Families in the Grave: What Ancient DNA Can Really Tell Us About Prehistoric Society

Ancient DNA can now reconstruct pedigrees extending across several generations, identify outsiders who married into communities and reveal whether burial grounds were organised through maternal or paternal descent. Yet a biological family tree is not a social system. Recent discoveries in China and Britain show both the extraordinary power of archaeogenetics and the danger of confusing descent, residence, inheritance and political authority.

At the Neolithic settlement of Fujia in eastern China, two burial grounds lay apart from one another. Archaeologically, that separation was clear enough. Genetically, it proved to be much more consequential. Every one of the fourteen successfully analysed individuals from the northern cemetery carried mitochondrial haplogroup M8a3. In the southern cemetery, forty-four of forty-six belonged instead to D5b1b. Men and women followed the same rule: where a person was buried corresponded overwhelmingly to the maternal line from which that person descended. The paternal picture was almost the reverse. Among thirteen men for whom Y-chromosome lineages could be resolved with sufficient confidence, diversity was high.

This is an unusually clean archaeological pattern. It is also easy to overinterpret.

Fujia, a late Dawenkou culture settlement in Shandong, covered roughly 37 hectares and included residential areas, pottery kilns and more than 500 excavated burials. Radiocarbon modelling places the cemeteries between about 2750 and 2500 BC and suggests that they remained in use for roughly 250 years, perhaps ten generations. The published study obtained authenticated genome-wide data from sixty people: fourteen from the northern cemetery and forty-six from the southern. Some of these people were closely related across the cemetery boundary, so the two groups were not separate biological populations. They appear instead to have belonged to one intermarrying community whose dead were sorted according to maternal ancestry.

That makes Fujia one of the strongest prehistoric cases yet for matrilineal organisation. It does not demonstrate a prehistoric matriarchy.

That distinction is more than terminological. Matrilineal descent means that membership in an enduring social group is reckoned through women. Matrilocal residence concerns where couples live after marriage, usually with or near the woman's family. Neither necessarily tells us who owned land, who controlled stored food, who conducted rituals, who arranged marriages, who exercised political authority or who made decisions within households. A society can be matrilineal without women collectively ruling it, just as a patrilineal society need not exclude women from wealth or authority.

Ancient DNA is becoming good at detecting the first of these phenomena. The others require archaeology.

Two cemeteries, two maternal ancestries

The strength of the Fujia case lies in the conjunction of different kinds of evidence rather than in mitochondrial DNA alone.

The genomic results show extensive biological connections both within and between the cemeteries. Individuals within each burial ground were, on average, more closely related to one another, but relatives also crossed the boundary. The investigators identified second- to sixth-degree connections between the two groups, consistent with repeated intermarriage. Nevertheless, the rule governing burial remained strikingly maternal. Even secondary burial did not erase it. In the southern cemetery, a substantial proportion of the dead had been moved or reburied after decomposition, yet their final placement continued to correspond to maternal lineage.

That last observation deserves attention. Secondary burial involves deliberate treatment of ancestral remains. If bones were transferred after the original funeral and nevertheless returned to the appropriate maternal cemetery, lineage affiliation was not merely an accidental result of households burying their own dead nearby. Someone remembered, or could determine, where these people belonged.

The stable-isotope evidence complicates another tempting interpretation. Strontium ratios from the teeth of twenty individuals were narrowly distributed and compatible with the geological environment around Fujia. Oxygen-isotope evidence likewise suggests broadly shared water sources. There was no statistically significant difference between male and female strontium values. The study therefore found no clear signature of systematic long-distance movement of husbands into Fujia, the pattern one might expect in a straightforward matrilocal marriage system. Instead, Fujia appears to have been relatively endogamous and geographically restricted.

So the safest conclusion is quite specific: maternal descent structured burial affiliation over many generations. Whether married men routinely lived with their wives' maternal relatives is much harder to establish. Nor can the cemetery tell us by itself whether property descended through women or whether senior women exercised unusual political authority.

The chronology matters here. Nineteenth-century writers including Johann Jakob Bachofen and Lewis Henry Morgan proposed sweeping evolutionary schemes in which early societies passed through maternal forms of organisation before patrilineal institutions developed alongside property and political complexity. Modern anthropology long ago dismantled any simple universal progression of that kind. Fujia does not rehabilitate it. Instead, it demonstrates something more interesting: prehistoric kinship systems were genuinely variable.

A second Chinese case makes that point particularly well. At Shimao and related settlements on the Loess Plateau, dating broadly to around 2300–1800 BC, a large archaeogenomic programme reconstructed pedigrees extending up to four generations. Here the dominant pattern was patrilineal. At some cemeteries, high-status male founders were followed by male descendants who also received wealthy graves and, in some instances, sacrificial attendants. The researchers interpret the combination of genealogy, grave goods and burial hierarchy as evidence that status and wealth could pass through male lines. Yet high-ranking women were also present, reminding us again that descent system and gendered political standing are not interchangeable categories.

Within Neolithic China alone, then, archaeogenetics is already revealing incompatible social arrangements rather than a single ancestral model.

When husbands move: Iron Age Dorset

The evidence from Late Iron Age southern Britain concerns a related but different question.

At Winterborne Kingston in Dorset, archaeologists recovered genomic data from fifty-five burials, with forty individuals preserved well enough for detailed analysis of inherited DNA segments. Thirty of those forty had at least one detectable relative of approximately seventh degree or closer. More than two-thirds of the genetically identified members of the core kin group belonged to a rare branch of mitochondrial haplogroup U5b1. Y-chromosome diversity, by contrast, was high.

The pedigree gives this statistical pattern human form. An adult woman, her daughter, adult granddaughters and further descendants were all buried within the community. Men were much more likely than women to lack detectable relatives there. Six people with no identified genetic connection to the main kin group were all male; several nevertheless received ordinary local Durotrigian burial treatment and locally produced pottery. They look less like excluded strangers than integrated members of the community, plausibly men who had married in.

This is much closer to direct evidence for matrilocal residence. Computer simulations comparing the observed distribution of mitochondrial, Y-chromosome, autosomal and X-chromosome relatedness with alternative residence systems fitted male-biased movement much better than patrilocality. The authors estimated female outward migration at close to zero, with substantially greater male movement.

Nor was Winterborne Kingston alone. When the researchers compared mitochondrial diversity across 156 European archaeological sites, the strongest concentrations of repeated maternal lineages occurred disproportionately in British Iron Age communities. Similar patterns appeared at sites in Somerset, Dorset, Oxfordshire, Cornwall and Yorkshire. Genetic relatives detected between different settlements generally did not share mitochondrial lineages, whereas within individual sites maternal lineages were frequently shared. One economical explanation is that men moved between communities whose resident cores were organised around women.

Even here, however, “matrilocal” should not quietly become “matriarchal”. The authors themselves explicitly caution that female-centred residence does not require female political dominance. Classical accounts of powerful British women, and the relatively rich furnishing of some female Durotrigian graves, are relevant contextual evidence, but they cannot turn a residence pattern into a constitutional system. Roman observers also had their own expectations about supposedly exotic northern societies.

What the DNA does establish is that the long-standing European assumption of predominantly patrilocal prehistoric communities cannot simply be extended into Iron Age Britain.

A pedigree is not a family

The most useful lesson from archaeogenetics may come from cases in which biological and social relationships fail to coincide neatly.

At Hazleton North, an Early Neolithic chambered tomb in Gloucestershire, DNA from thirty-five individuals allowed twenty-seven of them to be placed within a five-generation pedigree. Every identified transmission connecting successive generations of the principal lineage passed through men. Adult daughters of lineage males were absent, while women who reproduced with those men were present, consistent with patrilineal descent and female exogamy.

Yet the tomb was not simply a container for biological descendants. Four males had mothers who reproduced with members of the principal lineage but fathers outside it. They were nevertheless incorporated into the lineage burial structure. Another eight people were not close biological relatives of the central pedigree at all. The authors therefore raised the possibility that social kinship extended beyond genetic descent.

This is exactly where archaeological interpretation must resist the seductive precision of a reconstructed family tree.

Geneticists can determine with considerable confidence that two people were parent and child, siblings or more distant biological relatives. They cannot read the vocabulary by which those people understood the relationship. A genetically unrelated man might have been an adopted son, foster child, dependant, ritual relative, household member, political ally or something for which modern European kinship terminology has no good equivalent.

The large Neolithic cemetery of Gurgy “les Noisats” in France makes the same point differently. Researchers obtained genome-wide data from ninety-four individuals and reconstructed two principal pedigrees, the larger connecting sixty-four people across seven generations. The burial population was strongly patrilineal. Adult sons remained prominent in the cemetery, whereas adult daughters were strikingly scarce: twenty-seven adult male lineage offspring compared with only six females. Strontium values from several unrelated or genealogically external women differed from those of local males, consistent with women arriving from elsewhere.

That is persuasive evidence for female exogamy. It is not perfect evidence for a rigid rule. Several adult lineage daughters did remain at Gurgy, and the burial population contained fewer adult women overall than the reconstructed reproductive relationships imply should have existed. Some women may therefore have received different funerary treatment or been buried elsewhere. The researchers explicitly caution that a cemetery is a selected population and need not mirror the living community.

This problem runs through every attempt to reconstruct prehistoric society from graves: we study the people whom a community chose to bury in an archaeologically visible way. We do not automatically have its entire population.

From marriage to inheritance

Inheritance is harder still, because DNA carries genes, not property rights.

Occasionally, however, biological pedigree and archaeological wealth line up strongly enough to make hereditary status a reasonable interpretation. In the Early Bronze Age Lech Valley of southern Germany, genomic, isotopic and archaeological analysis found multigenerational high-status core families alongside biologically unrelated people whose graves were considerably poorer. The association between lineage membership and rich burial from childhood onwards has been interpreted as evidence that social position could be inherited rather than acquired solely during an individual's lifetime.

At Chaco Canyon in the American Southwest, nine people buried in the unusually rich Room 33 crypt at Pueblo Bonito shared an identical mitochondrial genome. Nuclear DNA from the best-preserved individuals demonstrated mother–daughter and grandmother–grandson relationships, establishing a maternal lineage that persisted in this elite mortuary setting from about AD 800 to 1130. That is strong evidence that membership in an elite descent group was transmitted through women. Whether the lineage supplied political leaders in every generation remains a further inference, because the genetic data cannot identify offices or titles.

The same caution applies to the spectacular Hallstatt elite burials of southern Germany. Genome and isotope analysis has identified biological connections between wealthy individuals buried as much as 100 kilometres apart. A probable uncle–nephew relationship between two exceptionally rich burials is compatible with dynastic succession through a maternal line: a man's sister's son can succeed him without father-to-son inheritance. But “compatible with” matters. The pedigree provides a plausible institutional mechanism; it does not preserve the rules of succession themselves.

DNA becomes evidence for inheritance only when archaeological status markers follow genealogical lines in repeated and chronologically convincing ways.

No prehistoric family system

The accumulating evidence is making one old question increasingly difficult to ask: What was the prehistoric family like?

There was no single prehistoric family.

Around 4850–4500 BC at Gurgy, a patrilineal community incorporated women from elsewhere. Around 2750–2500 BC at Fujia, two maternal clans maintained separate burial identities while intermarrying within a geographically restricted community. Later at Shimao, powerful male pedigrees intersected with increasingly formal social hierarchy. In Iron Age Britain, several communities appear to have retained women and received men. None of these arrangements can safely be treated as an evolutionary stage through which all societies passed.

Indeed, a 2025 study of Neolithic and Copper Age communities in the Carpathian Basin found substantially different kinship organisations among broadly contemporary populations separated by only about 100 kilometres and sharing much of their material-cultural world. One community was relatively closed, with numerous biological connections and indications of patrilineality; another was much more genetically diverse and connected to outside populations. Similar pottery did not entail similar marriage networks.

That finding may ultimately be as important as any spectacular seven-generation pedigree. Archaeological cultures are analytical constructions based largely on material similarities. They should not automatically be converted into tribes, peoples or societies with uniform marriage rules.

Ancient DNA has therefore changed the scale at which kinship can be studied. Where archaeologists once inferred household organisation indirectly from grave clustering, architecture, sex ratios or artefact distributions, they can now sometimes identify a grandmother, distinguish siblings from parent and child, recognise the arrival of spouses, or follow one descent line for centuries.

But the method becomes most archaeologically useful precisely where its limits are respected.

DNA can reconstruct biological relatedness. Repeated biological relationships, spatial organisation and sex-specific mobility can reveal probable descent and residence systems. Isotopes can test whether individuals grew up locally. Grave goods and burial treatment may indicate differences of status. When these patterns coincide over generations, hereditary privilege becomes a reasonable hypothesis.

None of them, alone, identifies marriage law, household authority, land ownership, emotional bonds, adoption, ritual kinship or political power.

Fujia therefore deserves attention not because it proves that Neolithic China was once matriarchal, but because it demonstrates something more securely archaeological. For roughly ten generations, people belonging to one community appear to have remembered two maternal ancestries so consistently that even the dead who were moved after burial were ultimately placed with the appropriate maternal group. What those clans owned, how marriages were negotiated between them, who exercised authority within them, and whether women controlled resources remain unanswered questions.

The family tree is becoming visible. The society around it still has to be excavated.

Selected Sources and Further Reading

1. Wang, J., Yan, S., Li, Z. et al. 2025. “Ancient DNA reveals a two-clanned matrilineal community in Neolithic China.” Nature 643, 1304–1311. DOI: 10.1038/s41586-025-09103-x.
2. Cassidy, L. M., Russell, M., Smith, M. et al. 2025. “Continental influx and pervasive matrilocality in Iron Age Britain.” Nature 637, 1136–1142. DOI: 10.1038/s41586-024-08409-6.
3. Rivollat, M. et al. 2023. “Extensive pedigrees reveal the social organization of a Neolithic community.” Nature 620, 600–606. DOI: 10.1038/s41586-023-06350-8.
4. Fowler, C., Olalde, I., Cummings, V. et al. 2022. “A high-resolution picture of kinship practices in an Early Neolithic tomb.” Nature 601, 584–587. DOI: 10.1038/s41586-021-04241-4.
5. Chen, Z., Gardner, J. D., Sun, Z. et al. 2025. “Ancient DNA from Shimao city records kinship practices in Neolithic China.” Nature 648, 659–667. DOI: 10.1038/s41586-025-09799-x.
6. Mittnik, A., Massy, K., Knipper, C. et al. 2019. “Kinship-based social inequality in Bronze Age Europe.” Science 366, 731–734. DOI: 10.1126/science.aax6219.
7. Kennett, D. J., Plog, S., George, R. J. et al. 2017. “Archaeogenomic evidence reveals prehistoric matrilineal dynasty.” Nature Communications 8, 14115. DOI: 10.1038/ncomms14115.
8. Szécsényi-Nagy, A. et al. 2025. “Ancient DNA reveals diverse community organizations in the 5th millennium BCE Carpathian Basin.” Nature Communications. DOI: 10.1038/s41467-025-60368-2.

Suggested keywords

ancient DNA; archaeogenetics; prehistoric kinship; matrilineality; matrilocality; patrilineality; marriage networks; prehistoric families; burial archaeology; inheritance; Fujia; Iron Age Britain

Editorial summary: Dense archaeogenomic sampling is moving archaeology beyond population migrations toward the reconstruction of individual pedigrees and community organisation, but the newest results also make it increasingly important to distinguish biological descent from residence, inheritance and political power.