This is a blockbuster of a book. Nothing like it has been written since Lévi-Strauss’s Structures élémentaires de la parenté (1949) or Meyer Fortes’s Kinship and the Social Order (1969). Yet in the sweep of its evidence and argument, Godelier’s summa is more ambitious and far-reaching than either of these. It is at once a major intervention in the discipline of anthropology, and a work of the widest human interest. Kinship has the reputation of being the most technical department of anthropology, the least accessible to a general public. But while Métamorphoses synthesizes a huge range of complex materials, it is written in an unfailingly lucid style that makes no assumptions of professional familiarity with terms and debates about kinship, but always takes care to explain them in language anyone can understand. The book is both a monument of scholarship and a gripping set of reflections on universal experience. It is certain to be read and discussed for years to come.

Godelier introduces his work with a contemporary paradox. Traditional kinship patterns in the West are in dramatic dissolution today, as heterosexual marriage declines, biological and social parenthood become dissociated, homosexual unions are legalized. Yet in the same period, anthropology—where the study of kinship was once the basis of the discipline, ‘comparable to logic in philosophy and the nude in art’—has all but completely turned its back on it, since the rebellions against Lévi-Strauss of Leach (Rethinking Anthropology in 1961) and Needham (Rethinking Kinship and Marriage in 1971), followed by the clean sweep of Schneider (Critique of the Study of Kinship in 1984), to the point where it is scarcely even referred to by postmoderns like Clifford and Marcus. Can it be that anthropology has nothing to say about the upheavals going on around us? Godelier intends to show the opposite.

To set the scene, and exemplify some of his principal arguments, he first describes his own fieldwork among the Baruya, a little-known tribe in the highlands of Papua New Guinea, where he first arrived in 1967. Over the next two decades, he spent a total of seven years with them, amassing an experience in the field that has been surpassed by few, if any. There was a time when the work of Malinowski and his British-based pupils provided the ultimate in ethnographic achievements while that of anthropologists elsewhere, including France, was regarded as relatively superficial—Lévi-Strauss’s own rather brief forays among the Nambikwara being an example. Godelier has completely reversed the basis of that judgement. Here he develops an excellent summary of the kinship institutions of the Baruya, in a chapter that sets the tone for much of the book. Among his most striking findings was a secret ritual of the fellation of young males by boys, over a period of years, practised in the belief that women are a source of cosmic disorder, and that sexual relations with them ‘represent a permanent threat to men’, who thereby ‘risk losing their strength, their beauty and their superiority’. Hence the drinking of sperm by young men to re-engender themselves prior to marriage—a liquid that is transformed after marriage, in the passage through women, into milk for children. The belief system surrounding this requirement for maturity, he observes, constitutes a formidable ideological basis for masculine domination, whose ramifications in other domains of life he sets out with an intimacy of knowledge that inspires admiration and confidence. His is an ethnography of extraordinary depth.

After this empirical testing-ground, Godelier proceeds to a systematic survey of what he posits as the six fundamental components of any kinship system: filiation and descent; alliance (principally marriage); residence; terminologies; conception; sexual prohibitions (principally incest). Across some three hundred pages, the range of variation in each is explored with a wealth of vivid illustration. Out of this conspectus, and his own experience, Godelier draws several key conclusions. Contrary to common belief, kinship is never the basis of pre-class societies. Their cohesion always rests primarily on religious–political relations. There is no knowing in advance the importance of kinship in any given society, which can vary widely, but it cannot be either a prime mover or a self-standing system. Because they involve a distribution of power, kin relations are typically inscribed by other kinds of relation, but they lack the capacity to modify these. Changes in a kinship system generate only new kin relations, never modifications of caste or class relations. In no society is the function of kinship to organize economic, political or religious life. It is to govern descent and alliance.

What, then, of the prohibition of incest, in whose universal imposition of exogamy Lévi-Strauss—following Tylor’s dictum, ‘either marry out or be killed out’—saw the original passage from nature to culture defining human society as such? A more foundational role for kinship would be difficult to conceive. In two long chapters, Godelier addresses this famous issue. Little is left of Lévi-Strauss’s construction by the time he has finished, and still less of the ideas of Françoise Héritier, who succeeded Lévi-Strauss in his chair at the Collège de France. He notes that the incest taboo is not an invariable feature of all societies. The exchange of women between men is not a universal feature of kinship systems. Exchange itself is not the common basis of all matrimonial, let alone social arrangements: relationships of gift, sale and retention (non-exchange) are equally important. The notion that the origins of humanity are to be found in a ‘Big Bang’, with the ex abrupto invention of language as the exchange of words and marriage as the exchange of women, is a myth. Godelier shows that, when faced with incontrovertible evidence that many of the principal claims of Structures élémentaires de la parenté were untenable, Lévi-Strauss either changed the subject, or attempted to gloss them with inconspicuous rewordings and less than forthright retractions.

Among the discoveries that have made short work of Lévi-Strauss’s story of the foundations of society have been the findings of primate studies, to which Godelier devotes a sensitive and imaginative chapter. What these have shown is that both chimpanzees and bonobos (pygmy chimpanzees in the Congo), our nearest biological relatives, already live in ‘societies’ that exhibit a kind of sketch of human constraints: young females find sexual partners outside their immediate natal group, while young males must wait their turn until adults are willing to yield partners to them. Enforcing at once cooperation and hierarchy, these patterns appear to be the product of mechanisms of natural selection, though they coexist with homosexual pleasures among males and females alike, less obviously attributable to the same functions. The passage from nature to culture with homo sapiens thus cannot have been a sudden, discontinuous transformation, but must have been more evolutionary in nature. The critical novelty in human society, Godelier argues, is that males assume a parental role, something unknown among these primates, where only mothers look after children—fathers being unaware of their connection with them.

Where does this leave the taboo on incest? Rather than insisting that it is ubiquitous—in face of the facts of history, which show that brother–sister, father–daughter and mother–son relations have in some societies, such as Ancient Egypt or Achaemenid Persia, not only not been prohibited, but even enjoined—Godelier suggests that what is actually universal is something simpler. The sexual drive is fundamentally asocial: notoriously no respecter of rules, it even particularly delights in breaking them. Hence for society to be possible at all, it must be constrained. Any society requires therefore the existence of some sexual prohibitions as such. These, however, can take any number of different forms. If taboos on incest are far the most common of these, that is because they guard the door to the parenting unit that distinguishes human from primate societies:

There nowhere exists a society where the individual is authorized to satisfy all his sexual desires (and so also fantasies). And it is always at the threshold of the social units within which men and women cooperate to bring up children, whether or not they have given birth to them, that the most extreme forms of sexual permissiveness have been halted.

Métamorphoses de la parenté ends with a panorama of transformations in kinship today, which focuses principally on the West, where the changes under way are the most dramatic. Historically, Godelier maintains, humanity has exhibited a certain evolutionary tendency in its alliance systems, from a common ‘Dravidian’ starting-point, whose changes have been irreversible, but not (so far) unilinear—a pattern he surmises is likely to hold in the future too. Humans, however, are the only species co-responsible with nature for their own evolution. In the past they rarely acknowledged their own role in creating rules of kinship, but now they can scarcely do otherwise, as laws and customs governing relations between and within the sexes are in full mutation, with the spread of single parenting, homosexual marriage, artificial insemination and the prospect of cloning all now crowding onto the public agenda. In the last lines of his book Godelier reiterates that ‘what separates human beings definitively from primates, their cousins in nature, is that they not only live in society but can and must produce society in order to live’. It is one of the underlying messages of this work that in confronting the unexpected in that task today, the sang-froid of the anthropologist is needed.

If such are the main lines of Métamorphoses, how should they be situated? Godelier is right to say that anthropological studies of kinship have declined in recent years. But this has been partly due to the failure of leading practitioners, among them Needham and Schneider, to look beyond a restricted set of materials. By narrowing the field, they contributed to the virtual disappearance of interest in the area for which the discipline was best known, indeed that has sometimes been seen as its only ‘professional’ competence. But while a good many anthropologists were deserting the field (partly, one suspects, because it did require more professional competence), others in the social sciences—historians of the family, demographers, psychologists—were becoming more interested in it, a development not reflected in Métamorphoses. Within the discipline itself, Godelier’s book can be seen as an attempt to reconcile two theoretical stories personified respectively by Lévi-Strauss and Fortes, one from France and the other from Britain, that gave opposite importance to marriage and to filiation, lateral versus lineal ties, within the ‘atom of kinship’. Others have tried to do something similar, but none so thoroughly. This is a great achievement. But it is not a completely balanced one, since in practice Godelier devotes the greater part of his book to systems of marriage and their associated terminologies rather than looking at relationships of filiation or descent. Yet if to this extent his principal concerns remain characteristically French, the critical perspectives he brings to them draw on central elements in the Anglo-Saxon tradition.

Whether these correctives always go far enough is a matter of judgement. A look at Godelier’s handling of alliance and filiation, of kin terminologies and dowry systems, and not least of incest, will show some of the issues at stake. Thus, as we have seen, Godelier distances himself from Lévi-Strauss’s claim that kinship systems are always based on the exchange of women between men. If this is true in some societies, in others—of which modern Europe is an obvious example—there is no exchange at all. What seems to be needed, however, is a yet stronger recognition that even where there is such an ‘exchange’, it cannot be viewed as analogous to the exchange of a commodity, because what is being transferred is a bundle of rights and duties, moving in both directions. That is, a man obtains—perhaps through bridewealth or bride service—certain rights over a woman, but the woman too obtains certain rights—to receive food, ‘medical treatment’, sexual attention—from the man. That situation, although unequal, requires a more subtle and complex analysis than the simple economic metaphor of exchange usually allows. An alternative British tradition draws heavily on studies like Hohfeld’s Fundamental Legal Conceptions (1923) to analyse situations of this kind in terms of a quasi-juridical grid of rights and corresponding duties.

Although Godelier speaks of the exchange of brides between two men, in fact a woman rarely loses her status in her natal lineage, so we are in fact dealing with an exchange of rights in women rather than of women themselves. In noting that the gift of a bride is not cancelled out by a return gift between lineages, Godelier tacitly recognizes this. Because the person ‘given’ never ceases to be a member of her native lineage, only certain rights in her are transferred. But although Godelier by no means concentrates exclusively on obligations to exchange or to give, going out of his way to consider obligations to conserve and to transmit as well, this does not lead him to ask how the timing and the content of transfers affects kinship relations in fundamental ways. This is true not only of descent, to which he pays much less attention, but of marriage. If in the long term marriage systems often involve the exchange of rights over women and/or men, direct exchange is rare. More usually the system is mediated by bridewealth or other prestations (like bride service), the amount, nature and timing of which influence the resulting relationships very strongly.

There is certainly no shortage of empirical material about marriage in Métamorphoses. Godelier has undertaken a massive survey of variations in the institution across the world. Nevertheless there are some important gaps. Just as we do not hear much about the major societies of Eurasia, apart from rather general references to Hindu and Chinese marriage, so too we find little more than casual allusions to contemporary Europe and America until the last chapter. Surprisingly, Godelier ignores the demographic research which has been so important a factor in family studies, and made its impact in many related spheres. This is not to say it should be appropriated uncritically. Following the lead of Malthus, many demographers, like other social scientists, have drawn a sharp distinction between the European and non-European family systems, particularly those of Asia. But it is often an error to oppose the West and the Rest in a categorical manner, as large numbers of anthropologists, sociologists and historians continue to do. That leads to the kind of mistakes Malthus made about China. Comparable errors have been made by anthropologists—Durkheim treating the Chinese as exemplars of ‘primitive classification’, Dumont positing a decisive break between a hierarchical India and a more egalitarian West, or Lévi-Strauss bracketing early Chinese with Australian marriage systems.

Such positions have done considerable harm to social and historical studies. It did not require the work of Joseph Needham to make it clear that China had as advanced a civilization as Europe at least until the sixteenth century, and that comparisons between the two have to be seen in these terms. More generally, in certain significant ways Asian family systems resembled European, not least in the implications of what I have called ‘the woman’s property complex’ deriving from dowries. Unlike bridewealth transactions in marriages in Africa or New Guinea, which are largely exchangist (but may also involve parental contributions), the dowry of Eurasian societies is in most cases a downward transmission of wealth from the parents to the bride (and therefore varies according to parental holdings). Here marriage does not involve an ‘exchange’ of goods so much as a confirmation of status. It can be regarded as, in effect, a form of pre-mortem inheritance. This distinction is not always perceived as clearly in France where the word dot can be used for both transactions, exchange between affines (‘alliance’), and transmission to heirs (‘filiation’).

Godelier might have taken into account work on dowry systems that points to a common Eurasian repertoire of strategies of heirship and household management, but he sidesteps this issue. Part of the problem here may have to do with the primacy of fieldwork in the anthropological tradition, and a compounding rejection of speculation about the longer-term development of humankind. The choice of this path was justified by solid reasoning. Nevertheless a price can be paid for spending years living among a particular tribe, or even writing about it. One starts, and one tends to finish, with the nature of its descent groups or alliances, neglecting the externalities of time (history) and space (comparison).

A related bias derives from the salience of kinship terminology as the topic that gave anthropology its initial boost, because the data it furnished were distinctive, limited in variety and easy to collect. The American lawyer-turned-anthropologist Lewis Morgan lent this impetus by sending a questionnaire to us consuls around the world in the 1860s. But they provided limited information regarding the actual societies from which the terms were taken, even in respect of interaction between close kin. For that matter, was clanship patrilineal, matrilineal or bilateral? The answers were rarely clear-cut. As Godelier points out, terminologies of kinship are ‘independent of the system of descent’, of attachment to ancestors. Going further, he remarks that ‘the symmetric exchange of spouses between two groups’ cannot be deduced with any certainty from a terminological system. Such scepticism is not entirely justified; some correlations between terms and practices can be found. Godelier accepts that Eskimo-type terminology is associated with an absence of descent groups (bilateral systems), and thinks kin terms are of use for understanding some of the ‘marriage class systems’ of the Australian aborigines. But otherwise the classification of a terminology as, say, of an Omaha or Iroquois type is of marginal interest if one is concerned with analysing the interaction of members of a society at a domestic level, including marriage.

By and large Godelier’s arguments here are correct, even if too often he inconsistently continues to refer to ‘systems of kinship’ as types of terminology (‘the Dravidian type’ and so forth). Discussing kin terms deriving from Western Polynesia, where tamai means both father and father-in-law, he concludes that the two must therefore be treated as if they were the same. But this does not follow at all. Some English-speakers refer to their mother-in-law or their stepmother as ‘mother’, but that is very different from saying that they are treated identically. In general, there is no necessary correspondence between terms and relations. English kin terminology did not change when the Reformation altered the rules about who you could marry, limiting the ‘prohibited degrees’ of consanguinity.

So while Morgan’s systematic study of kinship through classification of its terminology was one of the founding acts of anthropology, establishing it as a discipline among the social sciences, the direction he set was to have considerable drawbacks. Moreover its canonization led to the neglect of what was actually his more significant work, Ancient Society, deprecated by many anthropologists of the twentieth century because of their antipathy to ‘conjectural history’. But whereas some of Morgan’s work, supposedly derived from the study of kin terminologies, was certainly speculative in a less than satisfactory way, much of Ancient Society was based upon archaeological evidence that was reasonably ‘hard’, and important for discussion of socio-cultural developments more generally. One consequence of this neglect is the curious failure of so many anthropologists to give full weight to economic factors in their work. Paradoxically, this is true of Godelier as well. For example, marriage in New Guinea is virilocal, defined by the residence of the husband, whereas in contemporary Europe it is neolocal, involving a new residence for the couple. The contrast cannot be unrelated to the fact that in the first case we are dealing with a society based on simple agriculture, in the second with an industrial economy where, residence apart, the command of specific plots of land matters for very few. Equally, cultivation of the land by the hoe or by the plough has different economic implications for kinship (and politics). Engels, famously much influenced by Morgan’s Ancient Society, understood the logic of such links between production and reproduction. The effect of Godelier’s position, by contrast, is rather to isolate marriage from other transactions in society, for example the transfers involved in inheritance, to which he gives no systematic attention.

That there is something paradoxical in this can seen from Godelier’s own account of his intellectual trajectory, since—as he explains—he started out as a Marxist whose primary concern was economic anthropology. This was at a time, however, when Marxist culture in France was heavily under the influence of Louis Althusser, whose theoretical outlook was hostile to most traditional conceptions of Marxism—insisting, in particular, that if economic relations determined the form of a society in the last instance, ‘the lonely hour of the last instance never comes’. In appearance a call for greater complexity, this in practice often led to an aversion to economic explanations of any kind. Traces of this Althusserian formation are still visible in Godelier’s work thirty years later, long after his early interests had changed direction, and can be seen in his treatment of the problem of incest. He was originally a pupil of Lévi-Strauss, who was responsible for orienting him towards New Guinea, and his starting-point is Lévi-Strauss’s conception that the logic behind the prohibition of incest lies in a compulsion to exogamy if human beings are to coalesce into a society, establishing external ties with one another rather than duplicating internal ones within the family. For Lévi-Strauss, kinship is fundamentally a system of alliance, in which the meaning of the incest taboo is that men ‘renounce their sisters to exchange them against the sisters of others’.

Such a theory poses at least three questions: of description, of extension and of explanation. Putting it another way: does it accurately capture the prohibitions on incest; if so, what is the range of its application; and how persuasive is its causal mechanism? The first thing to be said is that its definition of the problem is one-sided, for it says nothing about the taboo between affines, to use Lévi-Strauss’s vocabulary. Yet I early on argued for the need to distinguish between lateral prohibitions affecting ‘brother’ and ‘sister’, which are at the core of the explanation of Tylor, Lévi-Strauss and the exchangist approach, and lineal prohibitions between generations. These had become fused in the Christian concept of incasta, incest, leading to the mistake of seeking a general explanation of two distinct prohibitions, on sisters and on mothers. They required, however, complementary explanations, which were not to be found in Structures élémentaires de la parenté.

Godelier is well aware of such Anglo-Saxon objections to Lévi-Strauss’s original construction, citing them in his own thorough critique of it. His view of the subsequent twist given the subject by Françoise Héritier is still sharper. Her book Les Deux Sœurs et leur mère (1994) claims to have discovered what she calls a ‘second type of incest’, the taboo that forbids a man to sleep with two sisters or a mother and daughter. Héritier attributes the force of the Christian idea of incest not to the sin of ‘one flesh’—biological or social consanguinity—but ‘one gender’. Her notion has to do with the passing of substances between two sisters, or mother and daughter, by way of a male, what Godelier describes as a ‘mechanism of fluids’. Incest here is held to occur between the same sex, as the two women mix identical substances. This Héritier takes to be the conceptual basis of all forms of incest, a proposition based on her observation that the relationship between kinsfolk of the same sex is stronger than between those of the opposite sex (brother and sister). For her, the fundamental incest prohibition is that between mother and daughter. Godelier leaves no doubt that he regards this as an extravagant construction that ignores the explanations of actors themselves in the name of Lévi-Strauss’s postulate that the symbolic has primacy over the imaginary and the real—an idea responsible elsewhere in Lévi-Strauss’s work, he notes, for the aberrant notion that the subordination of women is founded on unconscious structures of symbolic thought embedded in the brain.

Passing to the extent of taboos on brother–sister unions, Godelier notes that for many years French scholars, under the influence of Lévi-Strauss, disregarded the evidence that such interdictions were not to be found in Ancient Egypt or Persia. Different kinds of ‘close marriage’, indeed, were characteristic of much of the Mediterranean area before the coming of Christianity. But although he spells out the way such patterns undermine any idea of the universality of incest prohibitions as conventionally understood—in Persia not only brother–sister, but father–daughter and even mother–son coupling was sanctioned—he attributes such ‘exceptions’ essentially to local cosmogonies, in which selected humans could imitate the conduct of gods. In royal families, this would have been one element of the situation. But there were more terrestrial considerations as well. Godelier cites Herrenschmidt’s report of a Persian tale as late as the eleventh century ad, in which a mother says to her daughter: there is no one in Iran worthy of you except the prince Virou, your brother. Conservation of rank was certainly important in such cases. But in Ancient Egypt, as Keith Hopkins has shown, brother–sister marriages extended throughout the population. What needed to be conserved was not just rank, but property. For this was a society in which the status of women was high, and women possessed their own goods through the dowry. Irrigated land was extremely valuable, and many were involved in the conservation of differentiated property.

‘Close marriage’ of this kind was thus not simply a matter of the continuity of the group, as Godelier implies. A paradigmatic case can be found in Ancient Israel, when the daughters of Zelophehad were given the right to inherit from their father if he had no male heirs, but at the same time told they must marry within the clan, so the property would not be dispersed. It was the same with the epikleratic marriages of Ancient Greece, where an heiress had to marry within the kin group, to a father’s brother’s son—as in the contemporary Arab world. Close marriage conserves both rank and property, in a way often seen as ‘incestuous’ in other systems. From the Bronze Age onwards, in my view, stratified urban societies attempted to preserve the status not only of sons but of daughters, by means of a dowry which allocated sisters part of the parental wealth (rarely equal to that of their brothers). There was thus a general tendency in these societies to marry into the same wealth or status group, even occasionally into the same family.

Godelier resists such explanations. He prefers to look for the origin of close marriage in religious beliefs, and criticizes Keith Hopkins and myself for too ‘Marxist’ a view of incestuous unions in Ancient Egypt. A fascination with the ideological runs throughout his book, in which religious and political relations—but there is more on religion than politics here—are accorded primacy over all others in creating social cohesion. In part, this preoccupation goes back to an earlier phase in his career, when he was more committed to Marxism himself, and seeking an answer to the question of how classes could arise in pre-class societies—something that, he argued, no economic mechanism could explain, but only the consent-inducing powers of divine authority. But it also relates to his view of the failure of Mauss in his essay on the gift (which inspired so much of Lévi-Strauss’s thinking) to realize that there are objects that cannot be exchanged in many simple societies. Some, like land, can only be transmitted—that is, inherited by kin. Godelier show little interest in these. Others are ‘sacred’ objects—not, of course, necessarily physical—that are conserved rather than transmitted. With these he is much concerned.

Criticizing an observation of mine that religion was an aspect of the imaginary, Godelier insists that it is, on the contrary, ‘a social reality’. But, of course, what is imaginary may also be real, if it is real in its consequences. Religion is no less ‘real’ than such products of the imagination as Madame Bovary or Our Mutual Friend. But dogmas typically have much greater consequences than novels, because they enable the creation of ‘great organizations’. Religions are in that sense mighty historical realities. That does not mean we are unable to consider such notions as that the Jews are the Chosen People, that Christ is the Son of God, that Allah spoke to Muhammad, as imaginary. These are ideological constructions, and even the most ardent believers must agree that although their religion may not be imaginary, the others must necessarily be so. They cannot all be ‘true’, ‘real’, in that sense.

If such beliefs are imaginary, then there is even more room for invention than Godelier suggests, given that there is no tight functional or structural link between them and other aspects of the society. Notions of the creation of the individual, like notions of the creation of the world, display internal contradictions that give rise to variations in such myths, which are sometimes more theocentric, sometimes more anthropomorphic, as I have tried to show in the case of the Bagre recitation of the LoDagaa in West Africa. Typically, the problem in the past has been that, owing to the constraints of time and the nature of fieldwork in oral cultures, only one version of a myth has usually been recorded, providing an inevitably restricted window on the nature of the imagination in such societies that misses its range of variation and contradiction. Moreover, if the information has been obtained by deliberate enquiry rather than by observation or overhearing, then subjects may be pushed into a simpler and more single-stranded response about, say, the nature of the soul, than does justice to the complexity of their ideas about it. It is difficult to account for the range and variety of beliefs of this kind over relatively small areas unless one appreciates the fertility of people’s imaginations.

More generally, most anthropological—for that matter, many sociological and historical—accounts represent dominant ideologies as singular and as unambiguous, both in content and interpretation. The mythology of the cattle-keeping Tutsi, making them the eldest of three brothers created by God, places them unambiguously at the top of a hierarchy in which Bantu farmers and Twa hunter-gatherers rank below them. However, an alternative version, preferred by the other two groups, sees the brothers as equals, at least before God. Similarly, Hindu myths in India represent only part of the total ideological scheme, which includes not only Jain and Buddhist rejections of the caste system, but untouchable groups who are and have long been yet more rejectionist, even atheistic. Examining in depth notions of conception and birth in a range of human societies, at one level Godelier shows their remarkable diversity (along with certain more widespread commonalities). At another, he himself notes that they are not always uncontested, as when Baruya women remain unconvinced that their milk is in fact their husband’s sperm.

Arguing that kinship changes only through ‘the evolution of social relations which in themselves have nothing to do with kinship’, Godelier again singles out religion as one of the most significant forces inducing people to modify their ways of regulating alliance and descent. He cites as an example Muslim encouragement of marriage to the father’s brother’s daughter. But while it is true that Islam has been associated with this form of marriage, the practice preceded its coming, since it was found in Ancient Greece and Ancient Israel—in circumstances probably connected with transmissions of property, which had little to do with any specific religious creed. However, a particular religion might well insist upon a particular marriage as a way of maintaining the distinctiveness of its congregation and at the same time increasing its control of it.

According to Godelier, basing himself here on the work of Héritier, Christian rules on kinship went back to the adoption by the Eastern Church of Roman law which he maintains—exceptionally for the Mediterranean—forbade ‘all marriage between consanguines to the seventh degree calculated by the Roman manner’. But Roman customs on forbidden marriages evolved over time, and certainly changed as Christianity became more dominant. The Roman law of which Godelier speaks was already influenced by the coming of Christianity. According to my reading, earlier Roman society often allowed close marriage, including that of cousins. Why would it differ so radically from Greece, Judaism, Egypt and other Mediterranean societies, until Christianity laid down its divergent rules? Elsewhere, contradictorily and in my view more accurately, Godelier concedes that in Rome, too, marriage with close relations, for example first cousins, was permitted.

What of the role of religion in all this? For Godelier, the Christian insistence on distant marriage was a matter of ‘reducing to a minimum the accumulation and mixture of identical flesh marked by original sin’. That was not a concern of the Hebrews, as Augustine’s strenuous efforts to gloss the myth of Adam and Eve—Godelier quotes a splendid passage—remind us. Can the contrast between the Jewish practices of close marriage in the Old Testament, and the multiplication of prohibitions in early European marriage by Christianity, be simply explained by differing theological attitudes to desire? Given the formidable extent of the debt of Christian to Jewish religion, it seems unlikely. I have attempted a more concretely based explanation, pointing to the interests of the Church in channelling wealth away from kinship groups to the ecclesia. The requirements of building a ‘great organization’ appear to offer a more plausible logic for the banning of seven degrees of consanguinity than fears of the doubling up of original sin. Such reservations notwithstanding, Métamorphoses de la parenté is a hugely impressive work. There has never been a book that adequately covers the range of human kinship and domestic organization. This is as near as anyone has got.