The Elementary Structures of Kinship Claude Levi Strauss

The Elementary Structures of Kinship - Claude Levi-Strauss (Highlight: 134; Note: 0.

───────────────

II. The Problem of Incest

So much for natural sanctions. As for social sanctions, they are based so little upon physiological considerations that among the Kenyah and Kayan of Borneo, who condemn marriage with mother, sister, daughter, father’s sister or mother’s sister, and with brother’s daughter or sister’s daughter, ‘in the case of those women who stand to him in any of these relations in virtue of adoption, the prohibitions and severe penalties are if possible even more strictly enforced’

loth

Consequently, in a small, stable, endogamous population, as exemplified by many primitive societies, the only risk in marriages between consanguines arises from the appearance of new mutations, a risk that can be calculated since the rate of appearance is known. But the chances of finding a recessive heterozygote within the group are slimmer than would attend marriage with a stranger

The economic systems of some primitive or archaic societies severely limit population size, and it is precisely for a population of such a size that the regulation of consanguineous marriages can have only negligible genetic consequences. Without fully attacking the problem to which modern theoreticians can only hazard provisional and highly varied solutions,2 it can therefore be seen that primitive mankind was not in a demographic position which would even have permitted him to ascertain the facts of the matter.

betrothed

But the analogy between incest and suicide is only apparent, for if society prohibits them both this prohibition applies in the first case to a natural phenomenon found commonly among animals, and in the second, to a phenomenon which is completely foreign to animal life and which should be regarded as a function of social life. Society expressly forbids only that which society brings about. Next, and in particular, society condemns suicide because it considers it harmful to its interests, and not because it constitutes the denial of a congenital tendency. A better proof is that, while every society prohibits incest, there is none which does not make room for suicide and does not recognize it as legitimate in certain circumstances or for certain motives when the individual attitude happens to coincide with some social interest. Accordingly, the reasons why incest is prejudicial to the social order still remain to be discovered.

The Aleutian does not copulate with his wife during her menstrual periods for fear of bad hunting, but if a father sees his daughter during her first menstrual period, she risks becoming blind and mute. The dangers are all for her, not for him.2 As a rule, a woman is impure during her menses, not only for her clan relatives, but also for her exogamous husband, and for everyone in general

Durkheim does not propose any law which might account for the necessary transition in the human mind from the belief in totemic substantiality to the horror of blood, from the horror of blood to the superstitious fear of women, and from this fear to the setting up of exogamous rules. The same criticism can be levelled at Lord Raglan’s imaginary reconstructions. However, we have shown that there is nothing more arbitrary than this succession of transitions. Even if there were only these transitions at the origin of the incest prohibition, they would still have permitted other solutions, some of which at least should have eventuated, by the simple law of probabilities. For example, the prohibitions affecting women during their menstrual periods provided a very happy answer to the problem, and a number of societies could have been satisfied with it.

III. The Universe of Rules

Hence, the prohibition is not always expressed in terms of degrees of real kinship but refers to individuals who use certain terms in addressing one another. This remains true even of those Oceanic systems which permit marriage with a classificatory ‘sister’, but distinguish immediately between kave maori, or ‘real sister’, and kave kesekese, ‘different sister’, kave fakata-fatafa, ‘sister set aside’, kave i take ƞaeƞa, ‘sister from another place’,1 It is the social relationship more than the biological tie implied by the terms ‘father’, ‘mother’, ‘son’, ‘daughter’, ‘brother’, and ‘sister’, that acts as the determinant. For this reason especially, theories attempting to justify the prohibition of incest by the harmful consequences of consanguineous unions (including numerous primitive myths suggesting this interpretation) can only be regarded as rationalizations.

Purity of soul, in the Vienna School sense, is never a factor in what might more readily be called a form of abortive polygamy rather than monogamy, for, in these societies as well as in those which favourably sanction polygamous unions, and in our own, the tendency is towards a multiplicity of wives. It was earlier indicated that the contradictory nature of the information about the sexual habits of the great apes does not allow any resolution, on the animal plane, of the problem of whether polygamous tendencies are innate or acquired. Social and biological observation combine to suggest that, in man, these tendencies are natural and universal, and that only limitations born of the environment and culture are responsible for their suppression.2 Consequently, to our eyes, monogamy is not a positive institution, but merely incorporates the limit of polygamy in societies where, for highly varied reasons, economic and sexual competition reaches an acute form. The very small degree of social unity in the most primitive societies accounts very well for these particular characteristics

unts very well for these particular characteristics.
Even in these societies, moreover, monogamy is not a general rule. The Nambikwara, semi-nomads of western Brazil, who live for most of the year by collecting and gathering, sanction polygamy for their headmen and sorcerers, The securing of two, three or four wives by one or two important persons in a band of sometimes less than twenty people necessarily obliges their companions to be celibate. This privilege by itself is sufficient to upset the natural equilibrium of the sexes, since male adolescents occasionally can no longer find wives available from among the women of their own generation. Whatever the solution given to the problem – homosexuality among the Nambikwara, fraternal polyandry among their northern neighbours, the Tupí-Cawahib – the growing scarcity of wives does not appear less serious a problem in a society however predominately monogamous it might be.1 But even in a strictly monogamous society, the considerations of the previous paragraph still retain their validity. This deep polygamous tendency, which exists among all men, always makes the number of available women seem insufficient. Let us add that, even if there were as many women as men, these women would not all be equally desirable – giving this term a broader meaning than its usual erotic connotation – and that, by definition (as Hume has judiciously remarked in a celebrated essay2), the most desirable women must form, a minority. Hence, the demand for women is in actual fact, or to all intents and purposes, always in a state of disequilibrium and tension

Homosexuality in some groups, polyandry and wife-lending in others, and finally, almost everywhere, the extreme freedom of premarital relations would prevent adolescents from experiencing any discomfort while waiting for a wife, if the wife’s function were limited to sexual gratification. But, as often noted, in most primitive societies (and also, but to a lesser extent, in the rural classes of our own society) marriage is of an entirely different importance, not erotic, but economic. In our society, the difference between the economic status of the married man and the unmarried man amounts almost solely to the fact that the bachelor has to replace his wardrobe more frequently. The situation is altogether different in groups where the satisfaction of economic needs rests wholly on the conjugal society and the division of labour between the sexes. Not only do man and wife have different technical specializations, one depending on the other for the manufacture of objects necessary for their daily tasks, but they are each employed in producing different foodstuffs. Accordingly, a complete, and above all regular, food supply indeed depends on that ‘production cooperative’, the household. The Pygmies, who consider women and children as the most valuable of the active part of the family group, say that ‘the more women available the more food’.1 Likewise, Hottentot women, during the marriage ceremony, chorus the praise of the groom and of the men who, like him, are looking for a wife, ‘since today they have enough to eat’.

IV. Endogamy and Exogamy

Marriage is an eternal triangle, not just in vaudeville sketches, but at all times, and in all places, and by definition. Because women have an essential value in group life, the group necessarily intervenes in every marriage. It does this in two ways: firstly in the form of the ‘rival’, who, through the agency of the group, asserts that he had the same right of access as the husband, a right upon which the union is conditional and which must be shown to have been respected; and secondly through the group as a group, which asserts that the relationship which makes the marriage possible must be social, that is, defined in group terms and not in the natural terms having all the consequences incompatible with collective life which have already been indicated. Considered in its purely formal aspect, the prohibition of incest is thus only the group’s assertion that where relationships between the sexes are concerned, a person cannot do just what he pleases. The positive aspect of the prohibition is to initiate organization.

Is it not true that in certain areas of Australia and Melanesia this prohibition is adapted to a virtual monopoly of the women by the old men, and, more generally, to polygamy, the results of which we ourselves have emphasized?

The chief receives several wives from the group. In exchange, he gives a guarantee against want and danger, certainly not to the particular individuals whose sisters or daughters he has married, and not even to those perhaps condemned for ever to celibacy by the exercise of his polygamous right, but to the group as a group, for it is the latter which has suspended the common law in his favour.

In point of fact, marriage rules do not always merely prohibit a kinship circle, but occasionally also fix one within which marriage must necessarily take place, under pain of the same scandal as would result if the prohibition itself were violated. There are two cases to be distinguished here: on the one hand endogamy, or the obligation to marry within an objectively defined group; and on the other, preferential union, or the obligation to choose as spouse an individual who is related to Ego in some particular way. It is difficult to make this distinction in classificatory kinship systems, for all individuals in a defined kinship relationship to each other, or to a given subject, fall into one class, and consequently it is possible to pass from preferential union to endogamy, properly so called, without any marked change. Any system of marriage between cross-cousins could thus be interpreted as endogamous if all the parallel cousins were designated by one term, and all the cross-cousins by another. This double appellation could hold good even after the disappearance of the marriage system considered, and consequently an exogamous system par excellence would make way for a new system, which, on the contrary, would offer all the appearances of endogamy. This artificial conversion of genuine exogamous systems into ostensibly endogamous systems may be seen in the field, and the difficulties it raises in the interpretation of certain Australian systems will be seen later.

If a girl cannot find a partner possessing the true faith, it is better for her to marry her father, for it is the possession of this faith which is the prime essential in their definition of a human being

V. The Principle of Reciprocity

Amundsen’s misadventures show the cost of losing the meaning of reciprocity. From the generous gifts that Amundsen gave the Eskimo, in return for their presents, they quickly concluded that it was to their advantage to offer all their goods as presents. It soon became necessary to decline any present and to resort to proper commerce.4 Likewise, Holm states that an exchange with one native opens a general claim on the part of all the others to the same gift: ‘The natives explained that they always gave people everything they asked for.

Through the uselessness of the gifts, and their frequent duplication because of the limited range of objects suitable as presents, these exchanges also take the form of a vast and collective destruction of wealth. Without calling upon the very significant theme in modern folklore of the millionaire lighting his cigar with bank notes, there are many little facts in this example to remind us that even in our own society the destruction of wealth is a way to gain prestige. The skilful merchant knows that a way to attract customers is to advertise that certain high-priced articles must be ‘sacrificed’. The motive is economic, but the terminology retains an air of mystery.

Doubtless gambling provides, in modern society, the most striking picture of these transfers of wealth with the sole purpose of gaining prestige. Gambling really requires a special study by itself, but here we shall confine ourselves to a brief statement. During the last hundred years, gambling has shown exceptional development each time the means of payment were found considerably to exceed the local availability of commodities. The fabulous gambling stories of the Klondyke of Alaska, during the mining expansion, are echoed by those of the Amazon region during the great rubber period. Thus it seems as if money, which we have become accustomed to regard as a simple means of obtaining economic goods, found, when it could not be used up in this way, another archaic rôle, formerly attributed to precious things, viz., as a means of gaining prestige by the value of the gift or sacrifice, which has actually been made or simply mooted.

everyone lends and borrows without too much concern for restitution

Thus the refinements of sharing or distribution appear with the urgency or the absence of the need.

But there is still a general model here. In the significant field of the offering of food, of which banquets, teas and evening parties are modern examples, the language itself, as in ‘to give a reception’, shows that among ourselves, as in Alaska or Oceania, ‘to receive’ is to give.

One ‘offers’ dinner to a person whom one wishes to honour, and this type of invitation is the most frequent way of ‘returning’ a kindness. The more the social aspect takes precedence over the strictly alimentary, the more stylized also is the type and presentation of the food offered; the fine porcelain service, the silverware, and the embroidered table-cloths, which are usually carefully stored away in the family cupboards and buffets, are a striking counterpart of the ceremonial bowls and spoons of Alaska, brought out on similar occasions from painted and decorated chests. The attitudes towards food are especially revealing.

Food serves the body’s needs and wine its taste for luxury, the first serving to nourish, the second, to honour.

wine is a social commodity

What has happened? The two bottles are identical in volume, and their contents similar in quality. Each person in this revealing scene has, in the final analysis, received no more than if he had consumed his own wine. From an economic viewpoint, no one has gained and no one has lost. But the point is that there is much more in the exchange itself than in the things exchanged.

When social distance is maintained, even if it is not accompanied by any sign of disdain, insolence or aggression, it is in itself a matter of sufferance in that any social contact entails an appeal, an appeal which is a hope for response. This is the fleeting but difficult situation resolved by the exchanging of wine. It is an assertion of good grace which does away with the mutual uncertainty. It substitutes a social relationship-for spatial juxtaposition. But it is also more than that. The partner who was entitled to maintain his reserve is persuaded to give it up. Wine offered calls for wine returned, cordiality requires cordiality. The relationship of indifference can never be restored once it has been ended by one of the table-companions. From now on the relationship can only be cordial or hostile. There is no way of refusing the neighbour’s offer of his glass of wine without being insulting. Further, the acceptance of this offer sanctions another offer, for conversation. In this way a whole range of trivial social ties are established by a series of alternating oscillations, in which offering gives one a right, and receiving makes one obligated, and always beyond what has been given or accepted.

We are faced then with a ‘total social fact’ – on a microscopic scale, it is true – the implications of which are at once social, psychological and economic. This apparently futile drama, which perhaps the reader will think has been given a disproportionate importance, seems on the contrary to offer material for inexhaustible sociological reflection

Perhaps it will be added that there is only one common characteristic between the prohibition of incest and the reciprocal gift, viz., the individual repulsion and social reprobation directed against the unilateral consumption of certain goods, but that the essential characteristic of reciprocal gifts, i.e., the positive aspect of reciprocity, is entirely missing in the first case, so that strictly speaking our interpretation could be valid only for exogamous systems (and particularly dual organizations) which present this reciprocal characteristic, but not for the prohibition of incest as practised in our society.

The marriage itself is an inherent part of the prestations which accompany it. It forms merely the central motive

wife-buying

But marriage by purchase is a special institution only in its form. In reality, it is only a modality of that basic system analysed by Mauss, according to which, in primitive society and still partially in our own, rights, goods and persons circulate within the group according to a continual mechanism of prestations and counter-prestations.

the total character, sexual, economic, legal and social, of this collection of reciprocal prestations which make up marriage

Will you take my gifts or not?’ The answer being, perhaps: ‘I will take them’, or ‘I have taken the gifts of another man. I don’t want to exchange with you.’ Even the wording of these overtures is fixed by tradition. This exchange of gifts initiates a whole series of reciprocal prestations which lead to marriage, or, rather, constitute the initial transactions of marriage, viz., work in the fields, meals, cakes, and so on

VI. Dual Organization

migratory movements

viz

VII. The Archaic Illusion

Nevertheless, we should pause and reflect for a moment on our premises.

To assert, as we did in the last chapter, that a historical or geographical study could not exhaust the problem of the origin of dual organizations, and that for a better understanding of these organizations we must take into consideration certain fundamental structures of the human mind, would be a meaningless proposition if we were unable to perceive exactly how these structures were made up and what the method was by which we might apprehend and analyse them.

What are the mental structures to which we have referred and the universality of which we believe can be established? It seems there are three: the exigency of the rule as a rule; the notion of reciprocity regarded as the most immediate form of integrating the opposition between the self and others; and finally, the synthetic nature of the gift, i.e., that the agreed transfer of a valuable from one individual to another makes these individuals into partners, and adds a new quality to the valuable transferred.

She begins by emphasizing ‘the strength and urgency of the common wish of little children to have exclusive possession or at least the biggest share or main use of whatever properties are the centre of interest at the moment. The satisfaction of having things all one’s own is deep, the chagrin at others having more than oneself very bitter.’1 This attitude is felt not only for material objects, but also for immaterial rights, such as the hearing or singing of a song. Furthermore, ‘taking turns’ is one of the hardest lessons for children under five years to learn. All that the child ‘knows is that the others “have got it”, and he hasn’t. A few minutes is an eternity when one is eagerly waiting’

Hence, ‘the making and receiving of gifts remains the clearest, most unequivocal sign of love… In a thousand ways, every patient shows that the deepest layer of meaning of “being loved” is still receiving a gift; and the primal meaning of being hated is being deprived, being robbed, since to the infant this means being destroyed.’9 Beyond the intrinsic value of the thing given, there is the gift itself as a sign of love, and beyond that, the gift as a sign of the fact of being worthy of love

He who is denied feels that he has been denied because he is bad, because he is or has been hostile to the giver. This it is which brings poignancy to the child’s gratitude for gifts, and bitterness to his sense of loss when he is left out of the giving. The gift is not only a sign that the giver loves and does not hate; it is also a sign that the recipient is believed to be loving, not hating and hateful.’10
This explains the child’s desire to make enormous and magnificent gifts, ‘a polar bear … a real one’, ‘a big large engine’. It is essentially a wish to be potent in giving: ‘If one has “big large engines” to give away, one is indeed both safe and good. One is no longer the helpless puling infant, dependent upon the gifts of others, and driven by helpless anxieties to rage and jealousy… It is more blessed to give than to receive, because to be able to give is not to need.’1
These feelings apply to services, too: ‘How warmly the children appreciated services done them … Tommy … calls out to the children who are jointly carrying a large plant pot “Don’t go too quickly, so that I can help”.’ Such is the intensity of the ‘great pleasure from being able (i.e., powerful enough) to be unselfish’

Every field-worker who has had concrete experience of primitive children will undoubtedly agree that the opposite is more likely to be true and that in many regards the primitive child appears far more mature and positive than a child in our own society, and is to be compared more with a civilized adult. But this is not the question.
Once the distinction between the child and adult has been propounded – and, as we have seen, it must not be overestimated – what is the basic relationship to be established between their respective mental manifestations? Adult thinking is built around a certain number of structures which it specifies, organizes, and develops from the single fact of this specialization, and which are only a fraction of the initial summary and undifferentiated structures in the child’s thought. In other words, the mental schemata of the adult diverge in accordance with the culture and period to which he belongs. However, they are all derived from a universal resource which is infinitely more rich than that of each particular culture. Every newborn child provides in embryonic form the sum total of possibilities, but each culture and period of history will retain and develop only a chosen few of them. Every newborn child comes equipped, in the form of adumbrated mental structures, with all the means ever available to mankind to define its relations to the world in general and its relations to others.

In comparison with adult thought, which has chosen and rejected as the group has required, child thought is a sort of universal substratum the crystallizations of which have not yet occurred, and in which communication is still possible between incompletely solidified forms.

If we have recalled these facts it is because learning a language presents the same problems as the infant’s first steps into social life, and because these problems have received the same solution. The variety of sounds that the speech organs can articulate is almost unlimited. However, each language retains only a very small number of all the possible sounds. During the prattling period, before the introduction to articulated language, the child produces the total range of sound realizable in human language while his own particular language will retain only some of them. In the first few months of its life every child has been able to emit sounds which later he will find very difficult to reproduce, and which he will fail to imitate satisfactorily when he learns languages very different from his own.1 Every language makes a selection and from one viewpoint this selection is regressive. Once this selection is made, the unlimited possibilities available on the phonetic plane are irremediably lost. On the other hand, prattling is meaningless, while language allows people to communicate with one another, and so utterance is inversely proportional to significance.

prattling

For this reason we should speak more readily of the ‘polymorphism’ of child thought, using this term as does the psychoanalyst when he describes a child as a ‘polymorphous pervert’. But what is really meant by this? It means that the child presents in a rudimentary form, and co-existently, all the types of eroticism among which the adult will seek his specialization on the normal or pathological plane. Considering the relationship between the social attitudes of the child and the different types of organization in human societies the anthropologist should be similarly inclined to say that the child is a ‘polymorphous socialite’.

panmorphism

In these circumstances, it is easy to understand why ethnologists, psychologists and psychiatrists have each been tempted to set up parallelisms between primitive, infantile and pathological thought. In so far as psycho-neurosis can be defined as the highest form of mental synthesis on a purely individual level,1 the sick person’s thought resembles the thought of the child. These forms of thought no longer conform, or do not yet conform, to the selective structure of the particular group to which they belong. Consequently, the child and the sick person are both relatively free to elaborate their own particular synthesis. This synthesis is probably doomed to remain unstable and precarious, because its realization is on an individual plane and is not within the framework of any social environment.

VIII. Alliance and Descent

unilineal

where the greatest insult is to tell a man ‘to eat the excreta of his sister’. It can be wiped out only by the blood of the offender. But if the offender belongs to the opposite moiety, it is the sister herself who must be killed, and the original offender must also kill his sister if he wishes to re-establish his reputation.2 This native evidence is perhaps based upon myths but it corresponds to very similar observations made by Warner among the Murngin

vis-à-vis

The exchange of gifts .taking place on the occasion of the periodic settlement of grievances between the groups… . is not a business transaction – not a mere bartering – but a means of expressing and cementing friendship

Custom is not inconsistent in this case any more than it is in any other. But no understanding of it will come through considering merely its visible content and empirical expression. Custom is only a superficial aspect of the system of relations, which is what must be isolated.

matrilineal society

In other words, the woman will go to live in her husband’s village, sometimes far from her own people, while she and her children will always be strangers within the group with which they are nevertheless associated. If a society is both matrilineal and matrilocal, either permanently or temporarily, as often happens, the husband belongs to the despised class of ‘those-resulting-from marriage’ or ‘strangers’, as opposed to ‘the owners of the village’. Consequently, he is always aware of the precariousness of his residential title as compared with his wife and children

patrilocal

matrilocal

IX. The Marriage of Cousins

The Marriage of Cousins

ut there is no need, in this theoretical case, to postulate any precise type of institution. If there is any real case corresponding to this theoretical one it is clearly that of primitive bands composed of biological families set closely side by side, or, on the contrary, without regular contacts, and still at a very elementary stage of organization. As a matter of fact, our interpretative diagram does not imply the existence of stable institutions, or the establishment of a particular rule of descent or residence. It merely implies that women are regarded as valuables – a psychological attitude sufficiently attested to in the great majority of primitive societies, and by the relationships between the sexes at the animal level – and the apprehension by the individual of reciprocal relationships of the type

X. Matrimonial Exchange

it is a general fact in Australia that a man cannot hope to obtain a wife if he has no sister, daughter or god-daughter to give in exchange. Brough Smyth draws a moving picture of the condition almost of despair to which the bachelor who fails to obtain a wife is reduced in spite of himself in Australian aboriginal society: ‘A man who has no female relations that can be exchanged for a young woman of another tribe leads an unhappy life. Not only must he attend to his own wants, and share the discomforts of the bachelors’ quarters, but he is an object of suspicion to the older men, who have perhaps two or three young wives to watch … There is the discontent and unrest of such a life, which makes him a dull companion, a quarrelsome friend, and a bitter enemy.’

This suggests, what all the evidence tends to confirm, that these various groups have been produced by the deliberate and repeated bisection of a community, first into two, then into four, and finally into eight exogamous and intermarrying groups or classes; for no one, so far as I know, has yet ventured to maintain that society is subject to a physical law, in virtue of which communities, like crystals, tend automatically and unconsciously to integrate or disintegrate, along rigid mathematical lines, into exactly symmetrical units

The exchange relationship comes before the things exchanged, and is independent of them. If the goods considered in isolation are identical, they cease to be so when assigned their proper place in the structure of reciprocity.

we have not studied cross-cousin marriage as a primitive, archaic, relatively ancient, or recent expression of this form, but as a special case showing particularly clearly that reciprocity is present behind all marriages. We have given proof of this characteristic of cross-cousin marriage by adopting a double viewpoint: firstly by showing that the dichotomy of cousins into assigned and prohibited spouses may be directly deduced from the relationship between two or more families, once this relationship is conceived of as a structure of reciprocity; and secondly by emphasizing that through its logical qualities, the institution of cross-cousin marriage occupies an exceptional position, placed as it were at the bifurcation leading to two extreme types of reciprocity, viz., dual organization and the prohibition of incest.

XI. The Classical Systems

Systems with two matrilineal exogamous moieties
2.  systems with two patrilineal exogamous moieties

four-section systems:
(a)  with named matrilineal moieties
(b)  with named patrilineal moieties
(c)  without named moieties
4.  eight-subsection systems
5.  systems with four named patrilineal semi-moieties
6.  systems with two endogamous alternating divisions, and
7.  systems without named divisions

I.  Systems without classes:
(a)  unilateral marriage without the exchange of sisters
(b)  systems without classes or matrilineal dichotomy
(c)  systems with matrilineal clans
(d)  systems without matrilineal clans, and
(e)  systems with alternating generations.
II.  Systems with classes:
A.  Eight-class systems:
(a)  eight named subsections
(b)  four named sections divided into eight unnamed subsections, and
(c)  four patrilineal semi-moieties divided into eight unnamed subsections.
B.  Four-class systems:
(a)  four sections, and
(b)  two patrilineal moieties divided into four unnamed sections.
C.  Two-class systems:
(a)  two matrilineal moieties.2

Radcliffe-Brown proposed the use of more specialized terms, viz., ‘moieties’ when there were only two divisions in the group; ‘sections’ when there were four; and finally ‘subsections’ when there were eight.

In this work we have attempted to define two methods for determining the spouse, viz., the method of classes and the method of relations.1 We then showed that these two methods never correspond exactly and that even in the simplest system, such as exogamous moieties, inter-individual relationships must always be taken into consideration. None the less, the very fact that classes do exist proves, it seems to us, that they are not entirely without use and that there should be at least a certain degree of equivalence between the two systems. Not all the members of the class are possible spouses, and even between possible and preferred spouses there are differences which can only be explained in terms of consanguinity, whereas such differences are non-apparent in terms of class. Nevertheless, no possible spouse is to be found outside the class, and this alone indicates that from the point of view of marriage rules the class does have a function. This function is simple and wide when the number of classes is small, but gains in refinement and precision, without ever becoming perfect (since the number of kinsmen is theoretically unlimited, while a system with too many classes would be useless because of its complication), when this number increases. But the function is the same in all cases, viz., to act as a ‘sorting-office’ in the determination of the spouse. Seen in this way the Kariera system (including the division into four sections and the kinship system) might be said to be a particularly satisfactory logical model since the first relatives selected by the method of classes are also the preferred spouses according to the method of relations. This, incidentally, seems to be the reason why Radcliffe-Brown made the Kariera system the prototype of his kinship system and called it type I.

The relationship of consanguinity means more than class in the determination of the spouse, but the natives regard the violation of class exogamy with the same horror as they regard marriage with a relative of a forbidden consanguineous relationship. If the basis of the problem is to be found in the consideration of degrees of consanguinity, rather than in classes, the function of the latter is not substantially different, even though the relation between them still remains to be specified.

XII. The Murngin System

the Murngin system is not a direct codification of the marriage rule characteristic of this society, hut rather that it is the result of a sort of compromise between a pre-existing marriage rule and a fully developed class system introduced from outside.

While the Kariera system employs twenty-one different kinship terms, the Aranda employs forty-one, and the Murngin system seventy-one. How is it that the Murngin system, which brings about an intermediate dichotomy between that of the Kariera and the Aranda systems, requires almost double the number of kinship terms as that of the more complex system?

The second criticism brought against me – that I mistakenly postulated the circularity of the system – derives from a misunderstanding between what is meant by model and what is meant by empirical reality. The model of a generalized system necessarily implies a certain circularity although this may be simple or complex, and may assume various forms. But the empirical reality is far more flexible. Of all the empirically observable cycles of alliances, a certain proportion will be found to be circular, either short term (three are absolutely necessary, at least four among the Murngin because of the division into moieties) or long term; and others, which never ‘join up’ because they are ‘lost’.

XIII. Harmonic and Disharmonic Regimes

The subdivision of each generation into two age classes, ‘older’ and ‘younger’, is thus in direct correlation with the possibility of two men vying for the one woman. This competition is avoided by limiting their respective claims to two different age classes, the members of which are either real or classificatory parallel cousins. We shall have frequent cause to show that this dichotomy of generations always appears in such circumstances, and that it must be seen as a normal function of systems of alternate marriage. In fact, among the Cape York tribes, there is not one alternate marriage, but two. In association with matrilateral marriage there is, in a weakened form it is true (since in the second case the cousin can only be classificatory), a patrilateral type of marriage.

In the Manchu system, a man can marry a woman belonging to a generation senior to his own, or to the older branch of his own generation, while women from the younger class of his generation, or members of younger generations, are prohibited.4 The ‘age spiral’ of the Wikmunkan has its symmetrical but inverse model among the Manchu. Our interpretation of the Manchu system as resulting from a change from generalized to restricted exchange will thus confirm similar interpretations given by us for so-called ‘aberrant’ Australian systems (which are aberrant only because of an incomplete classification).

But harmonic regimes do exist, and short of producing indefinitely an increasing number of reducible societies,1 they have to be based on patrilateral or matrilateral marriage, but not on bilateral marriage, which is at the immediate disposal of disharmonic regimes alone, A harmonic regime can thus choose between two systems of marriage, patrilateral or matrilateral. Each of these two systems is itself compatible with two modes of descent, patrilineal or matrilineal

That success should have crowned the bold hypothesis of the great English sociologist may justifiably encourage all those who believe that an internal logic directs the unconscious workings of the human mind, even in those of its creations which have long been considered the most arbitrary, and that the appropriate methods to be applied to it are those usually reserved for the study of the physical world. At the same time, however, a doubt intrudes itself, certainly not concerning the reality of the Kariera system, but about its exclusive existence in the enormous territory which has been assigned to it.2

In any case, the one fact remains. A formal study of the notion of exchange, such as sociologists have so far employed, has shown us that it did not succeed in embracing the facts in their integrity. Rather than deciding to lend a sterile discontinuity to phenomena which are, after all, of the same type, we have preferred to seek a wider and modified conception of exchange in an attempt to arrive at a systematic typology and an exhaustive explanation. It is the Australian facts, i.e., those taken from the classic region of restricted exchange, which have forced us to develop the notion of exchange and have, as it were, imposed upon us the notion of generalized exchange. What is the connexion between restricted and generalized exchange? Are they to be seen as two independent forms, yet capable of interacting one upon the other when the chances of culture contact bring them together, or do they represent two related stages in one process of evolution? In so far as the concern is with the solution of regional problems, this is a problem for the ethnographer and cultural historian. Our own intention is limited to making a structural study of the two types and their interrelations, and we must now attempt to isolate, in a simple and directly observable form, the formula of generalized exchange, the theoretical necessity for which has become apparent even before we have succeeded in discovering it in the facts.

XIV. Appendix to Part One

In these few pages, written at Lévi-Strauss’s request, I propose to show how a certain type of marriage laws can be interpreted algebraically, and how algebra and the theory of groups of substitutions can facilitate its study and classification.
In the societies in question, the individual men and women are divided into classes, each person’s class being determined, according to certain rules, by those of his parents. According to the respective classes of a man and of a woman, the marriage rules indicate whether they can marry or not.

This mathematical examination of the Murngin system calls for a number of comments. Firstly, the discovery that a Murngin type system, if it functions under the strict conditions which alone allow it to be interpreted systematically, gives rise to the fission of the group into two irreducible societies, shows that a system of generalized exchange cannot evolve beyond its own formula. Doubtless, the system can be conceived of with any number of classes, but whether there be 3, n, or n + 1, the structure remains unchanged. If an attempt is made to change the structure there are two possibilities open: either the change is made, and the formula for generalized exchange comes to an end (as with the Dieri system), or the formula is preserved, and then it is the change of structure which proves illusory. In acquiring eight subsections, the Murngin system succeeds only in functioning under the theoretical conditions for two juxtaposed systems of generalized exchange, each with four classes. We came to this conclusion in the previous chapter, using a purely structural analysis, and it is confirmed by the mathematical analysis. The extent to which Elkin was mistaken can be seen when, letting himself be misled by a Malinowski-inspired empiricism, he wrote: ‘And generally speaking, the study of the merely formal element in Australian kinship is hardly worth doing … After all, it is of little satisfaction and affords no real understanding of the life of the tribe

In any case, there is one tribe in which we know that a system of generalized exchange in the course of evolving tends to bring about the subdivision of the group into sub-societies. This is the Apinayé of Central Brazil. It may be remembered that the tribe is divided into four kiyé, i.e., four ‘sides’ or ‘parties’. The rule of marriage is typical of generalized exchange: an A man marries a B woman, a B man a C woman, a C man a D woman, and a D man an A woman. Furthermore, boys follow their father’s kiyé and girls their mother’s kiyé, i.e., kiyé A comprises, on the one hand, the sons of A men and of B women, and on the other, the daughters of D men and of A women. In these circumstances, as we have already noted, all A men, and all B women, originate from one type of marriage (between A men and B women) which they are obliged to perpetuate. Although cousins of the first degree are prohibited spouses among the Apinayé (which may be considered as a partial defence against the consequences of the system), it is none the less true that the kiyé, which seem exogamous, really function as endogamous units.1 It is probable that a careful examination of cases of the same type, which must be more numerous than a too exclusive attention to the ‘exogamous’ side of the phenomenon would allow us to suppose, would provide a useful method of approach for the study of systems of generalized exchange.

XV. The Givers of Wives

If, on the other hand, a nephew refuses to marry his cousin, he must pay a fine to the offended uncle. If the wife dies after marriage, her husband has the right to demand another woman from his parents-in-law (uncle and aunt), who give him a daughter or a kinswoman as wife. Moreover, in the case of adultery the seducer’s clan is jointly responsible for the fine payable to the offended husband.1
That one never marries an individual is brought out clearly in the customary form of the marriage proposal.

XVI. Exchange and Purchase

The first thing one is struck by in the Kachin marriage rule is its simplicity. It seems that the declaration of preferential union with the mother’s brother’s daughter is enough for the formation of a subtle and harmonious cycle, within which the great as well as the smallest social units automatically find their place, and, without compromising their general agreement, may also improvise more limited developments, such as the ternary cycles of feudal families, which fit so easily within the quinary cycle which embraces all the groups. Thus the simple formula of the division into mayu ni and dama ni seems rich in possibilities, and it suffices to bring a balanced order into a complex reality. To all intents and purposes this formula dominates the whole social life, and what Sternberg said of the identical formula governing Gilyak institutions might equally be said of it, viz., that the origin of these institutions must lie in some unique and simple principle, some categorical imperative acceptable to the native mind, and the seed from which the complex organization of native institutions has grown

‘It were very shameful and very serious if a young man had intercourse with a girl of his own family or of his dama tribe, even were she only a distant relative of his. According to the old people, those cases are very rare and always severely punished because such a union … can never become legitimate. But a dama boy may without dishonour, if not without danger, have intercourse with a mayu girl …’

Every marriage takes more or less the form of a sale in which the price of the wife varies according to her rank.’1 The importance given to this aspect of the problem in native thought is expressed in the songs of rejoicing at the birth of a child: ‘Let him grow up. Let him become father of many children’, the father exclaims if it is a boy; and if it is a girl: ‘Let her grow up. May she be able to be given in marriage, and bring to her family buffaloes, gongs, brandy, clothes, etc..’2 Gilhodes makes the following comment on this: ‘Kachins earnestly desire to have children: boys to continue and propagate the family, girls to draw profit from them especially at their marriage …’

Marriage is … a socially regulated act of hostility.’2 We have seen that the dama ni is all-powerful in exacting a wife, and any other spouse who may be substituted if the first one named is not suitable. On the other hand, however, the mayu ni jealously watch over their son-in-law. During the first year at least (even if the many payments have been made regularly), the young husband must help his parents-in-law ‘to prepare the rice-fields, and at another time, to rebuild their house’.3 In case of divorce, when the husband is to blame, he must give his wife a dah (sword) and a buffalo which is sacrificed by his parents-in-law to celebrate their daughter’s return, and if the divorce is the wife’s fault her family must return all the presents and give a buffalo which is sacrificed by the husband as evidence of the return of the presents. However, says Gilhodes, in the case of adultery it is always the man who is responsible and he it is who pays the fines, sumrai kha.4 He adds: ‘The punishments that follow adultery are enormous considering the poverty of the Kachins, and reduce to rapid destitution, the family of the guilty man.

The oldest brother, being married and having left the hmunpi, loses all his rights. In the case of five unmarried brothers, the order of succession is: first, the youngest; in his absence the oldest; otherwise, one of the other three, from the youngest to the second oldest, in order of birth. But if four of the brothers are married, the order changes: first, the youngest (the only bachelor), next, the oldest, and if not, in descending order to the second youngest. When the inheritance is large, the oldest brother recalls all the marriage prestations still due, pays the price of his own marriage and of all his younger brothers’ marriages, and if anything is left he takes two-thirds and gives one-third to his youngest brother. The order of births is equally important for the sharing of the price paid for the sisters. If there are four brothers and three sisters, the oldest and the youngest have the right to the first two prices of the brother (ta man), and the intermediate brother to the third. If there are three brothers and five sisters, the oldest and the youngest each take two ta man, and the intermediate brother only one. It can be seen that the privileged position of the youngest is clearly a function of the system of matrimonial prestations, and that the regulation of the right of inheritance described by Head, involving a detailed differentiation of the beneficiaries, is intimately connected with marriage.

Generalized exchange can provide a formula of organization of an exceptional clarity and richness, a formula which can be widened indefinitely and can express the needs of as complex a social group as may be imagined; its theoretical law can function uninterruptedly and without fail.2 The dangers which threaten it come from outside, from concrete characteristics, and not from the formal structure of the group. Marriage by purchase, by substituting itself, then provides a new formula which, while safeguarding the principle of the formal structure, furnishes the means of integrating those irrational factors which arise from chance and from history, factors which the evolution of human society shows to follow – rather than precede – the logical structures which are elaborated by unconscious thought, access to which is often more easily gained through very primitive forms of organization.

XVII EXTERNAL LIMITS OF GENERALIZED EXCHANGE

It is the girl who makes the marriage proposal, except when she marries her father’s sister’s son, when it is an automatic procedure. When there is no patrilateral cross-cousin, she must marry a man in a similar position in the paternal lineage. There is no bride-price.

Mills reports that marriage with the mother’s sister’s daughter, the sister’s daughter, and the father’s sister’s daughter is always prohibited even though the latter do not belong to the same phratry as Ego. On the other hand, marriage is possible with the mother’s brother’s daughter, and she or a woman from the mother’s clan is the recommended spouse. This type of marriage is not obligatory, but to act otherwise incurs the risk of offending the maternal clan. If a man whose first wife belonged to the mother’s clan remarries into another clan, he is obliged to pay a fine called lolang ’ntyakma, ‘price for not taking from the mother’s clan’.3 On the other hand, remarriage with the father’s widow, viewed with disfavour by the Lhota, is approved by the Sema

In this work we have continually interpreted the distinction between older and younger in the one generation as an indication, or a vestige, of an alternate system of marriage (i.e., in which two classes of men can compete for the same class of women), the solution to which was sought in the complementary allocation of the elder and of the younger. Accordingly, the force of this differentiation among the Lhota is an added indication that two types of marriage clearly correspond to the heterogeneity and complexity of the social structure. Two spouses are always possible, one corresponding to a formula of restricted exchange, the other to a formula of generalized exchange.

and brothers-in-law, A quarrel with an elder blood relation such as father, mother, uncle, aunt, elder brother, elder sister and so on is a serious thing, and is believed to entail illness, poor crops and other evil fortune,’1 i.e., the precise consequences which other Nagas attach to quarrels with the maternal uncle.

XVIII. Internal Limits of Generalized Exchange

The Gilyak are divided into patrilineal, patrilocal and exogamous clans. They practise marriage by purchase, the bride-price being paid to the bride’s father, or to her brothers

even marriage by purchase is a function of marriage with the matrilateral cross-cousin. It appears only when the relationship of the future spouses does not conform to the preferred type, the only one which is ‘absolutely pure and sacred’. Sternberg makes the following comment; ‘As soon as a son is born to her, the mother’s first care is to do all in her power to bring about his betrothal with the daughter of one of her brothers.’5 If she succeeds, there will be no kalim, or bride-price. On the contrary, the future wife’s father will give gifts to the bridegroom, given the extraordinary privileges a man has over his mother’s brother.6 As it happens, the bride-price is very high, and poor people cannot hope to obtain a wife outside the institution of preferential marriage. Moreover, sexual relations are sanctioned between a man and his unmarried angey, and although the husband’s rights must be respected with regard to a married angey, fraternal polyandry is practised by common assent. On the other hand, if there is a prohibition on sexual relations with a stranger, yet ‘from one end of the region to the other, wheresoever a man may find his angey, he finds, at the same time, a conjugal home. All he needs is to know the kinship relationships,’7 Legends strongly suggest that there was no marriage ceremony, this being easily understood since the prescribed degree is the necessary and sufficient condition for marriage.

in a system of generalized exchange, A is debtor solely to B from whom he receives his wives and, for the same reason, B is debtor solely to C, C to n, and n to A, when there is a marriage it is as if B had also a direct claim over n, C over A, n over B, and A over C.

There is certainly no necessity, or even excuse, for treating these peculiarities as matrilineal survivals. But do they not express a sort of female reaction to, or viewpoint on, the system taken as a whole? These patrilineal and patrilocal regimes, which condemn women to the hard fate of exile, theoretically for ever, in foreign households, often different in language and customs, do not exclude, however, a certain solidarity in the female line; perhaps they are even the cause of it.

Among the latter, it took three years for a husband to receive his first smile from his wife. Among the Barbarians, the newly wed woman (who enjoys complete sexual freedom) leads the life of a young girl for three years. We are told that her brothers at the time show themselves to be extremely jealous. In China, in songs about the unhappily married woman, the wife who cannot pride herself on getting on as well with her husband as with a brother, makes the curious exclamation: “My brothers shall not find out: they would laugh and mock me.”

Actually, the Gilyak are so deeply convinced that their marriage rules are based on nature that they even apply them to their domestic animal, the dog. They believe that among dogs as well, brothers and sisters cannot copulate. The rare exceptions which occur are attributed to the influence of an evil spirit, milk, and consequently, if a Gilyak chances to witness an incestuous act between dogs, he must kill the animal, for fear of committing the same offence himself. The execution of the dog is a religious ceremony. It is strangled and its blood is cast to the four corners of the earth. For the dog, however, this ceremony is only in punishment of the incest. It is curious to note that the Gilyak tolerate incest between dogs of consecutive generations, between mother and son, etc

XX. The Chao Mu Order

Warner and Davis’s conceptual system cannot help us grasp the structures of kinship any more than traditional grammar can perfectly express the operations of language, or Aristotelian logic permit the real functioning of thought to be understood.2 However, the fact remains that, like Aristotle or the grammarians, a people (or, more likely, a scholarly élite) have undertaken the same task, in a new field, of putting a certain type of data into analytical form. The very perfection and artificiality of the Chinese system are sufficient indication that it is not, and cannot be, the result of a spontaneous and unconscious evolution. Everything about it suggests that it is a promulgated system. It has been fabricated, and fabricated with a certain end in view.

XXI. Matrilineal Marriage

teknonymy

The term teknonymy was invented by Tylor to describe the custom whereby a person is called the father, mother, grandfather or grandmother, etc., of one of his descendants instead of by his own name. Lowie accepts this meaning in his treatment of the institution,1 and he interprets it as resulting from a terminological deficiency, viz., either the language has no specific term for the relative in question, or else the term cannot be used, provisionally or definitively, for reasons of etiquette.

XXIII. Peripheral Systems

Peripheral Systems

XXIV. Bone and Flesh

Bone and Flesh

It is only in a system of generalized exchange that marriage can assume, for the uninstructed observer, the superficial appearance of a gift. In a system of restricted exchange, the immediately perceivable aspect is that of the exchange of a daughter against a daughter, or brother’s sister against son’s wife. In a system of generalized exchange, by contrast, the transfer is never directly reciprocal. It seems that the ‘parents-in-law’ group give a woman – and give her without receiving anything in exchange – at least on the part of the ‘sons-in-law’ group. The operation thus has all the appearances of a gift, and of a non-solicited gift, since the latter is founded on a pre-established order. In this connexion, Kachin practice can be compared with Hindu theory, viz., the mayu ni who are sounded out react ill-humouredly, make a show of refusing, and make extravagant demands. But when they have a girl to marry, they are the ones who take the initiative, and appeal to possible husbands, by hanging out women’s clothes.1 Among the Gilyak, orthodox marriage (based on generalized exchange) is a marriage by right, to which is opposed negotiated marriage based on purchase.2

XXV. Clans and Castes

What picture then can we form of the development of the social structure of India? A system of clans sometimes patrilineal and sometimes matrilineal, but in either case governed by generalized exchange, facilitated either the hierarchical integration of a group of conquerors or the gradual differentiation of statuses within a homogeneous society. In this connexion it is not unimportant to note that Aryan society seems to have been organized as a harmonic regime, i.e., with patrilineal and patrilocal groups, in other words satisfying the theoretical conditions of a system of generalized exchange.5 By taking a purely historical point of view, Hutton, like Senart, concluded that the origin of the caste system was probably pre-Aryan. The Indo-European invaders, it is maintained, were content to crystallize a system of pre-existing prohibitions into a type of social hierarchy. From this came the scheme of Manu with the four varna organized within a hypergamous structure.1 If, as Mitra suggests,2 the four varna must be differentiated into two groups, the dwija or ‘twice born’ and the sudra, with a subdivision of the former group into three sections, we could recognize even more easily the model of the tripartite organizations which still exist among the Naga.

XXVI. Asymmetric Structures

A careful examination of the evidence would seem to show that the relation between uncle and nephew at marriage is especially a feature of Dravidian society’,2 he has in mind other phenomena concerning the particularly close assistance that the nephew receives from the uncle on the occasion of his marriage, that is to say precisely ‘in those ceremonies in which we should expect them to be prominent, if the uncle-nephew relation be a survival of the marriage regulation’.

The term for brother-in-law, used here by the husband to address his wife’s maternal uncle, expresses a kinship relationship which can but need not exist, but which is implied in the system as a permanent virtuality. It is highly likely that the rule mentioned above, according to which a man can marry only his older sister’s daughter but not his younger sister’s daughter, is intended to avoid the possible conflict between father and son, the younger sister’s daughter remaining available for her matrilateral cross-cousin, which conforms to the regulation of marriage in force among the Korava.

There are thus two types of relationship between the uncle and the nephew. There is, on the one hand, the assistance given by the uncle to his nephew when he marries, and, on the other hand, the maternal uncle’s matrimonial privilege over his sister’s daughter, which also creates for him a special type of relationship with his matrilateral nephew

In welcoming new relationships with ritual presents old relationships must not be left idle

In the beginning there was the vibe, the alliance of Nerhè and Rheko, the former of the Ver totem, the latter of the Iule totem. They exchanged their sisters, and had it been a bargain they would have called it quits. But this exchange is not a bargain, it is an arrangement for the future, a social contract: the child that each has by the woman received goes back to take the place that this woman has left among her mother’s people; from generation to generation new gaps are filled alternately in the same way

The gift or the exchange of daughters creates a relationship of affinity; it can also be said that the bond of affinity, already established by marriage with the sister, is a ground for later claiming the daughter

Wherever a simple method of direct or indirect exchange is practised, based on the elementary structures of reciprocity created by the cession of sisters and daughters, the brothers-in-law relationship, or, as we would prefer to say, the relationship of affinity, has an importance characterized by an acute ambiguity. The brothers-in-law depend on one another in a truly vital way, and this mutual dependence can create alternatively, and sometimes also simultaneously, collaboration, confidence and friendship, or else distrust, fear and hate. Most often, arbitration between these opposed sentiments is ensured by a strictly fixed social behaviour and a whole system of obligations and reciprocal prohibitions, of which the parents-in-law taboo (including under this term all those designated in English as ‘in-laws’) is only one element.

XXVII. Cycles of Reciprocity

a human group need only proclaim the law of marriage with the mother’s brother’s daughter for a vast cycle of reciprocity between all generations and lineages to be organized, as harmonious and ineluctable as any physical or biological law, whereas marriage with the father’s sister’s daughter forces the interruption and reversal of collaborations from generation to generation and from lineage to lineage. In one case, the overall cycle of reciprocity is co-extensive with the group itself both in time and in space, subsisting and developing with it. In the other case, the multiple cycles which are continually created fracture and distort the unity of the group. They fracture this unity because there are as many cycles as there are lineages, and they distort it because the direction of the cycles must be reversed with each generation.

XXVIII. The Transition to Complex Structures

The groom works for his parents-in-law, and he receives the counter-prestation of his gifts from his wife in the manifold forms of cooking, gardening, procreation of children and sexual gratification. The exchange of services takes place in both directions, and the double circulation of women and cattle according to an alternating rhythm ensures the union of groups and of generations through the years,

XXIX. The Principles of Kinship

Thus, it is always a system of exchange that we find at the origin of rules of marriage, even of those of which the apparent singularity would seem to allow only a special and arbitrary interpretation. In the course of this work, we have seen the notion of exchange become complicated and diversified; it has constantly appeared to us in different forms. Sometimes exchange appears as direct (the case of marriage with the bilateral cousin), sometimes as indirect (and in this case it can comply with two formulas, one continuous, the other discontinuous, corresponding to two different rules of marriage with the unilateral cousin). Sometimes it functions within a total system (this is the theoretically common characteristic of bilateral marriage and of matrilateral marriage), and at others it instigates the formation of an unlimited number of special systems and short cycles, unconnected among themselves (and in this form it represents a permanent threat to moiety systems, and as an inevitable weakness attacks patrilateral systems). Sometimes exchange appears as a cash or short-term transaction (with the exchange of sisters and daughters, and avuncular marriage), and at other times more as a long-term transaction (as in the case where the prohibited degrees include first, and occasionally second, cousins). Sometimes the exchange is explicit and at other times it is implicit (as seen in the example of so-called marriage by purchase). Sometimes the exchange is closed (when marriage must satisfy a special rule of alliance between marriage classes or a special rule for the observance of preferential degrees), while sometimes it is open (when the rule of exogamy is merely a collection of negative stipulations, which, beyond the prohibited degrees, leaves a free choice). Sometimes it is secured by a sort of mortgage on reserved categories (classes or degrees); sometimes (as in the case of the simple prohibition of incest, as found in our society) it rests on a wider fiduciary guarantee, viz., the theoretical freedom to claim any woman of the group, in return for the renunciation of certain designated women in the family circle, a freedom ensured by the extension of a prohibition, similar to that affecting each man in particular, to all men in general. But no matter what form it takes, whether direct or indirect, general or special, immediate or deferred, explicit or implicit, closed or open, concrete or symbolic, it is exchange, always exchange, that emerges as the fundamental and common basis of all modalities of the institution of marriage.

tends to ensure the total and continuous circulation of the group’s most important assets, its wives and its daughters.

The prohibition of incest is less a rule prohibiting marriage with the mother, sister or daughter, than a rule obliging the mother, sister or daughter to be given to others. It is the supreme rule of the gift, and it is clearly this aspect, too often unrecognized, which allows its nature to be understood. All the errors in interpreting the prohibition of incest arise from a tendency to see marriage as a discontinuous process which derives its own limits and possibilities from within itself in each individual case.

Motherhood is not only a mother’s relationship to her children, but her relationship to other members of the group, not as a mother, but as a sister, wife, cousin or simply a stranger as far as kinship is concerned. It is the same for all family relationships, which are defined not only by the individuals they involve, but also by all those they exclude. This is true to the extent that observers have often been struck by the impossibility for natives of conceiving a neutral relationship, or more exactly, no relationship. We have the feeling – which, moreover, is illusory – that the absence of definite kinship gives rise to such a state in our consciousness. But the supposition that this might be the case in primitive thought does not stand up to examination. Every family relationship defines a certain group of rights and duties, while the lack of family relationship does not define anything;

Throughout a considerable period, and in a large number of societies, men met in a curious frame of mind, with exaggerated fear and an equally exaggerated generosity which appear stupid in no one’s eyes but our own. In all the societies which immediately preceded our own and which still surround us, and even in many usages of popular morality, there is no middle path. There is either complete trust or complete mistrust. One lays down one’s arms, renounces magic, and gives everything away, from casual hospitality to one’s daughter or one’s property.’

It is inseparable from the things which people it.

The ethnographer pressed the point, asking what they would think or say if, through some impossibility, this eventuality managed to occur. Informants had difficulty placing themselves in this situation, for it was scarcely conceivable: ‘What, you would like to marry your sister. What is the matter with you anyway? Don’t you want a brother-in-law? Don’t you realize that if you marry another man’s sister and another man marries your sister, you will have at least two brothers-in-law, while if you marry your own sister you will have none? With whom will you hunt, with whom will you garden, whom will you go to visit?’

The man should teach his children, and this social vocation, practised naturally within the family group, is irremediably compromised if emotions of another type develop and upset the discipline indispensable to the maintenance of a stable order between the generations; ‘Incest would mean the upsetting of age distinctions, the mixing up of generations, the disorganization of sentiments and a violent exchange of rôles at a time when the family is the most important educational medium. No society could exist under such conditions,’

The writer also cites the case of a woman who, married to a two-year old baby and having a child by ‘a marriage companion’, i.e., an official and temporary lover, shared her attentions between the two babies: ‘When she was nursing her own child, she also nursed her infant husband … In this case the husband also readily took the breast of his wife. When I asked for the reason of the wife’s conduct, the Chukchee replied, “Who knows? Perhaps it is a kind of incantation to insure the love of her young husband in the future”.’3 At all events, it is certain that these apparently inconceivable unions are compatible with a highly romantic folklore, full of devouring passions, Prince Charmings and Sleeping Beauties, shy heroines and triumphant loves.4 We know of similar facts in South America.

Likewise, among the Tapirapé of central Brazil, depopulation has brought about a system of marriage with young girls. The ‘husband’ lives with his parents-in-law and the ‘wife’s’ mother is responsible for woman’s work.2 The Mohave husband carries the little girl that he has married on his shoulders, busies himself with household duties, and generally speaking acts both as husband and in loco parentis. The Mohave comment upon the situation cynically, and ask, sometimes even when, the person concerned is present, whether he has married his own daughter: ‘“Whom are you carrying around on your back? Is that your daughter?” they ask him. When such marriages break up, the husband often has a manic attack.’

I myself have been present, among the Tupí-Cawahib of the upper Madeira, in central Brazil, at the betrothal of a man about thirty years old with a scarcely two-year-old baby, still in its mother’s arms. Nothing was more touching than the excitement with which the future husband followed the childish frolics of his little fiancée. He did not tire of admiring her, and of sharing his feelings with the onlookers. For some years his thoughts would be filled with the prospect of setting up house. He would feel strengthened by the certainty, growing alongside him in strength and beauty, of one day escaping the curse of bachelorhood. Henceforth, his budding tenderness is expressed in innocent gifts. According to our standards, this love is torn between three irreducible categories, viz., paternal, fraternal, and marital, but in an appropriate context it reveals no element of disquiet or defect, endangering the future welfare of the couple, let alone the whole social order.

Marriage is thus a dramatic encounter between nature and culture, between alliance and kinship. ‘Who has given the bride?’ chants the Hindu hymn of marriage: ‘To whom then is she given? It is love that has given her; it is to love that she has been given. Love has given; love has received. Love has filled the ocean. With love I accept her. Love. let her be yours.’3 Thus, marriage is an arbitration between two loves, parental and conjugal. Nevertheless, they are both forms of love, and the instant the marriage takes place, considered in isolation, the two meet and merge; ‘love has filled the ocean’. Their meeting is doubtless merely a prelude to their substitution for one another, the performance of a sort of chassé-croisé. But to intercross they must at least momentarily be joined, and it is this which in all social thought makes marriage a sacred mystery. At this moment, all marriage verges on incest. More than that, it is incest, at least social incest, if it is true that incest, in the broadest sense of the word, consists in obtaining by oneself, and for oneself, instead of by another, and for another.

As far as Freud’s work is concerned, this timidity leads to a strange and double paradox. Freud successfully accounts, not for the beginning of civilization but for its present state; and setting out to explain the origin of a prohibition, he succeeds in explaining, certainly not why incest is consciously condemned, but how it happens to be unconsciously desired. It has been stated and restated that what makes Totem and Taboo unacceptable, as an interpretation of the prohibition of incest and its origins, is the gratuitousness of the hypothesis of the male horde and of primitive murder, a vicious circle deriving the social state from events which presuppose it. However, like all myths, the one presented in Totem and Taboo with such great dramatic force admits of two interpretations. The desire for the mother or the sister, the murder of the father and the sons’ repentance, undoubtedly do not correspond to any fact or group of facts occupying a given place in history. But perhaps they symbolically express an ancient and lasting dream.1 The magic of this dream, its power to mould men’s thoughts unbeknown to them, arises precisely from the fact that the acts it evokes have never been committed, because culture has opposed them at all times and in all places. Symbolic gratifications in which the incest urge finds its expression, according to Freud, do not therefore commemorate an actual event. They are something else, and more, the permanent expression of a desire for disorder, or rather counter-order. Festivals turn social life topsy-turvy, not because it was once like this but because it has never been, and can never be, any different. Past characteristics have explanatory value only in so far as they coincide with present and future characteristics.

These bold assumptions concerning the thesis of Totem and Taboo, and the accompanying hesitations, are revealing. They show a social science like psychoanalysis – for it is one – still wavering between the tradition of an historical sociology, looking, as Rivers did, to the distant past for the reason for the present-day situation – and a more modern and scientifically more solid attitude, which expects a knowledge of its future and past from an analysis of the present. Moreover, the latter is clearly the practitioner’s point of view. But it cannot be overemphasized that the path followed in delving into the structure of the conflicts to which a sick man is prone, in order to recreate its history and so arrive at the initial situation around which all subsequent developments took place, is contrary to that of the theory as presented in Totem and Taboo. In the one case, the progression is from experience to myths, and from myths to structure. In the other, a myth is invented to explain the facts, in other words, one behaves like the sick man instead of diagnosing him.

Despite these presentiments, only one social science has reached the point at which synchronic and diachronic explanation have merged, because synchronic explanation allows the reconstitution of the origin of systems and their synthesis, while diachronic explanation reveals their internal logic and perceives the evolution which directs them towards an end. This social science is linguistics, regarded as a phonological study.1 When we consider its methods, and even more its object, we may ask ourselves whether the sociology of the family, as conceived of in this work, involves as different a reality as might be believed, and consequently whether it has not the same possibilities at its disposal.

The progress of our analysis is thus close to that of the phonological linguist. What is more, if the incest prohibition and exogamy have an essentially positive function, if the reason for their existence is to establish a tie between men which the latter cannot do without if they are to raise themselves from a biological to a social organization, it must be recognized that linguists and sociologists do not merely apply the same methods but are studying the same thing. Indeed, from this point of view, ‘exogamy and language … have fundamentally the same function – communication and integration with others’.2 It is to be regretted that after this profound remark its author makes off in another direction, and assimilates the incest prohibition to other taboos, such as the prohibition on sexual relations with an uncircumcised boy among the Wachagga, or the inversion of the hypergamous rule in India.3 The incest prohibition is not a prohibition like the others. It is the prohibition in the most general form, the one perhaps to which all others – beginning with those cited above – are related as particular cases. The incest prohibition is universal like language, and if it is true that we are better informed on the nature of the latter than on the origin of the former, it is only by pursuing the comparison to its conclusion that we can hope to get to the meaning of the institution.

Certain facts taken from psychopathology already tend to suggest that the relations between the sexes can be conceived as one of the modalities of a great ‘communication function’ which also includes language. For certain sufferers from obsessions, noisy conversation seems to have the same significance as an unbridled sexual activity. They themselves speak only in a low voice and in a murmur, as if the human voice were unconsciously interpreted as a sort of substitute for sexual power.3 But, even if one is prepared to accept and use these facts only with reservation (and here we call upon psychopathology only because, like infantile psychology and social anthropology, it allows a more comprehensive way of experiencing the social universe), it must be acknowledged that they receive striking confirmation from certain observations on primitive customs and attitudes. One need only recall that in New Caledonia ‘the evil word’ is adultery, for ‘word’ should probably be interpreted as meaning ‘act’.4

These prohibitions are all thus reduced to a single common denominator: they all constitute a misuse of language, and on this ground they are grouped together with the incest prohibition, or with acts evocative of incest. What does this mean, except that women themselves are treated as signs, which are misused when not put to the use reserved to signs, which is to be communicated?
In this way, language and exogamy represent two solutions to one and the same fundamental situation. Language has achieved a high degree of perfection, while exogamy has remained approximate and precarious. This disparity, however, is not without its counterpart. The very nature of the linguistic symbol prevented it from remaining for long in the stage which was ended by Babel, when words were still the essential property of each particular group: values as much as signs, jealously preserved, reflectively uttered, and exchanged for other words the meaning of which, once revealed, would bind the stranger, as one put oneself in his power by initiating him, something of oneself and acquires some power over the other. The respective attitudes of two individuals in communication acquire a meaning of which they would otherwise be devoid. Henceforth, acts and thoughts become mutually solidary. The freedom to be mistaken has been lost. But, to the extent that words have become common property, and their signifying function has supplanted their character as values, language, along with scientific civilization,1 has helped to impoverish perception and to strip it of its affective, aesthetic and magical implications, as well as to schematize thought.

But that atmosphere of feverish excitement and sensitivity which engendered symbolic thought, and social life, which is its collective form, can still with its far-off vision kindle our dreams. To this very day, mankind has always dreamed of seizing and flxing that fleeting moment when it was permissible to believe that the law of exchange could be evaded, that one could gain without losing, enjoy without sharing. At either end of the earth and at both extremes of time, the Sumerian myth of the golden age and the Andaman myth of the future life correspond, the former placing the end of primitive happiness at a time when the confusion of languages made words into common property, the latter describing the bliss of the hereafter as a heaven where women will no longer be exchanged, i.e., removing to an equally unattainable past or future the joys, eternally denied to social man, of a world in which one might keep to oneself