he is our (distant) kinsman but for other reasons entirely. We don’t feed our dog because he is a mammalian co-descendant when at the same time we eat cows or pigs, who might even be closer genealogically. So how can altruistic behavior directed towards a kinsman be kinship behavior?

I suggest that the final element necessary for behavior to be identified as kinship must be that the degree of altruism is somehow correlated with the genealogical distance between the individuals involved. That is, the degree of help must be related to the number of birth links separating the individuals. If one favors a genealogically distant tribesman over one’s close relatives, the behavior is not kinship.

Interestingly, given this perspective, even sexual reproduction may be seen as kinship behavior. To the extent that sexual reproduction can be seen as one individual at least partially helping another to leave descendants, such altruism or mutual altruism—cooperation—might be said to be a form of kinship cooperation. After all, to be successful, mating must occur between individuals not too distant nor too close genealogically, for in the former there is the problem of sterile hybrids, such as mules, and in the latter, inbreeding depression. That is, individuals, especially females, should be selected to prefer to mate with kinsmen at an optimal genealogical distance.

For altruism or cooperation to be correlated with genealogical distance, such distance must be identified by the participants. How are kinsmen and number of birth links identified? Female mammals identify their offspring at their birth and offspring come to identify their mother and their siblings through their contact with their mother. Through their relationships with one another, they may come to identify each other’s offspring (their nephews and nieces), and their offspring may come to identify each other as cousins. On the other hand, when the male’s offspring are identified, they are identified by their birth to a female with whom he has had prior sexual relations. All this is true for humans and probably best explained by Hamilton’s inclusive fitness theory. But in addition, humans everywhere identify a wider range of kinsmen through the use of a set of kin terms, such as mother, father, uncle, aunt, first cousin, second cousin, etc. The literal use of kin terms always implies precise genealogical distance. For example, in a group that I lived with for some time in Papua New Guinea, the Hewa, like in many societies, the term for father, aita, is extended to all identifiable male kin of my father on his generation. However, two things must be pointed out: in order to use the term “father,” or aita, accurately it is necessary to know precisely whether a person is in fact on your father’s generation, and this requires accurate genealogical knowledge—the actual number of birth links to the common ancestor of your father and his kinsmen. Furthermore, this extended usage of the term, even though very precise, is clearly metaphorical, for they distinguish the person we would call the actual father in the same way we do, as aita papi (literally, “true father”). This extended metaphorical usage of the term is similar to our use of the term “uncle” for the husband of one’s genealogical aunt. The kinship-like relationship to that “uncle-in-law” is based not on birth but his existing marriage to our aunt. That is, he is only a metaphorical “uncle.” Again, the literal meaning of a kin term implies precise genealogical distance. Our term “first cousin,” for example, implies an individual who is exactly two birth links away from an ancestor from whom I have descended exactly two birth links. Our general term “cousin” implies an individual who is two or more birth links away from an ancestor from whom I have descended two or more birth links. Because the literal use of kin terms depend on precise genealogical information, their use is limited, perhaps often to only fifty or sixty individuals for any Ego [or individual person]. All societies in the world appear to have such a set of kin terms, and therefore their use, apparently, has contributed to the success of human ancestors in leaving descendants for tens of thousands of years.

In addition to the use of kin terms, individuals in most societies use another verbal system to identify kinsmen. Descent names, clan names, ancestral names, or, more familiar to us, family names, are used to identify a much broader range of kin than kin terms. A descent (or ancestral, or clan) name is distinguished by the fact that it identifies an individual with a line of ancestors, usually of one sex only, often a patriline [father’s male ancestral lineage], sometimes a matriline [mother’s female ancestral lineage]. The main function or significance of this name that identifies an individual with a line of mostly dead ancestors is that it is used to identify living co-descendants, sometimes genealogically very distant. For example, in some east African societies, such as the Nuer, one ancestral or clan name may identify ten thousand people as co-descendant. But such a name is used to identify many more individuals than those of one clan. For example, if one’s mother is of clan X, all those of clan X are also identified as one’s co-descendants. In addition, of course, the offspring and grandchildren of each of your clan members, regardless of their sex, are also identified co-descendants. To bring this fact home, let me give a more familiar example. If you attend a family picnic identified by your last name, say Palmer, each person will be a descendant of a Palmer (perhaps a grandmother named Palmer). The only other people present will be the spouses of these co-descendants. That is, with few [exceptions], all those attending a family picnic will be either identified co-descendants or their spouses.

Identification of co-descendants is important in every society, for most living kin are co-descendants. The actual residential groups in the New Guinea Hewa, for example, consist almost totally of co-descendants and their spouses. While the Hewa have patrilineal clans, clans are never residential groups. The residential groups in the Hewa always consist of a set of co-descendants and their spouses. The identification of this set is based on a clan name: either the co-descendants bear that name or they have an ancestor who bore that name. While the residential group is referred to by the clan or ancestral name that identifies their common ancestor, the clansmen themselves, those who actually bear the name, may reside in any number of residential groups in which they are identified co-descendant. Or they may even live with their spouse. What I am arguing is that the basic significance of a clan is not that it is a group, but that it is a name used to identify living co-descendants. Co-descendants are all kinsmen of one another and can live together, in contrast to a group of individuals, only some of who trace kinship to one another. For example, my cousins on my father’s side are not identifiably related to cousins on my mother’s side. In a society whose relationships are based on kinship, they could not live together.

Ancestral, or clan, or family names thus permit the identification of many co-descendants, far more than simply those who bear the name. For example, I am an identifiable co-descendant of every clan to which every one of my identified ancestors belonged—two parents, four grandparents, eight great grandparents, etc. In addition, in many societies with large clans, segments of clans are also identified by ancestral names, such as sub-clans or lineages. Such names serve to distinguish closer sets of co-descendants from more distant ones within the same clan. Thus, the function of lineage or sub-clan names seems clear: to discriminate sets of kin on the basis of their genealogical distance from you.

To summarize the difference between kin terms and ancestral names: they both serve to identify kinsmen, kin terms by specifying precisely the number of birth links separating individuals, ancestral names, by identifying individuals with a particular line of ancestors, are used to identify a set of living co-descendants. Individuals identified by either kin terms or ancestral names are assumed to be closer genealogically than those not so identified.

More distant genealogical relationships are identified by cultural traditions and even physical traits. A tribe is usually distinguished by a distinctive language or dialect. In many New Guinean tribes, or societies, males decorate their bodies in a tribal way, with the individuals of each clan decorating themselves somewhat distinctively from the other clans. Thus their decoration, which seems to be mainly aimed at attracting females, communicates their clan and tribal ancestry, allowing females thereby to evaluate their genealogical position for marriage.

For non-human animals, distant genealogical relationships involving sexual reproduction and competition are identified by physical and sometimes also cultural traits. Along the west coast of the United States male white-crowned sparrows sing distinctive songs that are learned from their fathers or other males in their neighborhood. A mating between individuals in the two different subspecies whose boundaries overlap near Seattle, Washington, should produce less successful offspring that matings within their subspecies. According to Barash, the northern subspecies nest only once a year and the males are appropriately very aggressive in defending their nests. They must wait a year before they can have offspring again. The southern subspecies, on the other hand, nest two or three times a year, and the males take fewer risks to protect their nest from predators, for they can produce another set of offspring in a few months. Inter-subspecies matings, therefore, should produce offspring exhibiting inappropriate and therefore less successful behavior, with the sons growing up to either defend their nest too much or not enough. Females should be able to identify their genealogical distance to a particular male by, among other things, the culturally learned songs he sings. Perhaps this effect—communication of a male’s ancestry—can account for the selection for the culturally learned songs among the white-crowned male sparrows in general. Something similar may have led to similar behavior in other birds that copy the sounds around them when young, such a parrots.

An ethnic group, like a tribe, is distinguished by common ancestry; individuals in an ethnic group, like individuals in a tribe assume they are a set of co-descendants. And given the regular occurrence of marriage though time within that set, the assumption is surely true. An ethnic group is also culturally distinctive. Their clothing and language may communicate their ancestry (such as the Amish, Mennonites, and Hutterites) and their family, or ancestral names often indicate their ethnic identity; a name ending in “ian,” for example, such as Melikian, often is seen by Armenians as identifying an individual as Armenian. Such cultural features are used by the individuals involved to identify genealogical distance, and on the basis of which they may systematically discriminate and favor members of their own ethnic group over outsiders.

But kinship identification alone is not sufficient for kinship behavior to occur. In traditional societies traditions often encourage at least some degree of altruism directed toward distant kin. And religion seems to play the key role in this area.

But before looking at the possible religious origins of this behavior, let me return for a moment to an issue already raised. Much human behavior, including much of our kinship behavior is traditional. That is, it is encouraged and transmitted by ancestors to their descendants. Such behavior, therefore, should respond to natural selection. When a tradition helps individuals to leave descendants it should tend to increase in frequency. I realize there is some considerable reluctance to use the term natural selection to explain the frequency of traditional behaviors or traits, perhaps because they are learned and copied from other individuals. But I can find no more appropriate term in English to account for why Hutterite traits, for example, have increased so dramatically in South Dakota and northern Nebraska in the last century or so. These Hutterite traits, themselves, have contributed directly to the fact that the average Hutterite female has had more than ten surviving offspring, meaning that the population has pentupled every generation. This increase has dramatically increased the frequency of Hutterite traits or traditions. The genes on which these traits depend may well be widespread in human populations, and not distinctive to the Hutterites. This same kind of argument can be used to account for the extinction of the Shakers, a religious group that forbade any sexual activity whatsoever. By not leaving any descendants, both Shakers and their distinctive traits exist no more. Hutterite traits have been extremely fit, Shaker traits have not. It seems to me, there is no simpler explanation for the great difference in the current frequency of these two sets of traits than natural selection, as proposed by Charles Darwin. Darwin was not aware of genes and therefore, apparently, knowledge of genes and their changed frequencies was not necessary to the understanding of natural selection, as used by Darwin. Furthermore, because a traditional trait, in contrast to a gene, can be transmitted at a 100 per cent frequency, that is to all one’s children, if it is very fit its frequency can increase very rapidly. A gene will only be transmitted to half one’s offspring. It seems to me that within a set of co-descendants, if an inheritable trait, on average, increases the descendants of that set of co-descendants more than alternative traits in that set, or in neighboring sets of co-descendants—or populations—then that trait should tend to increase in frequency. My aim here, however, is to try to identify the important relationship between kinship, ethnicity, and religion.

How did these widespread traditions come about? Religion may have been the catalyst that extends kinship altruism, or at least favoritism to distant kinsmen. It is possible that the basic religion in most, if not all tribal societies, including the Hewa, is ancestor worship. Ancestor worship, as Evans-Pritchard has pointed out, is a kinship activity. It is a cooperative ritual of living co-descendants, directing their worship, and attention to their common ancestor. What distinguishes such activity as religious is the supernatural claim of these co-descendants that they can influence and be influenced by their dead ancestor. Such cooperative rituals invariably increase the likelihood of future cooperation between the participants in the ritual. But just as importantly, the ancestor is invariably called “father,” or “grandfather,” thus implying that the participants are siblings, or at most close cousins of one another. The use of such terms is aimed, it seems, at encouraging that behavior which normally occurs between close relatives, to be directed towards more distant co-descendants. The use of “brother” for fellow clansmen (clan brothers) is part of this encouragement. Thus, the most important effect of ancestor worship seems to be that it encourages “family-like” cooperation between (often distant) co-descendants, and by promoting respect for their common ancestor, they promote the acceptance of their ancestor’s behavior, his or her traditions. Such worship may be extended to tribal ancestors, and thus family-like cooperation may be encouraged between tribal brothers.

This seems to be the state of religion in tribal societies and hence, on the basis of the evidence available, among our distant ancestors. The non-agricultural Australian Aborigines, for example, direct virtually all religious behavior toward their alleged ancestors. It has been pointed out, for example, that the word often translated as the “Dreaming,” to describe their religion, means literally “ancestors,” and is used to identify all things sacred, all things important to ancestors—all their rituals, boys newly initiated, women who have just given birth, all they describe as “law.”

Ancestor worship, which promoted close, or family-like cooperation between co-descendants not in the same family, but in the same tribe, probably characterizes the religion of most if not all tribal societies. Some few thousand years ago, in a few places in the world, a new religious phenomenon began to occur. On the basis of available evidence, prophets began to appeal to individuals from different tribes, arguing that they, the prophets, spoke for the ultimate ancestor of everyone, and that their followers, as children of this first ancestor, should begin to act like siblings of one another. That is, these prophets tried, and to some extent succeeded, to create close kinship behavior between individuals who were only alleged or supernatural co-descendants. These “modern,” proselytizing religions have had a tremendous impact on the world, for they have created huge groups of cooperating individuals out of different tribes. Zoroaster was the prophet of Zoroasterianism, which was the basis of the Persian Empire. Viracocha, from Tiahuanaco (in present Bolivia), seems to have been the prophet at the base of the Inca Empire. Mohammed was the prophet of Bedouins who were transformed into huge Muslim groups. Thus, modern religions, created by prophets, incorporate individuals of different tribes into one large family. Family kinship terms—father, mother, brother, sister, children—are regularly used to encourage family-like altruism within these groups. Because these cooperative relationships can endure, individuals come to depend on one another. This cooperation has often led to great social, economic, and technological specializations. Obviously, in many of these groups, many individuals have had great success in leaving descendants.

When modern religions stop proselytizing and become endogamous, when they stop bringing in new converts and marry only co-followers, they become ethnic groups, which can be almost identical to tribes. When competition occurs with outsiders, and is seen as a threat to one’s ethnic group, extreme sacrifice can be encouraged to protect “one’s own,” witness the two religious groups in Northern Ireland. Culturally they are almost identical, even in their religions. But because the two categories have been endogamous for centuries, a Catholic has no Protestant kin, nor do Protestants have Catholic kin. They are not locked together by kinship. When competition occurs, and individuals see a threat to their “kin,” they may sacrifice their lives in their defense. The same can be said of the situation in Sri Lanka, with the Sinhalese-speaking Buddhists and the Tamil-speaking Hindus. And the same is true of the situations in Yugoslavia, South Africa, and the carcass of the Soviet Union. Indeed, clashes between such kinship groups characterize much of the modern world.

What I have tried to do in this paper, using an evolutionary approach, is to identify the relationship between kinship, religion, and ethnicity. I have argued not only that kinship behavior implies altruism directed at a kinsman, no matter how distant, but the degree of altruism must be correlated with genealogical distance.

Humans everywhere identify a limited number of kinsmen by kin terms, and almost everywhere a greater number of kin by ancestral names. Ancestor worship promotes cooperation between these identified co-descendants. Modern religions, created by a prophet out of different tribes, extend this kinship-like cooperation to all co-followers. “Modern” religious groups have replaced or engulfed tribes all over the world. When such a religion stops proselytizing, and becomes endogamous, it becomes an ethnic group. Whether or not it ceases to proselytize or becomes endogamous it consists of individuals encouraged to favor their “kin” over outsiders and to defend them when competition occurs. Such defense may be extremely violent.