GoKawiilThere is a long history of research examining family structures between Late Antiquity and the Early Middle Ages, as well as the influence of Christianity, as key factors in the development of modern European kinship systems30,33,34,35,36. However, sources are mostly normative texts or document elite practices, whereas little is known about non-elite local societies beyond the findings of a small number of genomic studies, all from geographically, chronologically or culturally distinct regions37,38,39,40,41. To evaluate spatial clustering among kin, we analysed grave distribution at Altheim (Supplementary Fig. 14.1) and Büttelborn (Fig. 6). At Altheim, first- to fourth-degree relatives (N = 77–137) were buried significantly closer together than unrelated individuals (n = 3,007; Mann–Whitney U test, two-sided, P < 1.08 × 10−17 to 1.04 × 10−3), with spouses (n = 6) also interred in close proximity (Supplementary Information 14). At Büttelborn, spatial placement was also shaped mainly by close biological relatedness, with distant kin exerting little influence. For example, a father, mother and three children occupy a circular cluster separated from more-distant relatives, such as an aunt and distant cousins (Fig. 6). Assuming that burial proximity reflects social relations, these patterns are consistent with communities being organized around nuclear or stem families, sometimes complemented by extended kin, such as half-siblings. This is reflective of global patterns in agrarian societies, in which small core family units are typically embedded within broader kin networks42, and aligns with historical research on local societies since the early Roman Empire35 and in Southern Germany after about 750 ce (ref. (43).
Fig. 6: Spatial arrangement of reconstructed pedigrees in the Büttelborn cemetery. The alternative text for this image may have been generated using AI. Full size image Colour-coded pedigrees mapped onto individual graves; graves of unrelated individuals are shown in grey. Chronograph estimates of birth and death dates are shown for individuals with a 14C date. A corresponding depiction for Altheim is provided in Supplementary Fig. 14.1. Grave plan courtesy of Thomas Becker, hessenARCHÄOLOGIE, Außenstelle Darmstadt.
Drawing on reconstructed pedigrees from Altheim (Fig. 4) and Büttelborn (Fig. 6, Extended Data Fig. 7 and Supplementary Fig. 12.2), we examined which rules of residence and descent are compatible with such data. Pedigrees more often continued through sons (20 and 9 cases in Altheim and Büttelborn, respectively; Supplementary Table 14.3), although continuation through daughters also occurred (9 and 2 cases, respectively; Supplementary Table 14.3). At Büttelborn, five generations of male relatives (Fig. 6, red pedigree), were buried in a tight cluster. By contrast, much less spatial clustering can be observed for the individuals of a pedigree connecting mostly women over four generations (Fig. 6, orange). Across sites, women share significantly larger IBD segments with individuals buried elsewhere than men (Mann–Whitney U test: two-sided, P < 0.0442; Supplementary Table 10.3 and Supplementary Figs. 10.8–10.11). Within sites, however, they share far fewer IBD segments (>8 cM) in comparison to men (Altheim: 30.92 ± 5.28 versus 170.55 ± 13.08; Mann–Whitney U test: two-sided, P < 1.47 × 10−61; Supplementary Table 10.2), and their relatedness coefficient is 6-fold lower at Altheim and 1.5-fold lower at Büttelborn. These inferred patterns are consistent with a flexible patrilocal system in which most women resided near their husband’s family, whereas some men settled near their wives (Supplementary Information 14), particularly in the absence of a son. However, the high Y-chromosomal and mitochondrial diversity at both sites is incompatible with strict patrilineal descent and strict patrilocality44,45. Instead, it indicates a flexible patrilineal or a bilateral system with flexible patrilocality in which pedigrees were continued mainly through sons but occasionally also through daughters. We further observed that, at least in Altheim, unions were predominantly exogamous when the pedigree continued through sons (16 out of 21 cases), whereas endogamy was more common when continuity was through daughters (5 out of 9 cases), a pattern that echoes those documented in modern matrilocal or matrilineal societies46. Together, patterns of residence and descent in Early Medieval Southern Germany closely mirrored those established during the Late Roman period. Written sources from the Roman Empire show a constant trend towards a bilateral inheritance system: in Roman law, daughters could, under certain circumstances, inherit equally with sons in cases of intestate succession. Yet restrictions on the ability of women to bequeath property, together with other succession rules, generally ensured that assets passed through the male line. From the early centuries ce onward, however, inheritance through daughters was gradually reinforced, a development that continued into Late Antiquity47. Wills sometimes privileged male descendants who received a larger share, especially when land was concerned, but daughters were often treated equally and, in the absence of sons, even became primary beneficiaries (ref. 48, pages 62–71). In the fifth and sixth centuries, both Western Roman and Justinianic law, as well as legal practice in western post-Roman kingdoms, provided further security for the inheritance rights of daughters and their children by will or on intestacy, and thus strengthened bilateral succession (ref. 48, pages 71–72). At the same time, in Late Roman society, family tradition was increasingly conceived of as deriving from both paternal and maternal lines35.
In the Late Roman and post-Roman West, Christian societies increasingly emphasized lifelong monogamy: while divorce faced stricter legal regulation, widowhood was elevated and remarriage was considered morally problematic30,33,36,47. Marriages between close kin were condemned, prompting numerous prohibitions by church councils and secular authorities from around 500 ce onward34. Christianity thus reinforced earlier social trends: literary and epigraphic studies indicate that close-kin marriage was already largely avoided in the pre-Christian Roman West, and lifelong monogamy had become a widely promoted ideal30,34,49. The Lex Baiuvariorum (tit. VII, 1), issued in the eighth century but reflecting earlier norms, confirms strict prohibitions of incest and levirate unions for Bavaria, where historical and archaeological research indicates the continued influence of Christianity after the collapse of Roman rule13. In line with these written sources, our data suggests that lifelong monogamy, with limited divorce or remarriage of widows, was the prevailing norm in sixth century Southern Germany: we identified 68 probable single-partner unions but only five individuals in Altheim and Büttelborn (three men and two women) who had children with multiple partners, and none of the cases is chronologically incompatible with serial monogamy (Supplementary Fig. 3.12). Although we cannot entirely rule out that children from other partnerships either did not exist or were buried elsewhere, the substantial coverage of our sample at Altheim (38%, 90% credible interval: 22–55%; Extended Data Table 1, Supplementary Information 12 and Supplementary Fig. 12.4) makes it unlikely that such a pattern would have systematically escaped detection.
The near absence of long (>12 cM) runs of homozygosity (Supplementary Table 2.13) and the lack of shared IBD segments (>8 cM) between spouses support strict incest avoidance, excluding relationships closer than the sixth degree (Supplementary Table 11.1). Given the small estimated community size (maximum of around 70 individuals per settlement per generation; Extended Data Table 1 and Supplementary Information 16), such avoidance probably encouraged intermarriage across broader social networks. However, a comparable absence of relatedness between spouses was also observed in endogamous unions, that is, among partners from Altheim. Recent genetic research suggests that incest avoidance was also present in non-Christian societies, such as Iron Age populations from the British Isles50, and Avar societies in the Western Pannonian plains37,51. However, the latter show evidence for levirate unions, which are absent in Altheim and Büttelborn, in accordance with the Lex Baiuvariorum.
Overall, Altheim is characterized by patrilocal tendencies and either patrilineal or bilateral descent rules, with pedigrees continuing through daughters more frequently than at previously studied sites (Supplementary Table 14.4), except in a matrilocal Iron Age society in southwest England50. The rate of multiple reproductive unions is low (Supplementary Table 14.4), and only a Neolithic community in north-central France shows a lower rate52. These findings align with written sources and suggest that by the sixth century ce, Central European agrarian societies already maintained residence, and marriage practices that were to become characteristic of Latin Christian Europe.