The Church's crusade against cousin-marriage did not create the Western nuclear family

Emmanuel Todd gets his vengeance

Emmanuel Todd is not well known in the English speaking world. He surfaced in 1976 when in The Final Fall: an Essay on the Decomposition of the Soviet Sphere he predicted the coming Soviet collapse on the basis of an unprecedented rise in the infant mortality rate between 1970 and 1974. In later work, he would outline a radical vision of anthropological particularism orthogonal to the traditional racialism. What truly distinguished the world’s major socio-politico-cultural formations, he would argue, was family structure.

The Explanation of ideology: Family Structures and Social Systems (1985) was based on the insight that the world map of communism corresponded extremely closely with a specific family system, that of the exogamous communitarian family. Following Walt Rostow, the Kennedy intellectual, Todd argued that the violent process of modernization exposed societies to communist subversion in a brief window when the traditional institutions, such as the authority of the landlord, have decayed but new institutions and higher standards of living are still not in place. Todd’s corollary to Rostow’s dictum was that, while all societies are traumatized by the process of modernization, some societies were more exposed to communist subversion than others. And what determined this conditional exposure to the risk of communism subversion was a specific sort of traditional family system where marriage was exogamous (cousin marriage was taboo) and there was coresidence of the father and all his married sons.

In general, Todd’s classification scheme rests on three questions.

(1) Do married sons reside in the same household as their parents? If no sons do so, that’s a nuclear family system. If the eldest son does, that’s a stem family. If all the sons do, that’s a communitarian family. Denmark, northern France, and the Anglo-Saxon countries were traditionally nuclear and characterized by strong norms of independence of children and a high status of women. Germany, Japan and Korea were traditionally stem characterized by patriarchal and authoritarian norms. China, northern India and the Slavic world were traditionally communitarian and characterized by strong horizontal ties between brothers and egalitarian norms. Both stem and communitarian families are generally patriarchal and feature a low status of women.

(2) Is there a norm of an egalitarian distribution of property inheritance between sons? Societies with such a norm are egalitarian; societies without one, for instance, where the first-born is expected to inherit the property, are inegalitarian. The traditional inegalitarian nuclear family of England and Denmark is called the absolute nuclear family. Conversely, that of northern France is called the egalitarian nuclear family. By construction, stem families are inegalitarian and communitarian families are egalitarian.

(3) Is cousin marriage allowed, preferred or taboo? Societies where cousin marriage is preferred are endogamous; societies where they are taboo are exogamous; societies that don’t care much about the question are undifferentiated. Examples of the last are hard to find. But it is, as we shall see, an important configuration in Todd’s schema. The Greater Islamic Middle East is endogamous; Christendom is aggressively exogamous. The endogamous family type makes the development of arms’ length, impersonal institutions difficult because of strong clan ties. Exogamy implies greater social flux between kinship groups.

Both stem and communitarian family systems feature strong parental authority.

There has been a revival of interest in Todd’s family systems. In 2016, David Le Bris indirectly tested Todd’s predictions (the preceding black and white maps are taken from Le Bris’s paper). He showed that ‘inequality among siblings favors investment in physical capital, while a high status of women and strong parental authority favor investment in human capital.’ Already in 2007, Duranton et al. had shown that regional polarization within Europe is predicted by Todd’s classification. They look at ‘household size, educational attainment, social capital, labour participation, sectoral structure, wealth, and inequality’ and find that ‘medieval family structures seem to have influenced European regional disparities in virtually every indicator considered.’

Gutman and Voigt (2020) find that ‘countries in which authoritarian
family types dominate have much higher levels of the rule of law and
innovation than predicted by Todd’. Looks like the authors did not read Todd carefully. He specifically says that stem family structures are especially deferential to state authority and exhibit a different but very dynamic form of capitalism, as evident in Germany, Japan and Korea. Meanwhile, they also find that, consistent with Todd’s argument, ‘countries in which endogamy is frequently practiced display a high level of state fragility and have weak civil society organizations.’

While explicitly rejecting monocausal explanations of macrohistory, Todd argued that family systems are the deepest, most highly conserved social structures that persisted for thousands of years and strongly conditioned social evolution, including ideological and economic change. The persistence of regional family structures is crucial to Todd’s radical anthropological particularism. The deep structures of family systems are so important because they are highly persistent. The reason for their persistence is paradoxical. Despite being at the core of folkways, family structures persist because they are subconsciously and weakly held. So, migration does not erode them because migrants find it easy to abandon their traditional family structures and adopt those of the situated community.

The persistence of family structures is not well understood by contemporary social scientists. Neither is the full import of Todd’s radical particularism. China, India, Iran, Russia, Germany, and England and her settler colonies are all differentially situated with respect to their traditional family arrangements. How did they come to be so? On the question of the origins of family systems, Todd takes Peter Laslett’s discovery of the time-depth of the Western nuclear family as the point of departure. What Laslett discovered in 1972 was that the idea that modernization had eroded traditional large extended families in the West — central to Durkheim’s ‘law of contraction’ of the family — was nonsense. In the West, by which he meant north and west of Germany, he discovered that nuclear families had been the norm in the 17th century.

Todd takes this idea and runs with it. The principle of conservatism of peripheral areas (le principe du conservatism des zones périphériques) shows that archaic features survive on the periphery of continents, while more derived features from later expansions occupy the central region. This is a general principle in biogeography and was well understood by high racialist scholars who tried to infer population history from ethnology, physical anthropology and geography. The most fascinating instance of this tendency can be found in Griffith Taylor’s Environment and Race (1926), from which the following schematic map of racial history is taken. The specifics of the population history Taylor imagined haven’t held up. But recent advances in ancient DNA have more than corroborated Taylor’s racial catastrophism without the racial scaffolding — the dramatic role played by large-scale population pulses in deep history. Contemporary paleoanthropologists speak of the range expansions of breeding populations situated in time and space called paleo-demes.

Anyway, so Todd becomes convinced that the initial family type of Pleistocene hunter-gatherers was that of the undifferentiated nuclear family — undifferentiated in the sense of non-patriarchal and featuring a high status of women. What convinces him is the empirical evidence. From the Kalahari Khoisan Bushmen in South Africa to the Yukaghir of northern Siberia, from the aborigines of Tasmania and Australia to the Indians of Tierra del Fuego at the southernmost tip of South America, all geographic isolates with presumably deeply-conserved and archaic social structures sport undifferentiated nuclear families. In a truly provincializing-Europe move, Todd argued that the societies found at the western extremity of Eurasia similarly displayed archaic features — that is why the nuclear family was traditional in England, Holland, Denmark and northern France!

More generally, he argued that the arrow of time in the anthropology of family systems was inverted: it went from simple to complex and uniform to differentiated. What happened over the course of deep history is that family forms evolved from the undifferentiated exogamous nuclear family with weak kinship ties to the endogamous communitarian family with strong kinship ties and the macroregions of the world became progressively more differentiated in family structure. And this process was staggered — much more advanced closer to the center of innovation in the Middle East than further away. We may speak of the combined and uneven development of family structures in deep time.

Having inverted the traditional schema, Todd argued that modernization favored the archaic nuclear family over the advanced communitarian family because it is more socially fluid and found it easier to create impersonal social systems that would cannibalize the functions performed by the family — the sense in which Durkheim was right about the family withering away.

Schulz, Bahrami-Rad, Beauchamp and Henrich (2019) have argued that the exceptionalism of Western norms and moral attitudes is due to the erosion of kinship at the hands of the medieval Church in Europe. Specifically, they argued that the Church’s crusade against cousin marriage was responsible for the erosion of kinship, which was in turn responsible for the West’s peculiar hyper-individualism etc. They marshal a ton of cross-sectional evidence for their case. But the evidence is not dispositive. Unfortunately for them, their entire exercise assumes that the initial conditions of family structures in the West on which the Church acted were endogamous communitarian — thus directly contradicting Emmanuel Todd’s thesis.

Joseph Henrich, one of the coauthors of the paper and a professor at Harvard, wrote an entire book arguing that the Church’s crusade against cousin marriage was responsible for the Western mindset. He assumes that kinship structures were considerably thicker in the Western past. In particular, that cousin marriage rates were much higher in Western Europe at the time the Church began its crusade against endogamy. What Henrich and co are saying, in short, is that the Western nuclear family that Laslett found in the 17th century was a recent construction. And in this construction process, the Church was the prime mover.

In Henrich is right, Todd must be wrong about the archaic character of the Western nuclear family. Examination of the evidence shows that Todd is right and Henrich is wrong. The reason is simple — we can rule out the Henrich hypothesis. The alternate hypothesis is that the Church’s war against cousin marriage was directed at the nobles, who did indeed practice it in a manner that isolated them from the dominant family system of their societies. Just as socialism could spread easily over the exogamous communitarian anthropological base and found itself blocked on its boundaries and Islam likewise for the endogamous communitarian base, the Church’s influence may have been greatest in exogamous anthropological terrain. In other words, the alternate hypothesis inverts the causal arrow between family systems and the Church’s influence. The correlation between them is explained by a causal vector pointing in the reverse direction — the exogamous anthropological base explains the extent of medieval Christendom.

In forthcoming work, David Le Bris and Victor Gay show that the correlation between Church influence and cousin marriage vanishes if you control for Todd’s principle of conservatism of peripheral areas — operationalized by geodesic distance from the center of Neolithic innovation in Sumer. This is consistent with Neolithic time-depth being a mediator between Church policy and cousin marriage. But Neolithic time-depth cannot be a mediator between Church policy and cousin marriage because causes have to precede effects in time. Meanwhile, cousin marriage is more strongly correlated with distance from the center of Neolithic innovation than Church exposure. These conditional correlations suggest that Todd is right and Henrich wrong. But there is a much stronger reason to think so.

With the intellectual-technical revolution in ancient DNA, we can actually look at the initial anthropological conditions on which the Church acted. On the eve of the Church’s crusade against cousin marriage, what was the anthropological terrain? What was the dominant family type in northwestern Europe? The population history of the region has been marked by a series of major incoming migrations from the east, stratifying the populations in the lead up to the Church’s campaign. Over a small layer of Pleistocene hunter-gatherers sits a larger layer of Near Eastern farmers, which is in turn dwarfed by the Indo-European layer. On top of this basic structure, you have later local population migrations and intermixing. Things are messy. But a close enough approximation of the initial anthropological conditions is available. Much of the contemporary population of northwestern Europe can be traced to the Germanic barbarians of the late Roman period, who in turn, have been phylogenetically traced to a single paleo-deme — the Corded Ware Culture (CWC) in the Third Millenium BC. What was the family structure of this paleo-deme?

The question got a definite answer with the archeological investigations of Haak et al. (2008). They reported excavations from a 4,600 year-old burial site from Eulau in Saxony-Anhalt, Germany. Thirteen individuals, victims of a violent raid, with the bone lesions to show for it, had been buried at the site very carefully. The grave goods were unspectacular, suggesting that they were not elites. The paleo-deme had a definite ritual in their burial practices: ‘During the CWC, the dead are placed in a flexed position, with men’s heads oriented to the west, while the women’s heads are to the east. As a consequence, males are resting on their right sides, females on their left, with both sexes facing south.’

The 13 individuals were buried in four separate sets, each composed of a clear nuclear family. ‘The arrangement of the dead seems to mirror their relations in life. The latter is reflected by the face-to-face arrangement of several pairs of individuals and the positioning of their arms and hands, which are interlinked in several cases.’ DNA sequencing allowed the authors to establish that ‘biological relationships can be inferred from the subtle positioning of the dead within multiple burials’. Specifically, ‘physical closeness of individuals implies biological kinship and a face-to-face orientation indicates a parent offspring relationship’.

Isotope analysis of the tooth enamel yielded further anthropological information: ‘Strontium (Sr) isotopes in tooth enamel reflect dietary strontium derived from soils during childhood and will vary between individuals from distinct geological regions. They can be used to infer the childhood geological location of an individual and hence identify subsequent movement’. Turns out, all children and males were local but the females spent their early lives elsewhere, clearly indicating exogamy and patrilocality.

So the paleo-deme most directly ancestral to the population of northwestern Europe at the time of the Church’s campaign was exogamous, patrilocal and nuclear. Ie, Todd is right. The Western nuclear family is an archaic feature that had been conserved on the northwestern periphery of Eurasia. While the ancient Near East and the Greco-Roman world of Antiquity evolved thicker kinship structures that would show up in the Bible and Church doctrine, the vast populations outside the island of civilization in the central region conserved the archaic nuclear family.

The absolute nuclear family would appear with the top-down spread of primogeniture from the nobles to the commoners. The Church’s war against cousin marriage did not manage to eradicate the practice among nobles — the Hapsburgs were spectacularly inbred as late as the 20th century. Perhaps the Church’s contribution was to contain the practice of endogamy within the nobility? That possibility is consistent with the evidence available as of writing. What can be ruled out is the idea that the Church created the Western nuclear family. It could not have, for the simple reason that the Western nuclear family predates the Church by thousands of years.