Evolutionary Insights by Anthropology.net

In the oldest levels of a neighborhood called DK-G South, the houses are large. The largest date to around 2500 BC and cover more than 160 square meters of floor plan. This isn’t surprising for a Bronze Age city. What’s surprising is what happens next.

Over the following four centuries, as Mohenjo-daro continued to grow and fill in, housing inequality fell. Not marginally. By around 2100 BC, the Gini coefficient of residential disparity in DK-G South had dropped to 0.23, a figure typical of Neolithic farming villages, not major urban centers. The gap between the largest and smallest homes had narrowed to something archaeologists usually associate with societies that don’t yet have cities.

Mohenjo-daro had tens of thousands of residents. It had standardized brick construction, a citywide drainage network, a grid of streets, a sophisticated system of weights and measures, and a vigorous long-distance trade network that extended to Mesopotamia. It was, by any reasonable measure, one of the most functionally complex urban environments on earth in the third millennium BC. And it appears to have been getting more equal as it matured.

That finding is the core of a new study by Adam Green, Iqtedar Alam, and Cameron Petrie, published in Antiquity in 2026. Their analysis draws on excavation data from the earliest twentieth-century digs at the site, converting house footprints into Gini coefficients through a GIS-based approach. The Gini coefficient is a standard economic measure of distributional inequality: zero means total equality, one means total concentration. Their results put the city’s overall residential Gini at 0.44, well below contemporaneous Mesopotamian cities like Ur and Ugarit, which both score above 0.6. The ancient Greek site of Knossos sits at 0.86. The Classic Maya city of Palenque at 0.75.

Mohenjo-daro was already an outlier by those comparisons. But the diachronic data from DK-G South is what makes the study genuinely provocative.

What the houses say

The standard archaeological assumption is that cities generate inequality. The logic is intuitive: larger populations mean more specialization, more specialization means more occupational hierarchy, and hierarchy means wealth concentrates at the top. V. Gordon Childe codified this in his 1950 “urban revolution” framework, which treated social stratification as a defining feature of cities. Neoevolutionary theory went further, treating the state as essentially defined by stratification. If you had a city, you had elites; if you couldn’t find the elites, you weren’t looking hard enough.

Indus archaeologists spent much of the later twentieth century fighting this prior. Mohenjo-daro and the other major Indus sites lack the usual material signatures of hierarchical power: no palaces, no royal tombs, no temples accessible only to a priestly class, no statues of rulers. The arguments went back and forth. Some scholars concluded the Indus must have been a complex chiefdom or non-state. Others countered that a civilization that built cities at this scale must have had elites, even if the evidence was missing. Absence of evidence wasn’t evidence of absence, or so the argument ran.

What Green and colleagues have added is a quantitative dimension to a debate that had stayed largely qualitative. And the quantitative data doesn’t just confirm the impression that Mohenjo-daro was relatively equal. It shows the city moving in a direction opposite to what the theory predicts.

The method has known limitations, which the authors are explicit about. Gini coefficients of house size are not a comprehensive picture of social inequality. They don’t capture differences in movable wealth, in gender or kinship status, or in access to political power that doesn’t manifest architecturally. A wealthy person might not invest in a larger house. A small structure might house a powerful family. The measure also struggles with people who have no residence at all, the poorest members of any ancient city. These caveats matter.

But the measure does capture one real economic dimension: access to labor and resources at the household level. And it allows direct comparison across sites and time periods, which is why the Global Dynamics of Inequality (GINI) project has assembled this kind of data from hundreds of ancient settlements worldwide. Mohenjo-daro now sits in that database, and what it shows is distinctive.

The governance question

The DK-G South sequence tracks five periods from roughly 2500 to 2100 BC. The earliest period shows a Gini of 0.39 and a median house size of 161 square meters. By the latest period, the Gini is 0.23 and the median has dropped to 141 square meters, but with much less variance. The log standard deviation of house areas falls from 0.73 to 0.44 over the same span. Houses weren’t just smaller on average; they were converging toward a common size.

This convergence happened alongside something else: the city’s residences became more uniformly aligned with Mohenjo-daro’s street grid. In the earlier levels, structures are arranged more loosely. In the later levels, they lock into the urban plan. Green and colleagues read this correlation as suggestive of governance, that the same collective decisions that built and maintained public infrastructure also shaped residential space in ways that narrowed the gap between households.

What that governance looked like in practice is uncertain, and the study doesn’t overclaim. There are no deciphered texts from the Indus Civilization, no royal inscriptions, no administrative archives of the kind that survive from Mesopotamia. But the material record offers indirect evidence for distributed authority. Indus seals, small stone stamps that likely facilitated exchange and credit, were found primarily in private residences at Mohenjo-daro rather than in temples or central administrative buildings. In Mesopotamia, equivalent objects concentrated in institutional contexts, where a palace or temple controlled trade. At Mohenjo-daro, the tools of economic governance were spread across the city’s households. Green has previously characterized this contrast in terms of where the power to govern exchange sat: distributed rather than monopolized, and for that reason probably harder to weaponize against ordinary residents.

The same logic extends to the standardized weights and measures used across the entire Indus region. This wasn’t a system enforced from a single administrative center. It was a shared protocol adopted by communities across a vast area, a form of coordination without centralized control.

Green and colleagues point to the decline in residential disparity coinciding with more structured urban development, and suggest that the same governance producing civic amenities may also have constrained how unevenly residential space was distributed. Green has described Mohenjo-daro as having had “deliberative spaces” where collective decisions about the city were made. The new Gini data, tentatively, supports something like that interpretation. The declining inequality wasn’t random drift. It tracked with urban development in a way consistent with intentional decisions about how the city was organized.

One further pattern in the data is worth noting carefully. As inequality declined, productivity appears to have risen. The GINI project uses mean residence area as a rough proxy for household wealth and productivity. In DK-G South, median residence area bottomed out at about 98 square meters in the Late III Period, around 2200 BC, as the Gini was still falling. Then, in the latest period, when the Gini reaches its low of 0.23, median residence area rebounds to 141 square meters. Gregory Possehl had already noted, separately, that there is more evidence for craft production in the city’s later levels. The authors are cautious here, and rightly so: the sample sizes are small and the relationship isn’t statistically robust. But the pattern raises a pointed question about causality.

Development economics since Kuznets has generally assumed that growth generates inequality, and that only after growth reaches a certain threshold does redistribution become viable. Thomas Piketty’s analysis of capitalist economies complicates this picture but doesn’t overturn the basic direction of the arrow. The Mohenjo-daro data, if interpreted at face value, runs that arrow the other way: collective governance constrained inequality, and declining inequality may have contributed to a subsequent rise in productivity. The city’s most equal period appears to be its most productive.

That is a speculative reading, and the paper says so. But it’s the kind of speculation that makes a dataset useful beyond its immediate empirical scope. Mohenjo-daro isn’t a policy prescription. It is a natural experiment, running across four centuries, that complicates the idea that inequality is an unavoidable cost of urban complexity.

The site still holds a great many unanswered questions. The chronological resolution outside DK-G South is poor; most other excavated areas collapse multiple building phases into a single unit, which almost certainly inflates the overall Gini by mixing large early structures with smaller later ones. Future work revisiting structural chronologies from the rest of the site could sharpen the picture considerably. Projects like M-LAB at the Pratt Institute are already working to reinterpret Mohenjo-daro’s legacy datasets with more rigorous methods.

For now, the numbers from DK-G South say something that’s hard to dismiss: as this Bronze Age city built itself into maturity, its wealthiest households and its poorest ones were moving toward each other. That doesn’t happen by accident.

Further Reading

* Childe, V.G. 1950. The urban revolution. The Town Planning Review 21: 3–17. https://doi.org/10.3828/tpr.21.1.k853061t614q42qh

* Green, A.S. 2020. Debt and inequality: comparing the ‘means of specification’ in the early cities of Mesopotamia and the Indus civilization. Journal of Anthropological Archaeology 60. https://doi.org/10.1016/j.jaa.2020.101232

* Green, A.S. 2022. Of revenue without rulers: public goods in the egalitarian cities of the Indus civilization. Frontiers in Political Science 4. https://doi.org/10.3389/fpos.2022.823071

* Green, A.S. et al. 2025. Kuznets’ tides: an archaeological perspective on the long-term dynamics of sustainable development. Proceedings of the National Academy of Sciences USA 122. https://doi.org/10.1073/pnas.2400603121

* Kohler, T.A. et al. 2025. Introducing the special feature on housing differences and inequality over the very long term. Proceedings of the National Academy of Sciences USA 122. https://doi.org/10.1073/pnas.2401989122

* Kuznets, S. 1955. Economic growth and income inequality. The American Economic Review 45: 1–28.

* Piketty, T. 2014. Capital in the Twenty-First Century. Cambridge, MA: Harvard University Press.

* Possehl, G.L. 2002. The Indus Civilization: A Contemporary Perspective. Walnut Creek, CA: AltaMira.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

Jar 1 at Site 75 sat in forest roughly 70 kilometers northeast of Phonsavan, on the Xieng Khouang Plateau in northern Laos. It was already in poor condition when researchers found it: the sides partially collapsed, the interior open to the elements. But the base was intact, and within the remaining walls, the original sediment deposits had been preserved. Over three field seasons between 2022 and 2024, a team led by Nicholas Skopal of James Cook University excavated it completely.

What they found was dense. Bones packed without obvious vertical structure, skulls clustered toward the jar’s edges, long bones bundled together. When the osteologists worked through the assemblage, counting from the most commonly repeated skeletal elements, they arrived at a minimum of 37 individuals. The highest count came from dentition: 469 observed teeth, producing an MNI of 37. The youngest was approximately 18 months old. Adults of varying ages were present throughout.

“The big jar we’ve found is unique,” Skopal told reporters, “and I’ve seen a lot of jars.”

That uniqueness matters. Hundreds of large stone vessels are scattered across more than 120 known sites on the Xieng Khouang Plateau. The most studied concentration is near Phonsavan, the so-called Plain of Jars, where some jars reach three meters in height and weigh several tons. The first systematic survey was done by French archaeologist Madeleine Colani in the 1930s. She rejected the idea that the vessels were storage containers and proposed a funerary role. A local legend holds that giants used them to brew rice wine. Scholarship since Colani has leaned toward mortuary use, but the physical evidence has been thin. A few jars contained ash and burned bone fragments. Some had burial pits dug nearby. Most were empty, which made the funerary hypothesis difficult to confirm with any confidence. If each jar was a tomb, why did so few of them hold human remains?

Jar 1 makes that question answerable, and the answer involves movement.

A Multistep Death

The critical detail about the bones inside Jar 1 is their disarticulation. Most joints had already separated before the remains were placed inside. The bones had not come from bodies laid directly into the jar. They came from bodies that had decomposed somewhere else first. This is secondary burial: the deliberate reinterment of already-processed skeletal material after an initial phase of decomposition away from the final deposit.

The paper’s most consequential observation follows from this. About 500 meters west of Jar 1, the team found a second cluster of seven smaller stone jars. None of them contained human remains. The ceramics found near these smaller vessels are consistent with the assemblage from Jar 1’s surrounding context, suggesting the two locations were used at the same time by the same community. The working hypothesis is that the smaller jars served as the initial stage: a newly dead person would have been placed in one, the flesh allowed to separate from the bone, and then the cleaned skeletal remains transferred to the large jar.

Skopal described the logic to Science News:

“Maybe they used those [smaller] stone jars to ‘distill’ the bodies, so when someone died, they might have put the body in there so all the flesh came off. Then they took the bones and they put them in this big jar, so it’s almost like a crypt.”

But the process may not have ended there. The team noted a discrepancy between the minimum individual count from dental remains and the lower count derived from cranial and postcranial elements alone. Bones appear to be missing. Some may have been selectively removed from Jar 1 after deposition, possibly for reinterment at a habitation site or place of worship. Stone slabs found inside the jar are comparable to those used as markers for secondary bundle burials at other Plain of Jars sites. Comparable practices of secondary processing and relocation of skeletal remains are documented across northeast India, Myanmar, Thailand, and Vietnam during the Iron Age and Historical periods. The large jar, on this reading, was not the end of the journey. It was a stage in an ongoing process.

The radiocarbon dates establish how long that process ran. Eight bone and tooth samples returned results spanning approximately cal AD 890 to 1160, a range of roughly 270 years. Charcoal from pottery deposited in an adjacent trench dated to cal AD 890 to 1020, consistent with the early end of the skeletal dates. The jar was not filled in a single event. Generations of a family, or an extended kin group, returned to it repeatedly.

“The number of individuals also suggests the jars were owned by family or extended family groups,” Skopal noted in the press release accompanying publication. “They likely served as places where ancestral rites were performed over generations.”

Four skulls were found with their mandibles still in close association, a small departure from the general pattern of disarticulation. The jaw is typically among the first elements to separate during decomposition. Their preservation together raises the possibility that in some cases, a recently dead or only partially processed individual was placed inside the jar directly, perhaps with greater ceremony. The osteological analysis is ongoing.

One other detail is worth holding onto. At least two individuals show deliberate tooth ablation: the intentional removal of specific teeth during life. One person lost both upper lateral incisors; another lost both lower central incisors. Tooth ablation is documented widely across prehistoric and historical mainland Southeast Asia and southern China, where it appears associated with social identity and life transitions. Its presence at Site 75 places this community within a broad pattern of shared bodily practice across the region.

What the Beads Know

Twenty glass beads came out of the jar during wet-sieving of the sediments. Orange, turquoise-blue, dark blue, black, red-over-orange. Twelve were selected for chemical analysis by laser ablation mass spectrometry at the Field Museum in Chicago, and the results upend any assumption that the Xieng Khouang Plateau was a remote backwater.

Nine of the beads matched a mineral soda-high alumina composition produced in eastern South India, a glass type that circulated across South and Southeast Asia from roughly the end of the first millennium BC through the eleventh century AD. Two beads had a soda plant-ash composition whose chemistry aligns most closely with Mesopotamian glass, the type that appears at Southeast Asian sites dating from the eighth or ninth century through the eleventh century AD, before later sites begin yielding Egyptian glass instead. The team notes that the temporal overlap between the Mesopotamian glass signature and the radiocarbon dates from the jar is consistent. One dark-blue bead had a potash-rich composition more commonly dated to 300 BC through AD 300, most likely from southeastern China or northern Vietnam. The team treats it as a curated antique, a much older object that had been kept, traded, or passed down.

No glassmaking tradition is known from the Lao highlands. These beads arrived from elsewhere, through exchange networks that extended from the Laotian interior to the Indian subcontinent and the Near East. They are not rare luxury goods in the way that term sometimes gets misused in archaeology; they are trade goods, the kind of thing that moved in bulk along commercial routes. Their presence in a burial jar in upland Laos means the people using that jar were connected to those routes.

The timing sharpens the point. The period cal AD 890 to 1160 sits within an era of expanding commercial connectivity across Asia. The Song Dynasty in China was actively intensifying long-distance trade. The Khmer Empire was at its height. The Dali Kingdom in Yunnan extended into parts of northern Laos. The Đại Việt and Champa kingdoms controlled Vietnam. The Pagan Kingdom held Myanmar. All of these states were embedded in overland and maritime exchange networks that moved goods, technologies, and cultural practices across vast distances. The team proposes that the jar sites may have sat at nodal points along routes connecting East and Southeast Asia, the plateau elevated enough to serve as a waypoint, the communities there integrated into the wider system. The Mesopotamian beads in Jar 1 are physical evidence for the highland end of that connectivity.

Miriam Stark, an anthropologist at the University of Hawaii not involved in the study, told Science News that the collective mortuary assemblage was exactly what she had hoped to see discovered. But she pressed on the gap that the data still cannot close:

“I do wonder, where did these people live?”

No settlements associated with the Plain of Jars tradition have ever been found. No houses, hearths, or refuse deposits. The entire archaeological record of these communities consists, almost entirely, of what they did with their dead. The jar sites are the only material evidence of who they were, and most of those sites are empty.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

Somewhere in the Altai Mountains roughly 59,000 years ago, a Homo neanderthalensis sat still while someone drilled into their tooth with a pointed piece of jasper.

We know this because the tooth survived. It came out of Chagyrskaya Cave in southwestern Siberia as specimen Chagyrskaya 64, a lower second molar from an adult individual, and when paleoanthropologist Alisa Zubova of the Russian Academy of Sciences and her colleagues examined it, something was immediately wrong. There was a hole in it. Not a small one — a large, irregular concavity measuring 4.2 mm long, 2.8 mm wide, and 2.6 mm deep, positioned at the center of the occlusal surface and reaching all the way down to the floor of the pulp chamber. Around the edges, visible at magnifications up to 500x, were fine linear striations. The kind left by a rotating tool.

“We were intrigued by the unusual shape of the concavity on the tooth’s chewing surface,” said Zubova in a statement accompanying the study. “It differed from the normal morphology of the pulp chamber and did not match the typical pattern of carious lesions seen in Homo sapiens. Moreover, distinctly visible scratches suggested that the concavity was not the result of natural damage but of intentional actions.”

Their analysis, published May 13, 2026 in PLOS One, presents a compelling case that Chagyrskaya 64 carries the oldest known evidence of deliberate dental treatment in human evolutionary history — predating the previous record holder by more than 40,000 years.

What the Tooth Actually Shows

Before accepting the drilling interpretation, the team had to work through the alternatives. A deep concavity in an old molar could plausibly result from severe wear, a fracture, or natural carious decay. None of these fit.

Dental trauma leaves sharp, fractured edges. The concavity walls in Chagyrskaya 64 are smooth and rounded. Extreme attrition can expose the pulp chamber, but it cannot widen it — and the upper portion of the concavity is broader than the pulp chamber below, which rules out wear as the cause. The comparison sample from the same stratigraphic layer, another Chagyrskaya molar recovered from identical sediment, shows intense occlusal wear but a completely flat chewing surface, no concavity, and uniformly mineralized dentin.

Micro-CT imaging told a different story for Chagyrskaya 64. The tooth showed pervasive demineralization of primary dentin, corresponding to Grade 4 or 5 on the Downer diagnostic scale — severe, deep caries extending into both inner and outer dentin. The absence of secondary dentin in a heavily worn tooth is itself anomalous; normally, the pulp responds to attrition by laying down reparative tissue. Here it hadn’t, because the infection had already destroyed it. There was no secondary dentin to form.

The morphology of the concavity also doesn’t match anything caries produce naturally. Bacterial decay works by chemical demineralization, creating soft tissue breakdown and irregular cavities. It does not excavate clean, sub-rounded depressions with parallel microstriations on the walls. Those striations, analyzed under scanning electron microscopy, are oriented parallel to the concavity walls and exhibit a corrugated, ridged structure consistent with a rotating tool edge. They appear in at least two discrete locations, at slightly different depths, suggesting the tool was worked from multiple angles.

And critically: over the striations, there is polish. Ante-mortem wear, from years of chewing after the procedure was done.

Drilling in the Dark

To test whether a jasper perforator could actually produce this kind of concavity, Zubova and colleagues ran a series of experiments on three modern Homo sapiens molars. Two came from undocumented Holocene archaeological collections. The third — an upper left third molar with an untreated cavity and no previous dental work — was contributed by one of the authors. The tools were fabricated from local jasperoid raw material, matching what was available at Chagyrskaya Cave.

The experiments were conducted by a single researcher experienced in Paleolithic stone knapping, with a small amount of water added to simulate the oral environment. The technique that worked was manual rotation — holding the pointed perforator between two fingers and drilling with a twisting motion, not scraping. Scraping spread force too broadly and produced only surface striations. The rotational method disrupted the tooth surface within seconds.

Penetrating through to the pulp chamber took between 35 and 50 minutes of continuous drilling. The research team notes this is probably a conservative estimate for the original procedure. Working inside a living person’s mouth, with limited visibility, constrained access, and an uncooperative patient, could easily double the time.

Neanderthals have thinner enamel than Homo sapiens, distributed over a larger volume of coronal dentin. The infection in Chagyrskaya 64 had already softened the dentin considerably. Both factors likely made the drilling faster. Still: somewhere between one and two hours of someone boring into your tooth with a sharp rock.

Raman spectroscopy of the tooth surface detected no residue of plant material, resins, or analgesics. The team notes this doesn’t rule out their use — if the individual lived and chewed on the tooth for years afterward, any herbal packing would long since have worn away. But it leaves open the possibility that no painkiller was used at all.

The experimental results matched the archaeological tooth closely. The groove profiles on both share a similar V-shaped morphology, with gently sloping outer walls. The width of the experimentally produced grooves (0.249 mm) closely approximates what’s preserved on Chagyrskaya 64 (0.294 mm). Macrotraces are more pronounced in the experimental specimens — expected, given that they hadn’t then spent 59,000 years in sediment with subsequent wear on top. Microtraces on the original molar are accordingly faint, but present.

The concavity itself is not a single depression. Micro-CT shows three partially overlapping depressions that together occupy the entire pulp chamber volume. Whether they were drilled in separate episodes or represent the practitioner widening the cavity as they worked isn’t clear from the preservation. The team’s experimental replication of a three-depression configuration suggests the smallest of the three was probably an early stage, with the two deeper ones expanded after initial penetration. One possibility is that the third depression targeted a separate area of carious tissue. Another is that it was an error — stone tools aren’t precision instruments.

The Broader Picture

Caries are rare among Neanderthals. Fewer than ten cases have been identified across the entire fossil record, from sites including Palomas, Bau de l’Aubesier, De Nadale, and Kebara. All the previously known cases involve minor lesions confined to enamel or upper dentin, none reaching the pulp chamber. Chagyrskaya 64 is in a class by itself.

The Chagyrskaya population appears to have been particularly susceptible. Another tooth from the same cave, Chagyrskaya 18, is a naturally shed deciduous molar from a Neanderthal child of about 9 to 11 years. It carries two carious lesions. The smaller of the two has already breached through the enamel into the upper dentin. The larger is still confined to enamel but actively demineralizing. Two caries cases in a small population is unusual for Neanderthals; the team suggests it may indicate the presence of specific cariogenic bacteria in the local oral microbiome rather than a dietary shift, since isotopic evidence from Chagyrskaya places these individuals squarely within the dietary range of other European Neanderthals.

Chagyrskaya 64 also carries a second ante-mortem modification: a pronounced interproximal groove on its distal surface, with microscopic parallel striations characteristic of Stage 4 toothpick use-wear. A nascent groove is also forming on the mesial surface. The same tooth was toothpicked on both sides and then drilled. The toothpick groove sits precisely where demineralization is concentrated in the cervical area — over a second, smaller carious lesion. The drilling concavity sits over the larger lesion on the occlusal surface. Two interventions, two lesions, two tools.

Toothpicking behavior predates H. neanderthalensis considerably. Interproximal wear grooves appear on teeth attributed to Homo habilis from Olduvai Gorge, and the same wear pattern has been documented in Japanese macaques (Macaca fuscata). The behavior itself isn’t cognitively remarkable. But drilling is something else. The rotational technique requires coordinated fine motor control, sustained application of force in a confined space, and — given the documented grain of the striations — deliberate adjustment of the tool’s angle during the procedure. The team notes that analysis of bone retouchers from Chagyrskaya Cave shows Neanderthals there typically held tools between two fingers during knapping, a technique already requiring significant digital precision.

The previous oldest evidence of dental intervention belongs to an individual from Ripari Villabruna in northeastern Italy, dated to around 14,000 years ago. That tooth, a third molar from a late Upper Paleolithic H. sapiens, shows traces of enamel scraping around a carious lesion. Micro-CT of the Villabruna tooth shows substantial retention of demineralized tissue — the scraping was superficial, limited largely to the enamel surface. It would not have relieved the pain of deep infection. Chagyrskaya 64 represents not just an earlier intervention but a more effective one, by a population whose tool-use repertoire apparently included the insight that drilling into the pulp chamber — deliberately and completely — would stop the pain by destroying the nerve.

Whether this constitutes knowledge in any formal sense, or something closer to problem-solving arrived at under extreme duress, is a question the tooth cannot answer. What it does answer is whether the capacity existed. The individual who drilled Chagyrskaya 64 identified a source of pain, selected an appropriate tool, executed a controlled invasive procedure, and their patient then went on to chew with the treated tooth for long enough to wear it smooth.

“What struck me, and continues to strike me, is what an incredibly strong-willed person this Neanderthal must have been,” said Lydia Zotkina, co-author of the study. “Now, every time I go to the dentist, I think about that guy.”

Further Reading

* Oxilia G, Peresani M, Romandini M, Matteucci C, Spiteri CD, Henry AG, et al. Earliest evidence of dental caries manipulation in the Late Upper Palaeolithic. Scientific Reports. 2015;5:12150. https://doi.org/10.1038/srep12150

* Kolobova KA, Roberts RG, Chabai VP, Jacobs Z, Krajcarz MT, Shalagina AV, et al. Archaeological evidence for two separate dispersals of Neanderthals into southern Siberia. PNAS. 2020;117(6):2879–85. https://doi.org/10.1073/pnas.1918047117

* Mafessoni F, Grote S, de Filippo C, Slon V, Kolobova KA, Viola B, et al. A high-coverage Neandertal genome from Chagyrskaya Cave. PNAS. 2020;117(26):15132–6. https://doi.org/10.1073/pnas.2004944117

* Skov L, Peyrégne S, Popli D, Iasi LNM, Devièse T, Slon V, et al. Genetic insights into the social organization of Neanderthals. Nature. 2022;610(7932):519–25. https://doi.org/10.1038/s41586-022-05283-y

* Faux P, Ding L, Ramirez-Aristeguieta LM, et al. Neanderthal introgression in SCN9A impacts mechanical pain sensitivity. Communications Biology. 2023;6(1):958. https://doi.org/10.1038/s42003-023-05286-z

* Estalrrich A, Alarcón JA, Rosas A. Evidence of toothpick groove formation in Neandertal anterior and posterior teeth. American Journal of Physical Anthropology. 2017;162(4):747–56. https://doi.org/10.1002/ajpa.23166

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

A tooth recovered from Zhoukoudian cave near Beijing in the early 1950s has been sitting in storage for decades. It belongs to a Homo erectus individual who died roughly 420,000 years ago, during a period when the world held at least four distinct human lineages simultaneously. H. sapiens had not yet emerged. Neanderthals were consolidating in the west. A poorly understood group we now call Denisovans was somewhere in Asia. And H. erectus -- the oldest hominin to have left Africa, a species that had already been roaming across three continents for well over a million years -- was still out there, still occupying the limestone hills north of what is now Beijing.

That tooth, and five others like it from two additional sites across China, have just yielded the first informative molecular data ever recovered from Homo erectus. The results, published this month in Nature by a team led by paleogeneticist Qiaomei Fu of the Chinese Academy of Sciences’ Institute of Vertebrate Paleontology and Paleoanthropology, connect these 400,000-year-old individuals to the Denisovans -- and, through them, to people alive today.

The breakthrough didn’t come from DNA. It couldn’t. Ancient DNA degrades quickly; even under ideal preservation conditions, retrievable sequences rarely survive beyond a few hundred thousand years, and the specimens from Zhoukoudian, Hexian in Anhui Province, and Sunjiadong in Henan Province are far older than what DNA analysis can reach. Instead, Fu’s team turned to proteins.

Proteins survive longer than DNA because they are structurally sturdier, and because tooth enamel -- the hardest tissue in the vertebrate body -- protects them especially well. Enamel is a mineral matrix, and proteins trapped inside it can persist for millions of years in fragmentary form. They are not genetic material themselves, but they are produced by genes and carry the signature of the genetic variants that shaped them. As University of Copenhagen biochemist Enrico Cappellini, who was not involved in the study, put it in comments to Science: proteins are,

“a direct expression of genes, so in a sense they carry genetic information. It’s the best proxy for DNA.”

Retrieving those proteins from a prized and irreplaceable fossil is a different kind of problem. Most paleoproteomic studies drill or cut into bone or tooth to extract a powdered sample -- destructive work that curators of important specimens understandably resist. Fu’s team used a method called acid etching: the tooth is wrapped in a waterproof film, a patch of enamel just a few millimeters square is exposed, then briefly treated with acid. The technique removes just enough material for analysis while leaving the specimen physically intact except for slight surface discoloration. The teeth from Zhoukoudian, Hexian, and Sunjiadong emerged with their morphology preserved.

What came out of that process was a catalog of 11 ancient enamel proteins covering hundreds of amino acid positions. Across all six individuals, from all three sites, the team identified two amino acid variants in a single protein: ameloblastin, which plays a key structural role in enamel formation.

Two variants, two stories

The first variant, designated AMBN(A253G), has never been seen before in any other hominin or primate. It wasn’t in the 1.77-million-year-old H. erectus tooth from Dmanisi, Georgia -- the only previous specimen to yield H. erectusproteins, though those had been too degraded to reveal informative variants. It wasn’t in Homo antecessor from Atapuerca. Not in Neanderthals, not in Denisovans, not in any modern human population. The variant appears to be specific to this population of Middle Pleistocene H. erectus from East Asia.

That exclusivity makes it useful. Fossils attributed to H. erectus across Asia are morphologically variable enough that researchers have sometimes disagreed about their relationships. The Hexian specimens, in particular, look morphologically closer to Indonesian H. erectus than to those from Zhoukoudian, and some researchers had proposed they might be more closely related to Denisovans. The AMBN(A253G) variant settles that debate. It appears in both Hexian teeth, just as it does in the Zhoukoudian and Sunjiadong teeth, clustering all six individuals together in a Bayesian phylogenetic tree with 100% posterior probability. Hexian belongs to H. erectus.

The second variant is where things get complicated. AMBN(M273V) had already been identified in Denisovans. The known Denisovan specimens include Denisova 3 from the cave in Siberia that gave the group its name, the “Dragon Man” skull from Harbin, the Penghu mandible from Taiwan’s seafloor, and a molar from Laos -- geographically scattered, morphologically distinct from Neanderthals and modern humans, linked almost entirely through ancient DNA and protein analysis rather than anatomy. When the Altai Neanderthal genome was analyzed, researchers found that the Denisovan lineage had received between 0.5% and 8% of its genome from a “super-archaic” hominin whose ancestors had diverged from the common lineage of Neanderthals, Denisovans, and modern humans more than a million years ago. They also found that roughly 15% of those super-archaic DNA regions later introgressed from Denisovans into populations in Asia and Oceania.

The AMBN(M273V) variant fits this pattern. It is absent from most modern human populations but present at 21% frequency in the Philippines, 1.17% in India, and 0.71% in Papua New Guinea -- exactly the populations expected to carry the highest proportions of Denisovan ancestry. It is homozygous in the more recent Denisovan specimens, Denisova 3 and Penghu 1, but heterozygous in two older Denisovans, the Harbin individual and Denisova 25 (dated to roughly 200,000 years ago). That pattern -- heterozygous earlier, homozygous later -- is consistent with an allele that entered the Denisovan lineage from outside and increased in frequency over time.

Fu’s team argues that the “outside” is H. erectus. The variant appears in all six H. erectus individuals across three sites spanning both northern and southern China. Its genomic region shows greater sequence divergence between modern Africans and Denisova 3 than between Africans and the Altai Neanderthal -- a signal that the region is older than the Neanderthal-Denisovan split, consistent with origin in a more diverged group. Denisovans diverged from Neanderthals approximately 380,000 to 470,000 years ago. The H. erectus specimens are dated to roughly 400,000 years ago. The timing, the geography, and the allele frequency pattern all point in the same direction: H. erectus populations in East Asia encountered Denisovans around or before this period, and something passed between them.

“Their shared habitats create opportunities for interactions,” Fu and colleagues write in the paper.

The scenario they reconstruct goes: H. erectus populations in East Asia, the same individuals whose teeth now sit in Beijing, Hefei, and Luoyang, passed the AMBN(M273V) variant to Denisovans through interbreeding, probably more than 400,000 years ago. Denisovans carried it forward, homozygosity increasing over time as it became more established in the population. Modern humans moving out of Africa and through Asia perhaps 50,000 years ago encountered Denisovans, interbred, and acquired it. A small proportion of people alive today -- primarily in island Southeast Asia and the Pacific -- carry a protein variant in their tooth enamel whose ancestry traces back to Homo erectus on the Chinese mainland almost half a million years ago.

What remains genuinely uncertain

Not everyone finds the causal chain fully convincing. Yale geneticist Diyendo Massilani, speaking to Science, noted that the study provides no direct evidence of admixture: each group could theoretically have evolved the M273V variant independently, through convergent mutation rather than shared ancestry. He characterized the admixture interpretation as hypothetical.

Kirsty Penkman at the University of York, a geochemist who specializes in ancient proteins and was not involved in the research, pushed back on that concern. Enamel proteins are structurally constrained -- they cannot mutate extensively without losing their function. A shared variant in an enamel protein is therefore less likely to represent independent evolution than a shared variant in a less constrained region of the genome might be,

“A biomineral protein can’t mutate too much,” she told Science, “because then it stops doing its job.”

The honest answer is that proteins cannot resolve this question the way genomic data can. Full H. erectus genome sequences would make the case definitively, but those sequences don’t exist and may never be recoverable. The DNA in these specimens is gone. What Fu’s team has done is work with what survives. John Hawks at the University of Wisconsin-Madison, also not involved in the study, made a point worth taking seriously,

“Scientists used to call this ‘the muddle in the Middle Pleistocene,’” he told LiveScience, “and now we know that muddling is just mixing.”

But he also raised a deeper question that the study implicitly surfaces: whether paleoanthropologists have been too loose with the label Homo erectus, grouping together fossils from the Middle Pleistocene in China that may actually represent Denisovan relatives or other populations not yet identified. The protein data confirms that the six specimens from Zhoukoudian, Hexian, and Sunjiadong belong together. It says less about where the boundaries of H. erectus as a category ultimately fall.

That is, for now, where the field sits. A small patch of acid-etched enamel, a handful of proteins, and two amino acid variants are the most informative molecular data ever extracted from a species that spent nearly two million years shaping the inhabited world. As Penkman put it: fewer than 20 hominins have ever yielded analyzable protein sequences. Each new specimen adds something. This study adds six at once -- and what they add is the outline of an encounter, written in the crystalline architecture of a tooth.

Further Reading

* Welker, F., et al. (2020). The dental proteome of Homo antecessor. Nature, 580, 235–238.

* Prüfer, K., et al. (2014). The complete genome sequence of a Neanderthal from the Altai Mountains. Nature, 505, 43–49.

* Hubisz, M.J., Williams, A.L., & Siepel, A. (2020). Mapping gene flow between ancient hominins through demography-aware inference of the ancestral recombination graph. PLoS Genetics, 16, e1008895.

* Tsutaya, T., et al. (2025). A male Denisovan mandible from Pleistocene Taiwan. Science, 388, 176–180.

* Fu, Q., et al. (2025). The proteome of the late Middle Pleistocene Harbin individual. Science, 389, eadu9677.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

When T. Whitworth described a set of human cranial fragments from Turkana, Kenya, in 1960, he found himself stuck. The skull was thick-walled. The forehead sloped. The mandible was heavy-built in a way that recalled older, more archaic specimens. But there was also a chin. A proper chin, the bony projection below the lower teeth that counts as one of the diagnostic markers of Homo sapiens.

Whitworth was honest about the contradiction. He noted the archaic features, noted the modern one, and didn’t resolve them into a clean interpretation. The fossil got filed away. It was described again in 1966 under the name Kabua 1, mentioned in subsequent surveys, and mostly left alone. Sixty years later, a team led by Abel Marinus Bosman has given it the quantitative analysis it never received, combined with new uranium-series dating. The picture that emerges is strange in an instructive way.

Kabua 1 is probably a Homo sapiens. But it is one that carries enough archaic-looking morphology to resist easy classification, and that difficulty, the team argues, is itself meaningful.

What the fragments preserve

The Kabua 1 material, held at the Natural History Museum in London, is fragmentary: pieces of the frontal, parietal, occipital, and temporal bones, along with maxillary and mandibular fragments and several teeth. Working from microCT scans, Bosman and colleagues produced six separate virtual reconstructions, each guided by a different reference cranium ranging from the Middle Pleistocene fossil Kabwe 1 (sometimes called Broken Hill) to recent Maasai individuals. The approach is deliberate. Because Kabua 1 is incomplete and taphonomically distorted, any single reconstruction encodes assumptions about the missing parts. Six different references give a distribution of possible morphologies rather than a single answer, which the team describes as analogous to priors in a Bayesian framework. It is a more epistemically honest way to handle fragmentary material.

The reconstructions were compared against a broad sample: Middle Pleistocene Africans and Eurasians, Neanderthals, Late Pleistocene H. sapiens, and 109 recent African H. sapiens crania from eastern and southern Africa. The analysis focused on the neurocranium, because neurocranial globularity, the roundedness and anteroposterior shortness of the braincase, is one of the most diagnostically useful markers distinguishing derived H. sapiens from more archaic hominins.

In the principal component analysis, the first axis captures the difference between elongated, low vaults and rounded, globular ones. Neanderthals and Middle Pleistocene hominins cluster toward the elongated end. Derived H. sapienscluster toward the globular end. All six Kabua 1 reconstructions fall within the H. sapiens range on this axis.

The second principal component complicates things. The Kabua reconstructions score low on it, overlapping with Middle Pleistocene hominins and Neanderthals as well as some H. sapiens. What this reflects, concretely, is a short nuchal plane, a less vertical frontal bone, and a mediolaterally narrow vault. The reconstruction based on the Ngaloba cranium sits closest to the Middle Pleistocene African group in morphospace, near where the fossil Saldanha overlaps with H. sapiens variation. The reconstruction based on Skhul 5 sits at the opposite extreme, exhibiting a markedly globular neurocranium.

The classification analyses add a further layer of nuance. Linear discriminant analysis classifies five of the six reconstructions as Middle Pleistocene African, and only the Skhul 5-based one as H. sapiens. The machine learning methods, which are more robust to the non-normal distribution of the data and less sensitive to sample size imbalances, are more mixed. As the H. sapiens reference sample grows, the reconstructions increasingly classify with that group, with the Ngaloba-based reconstruction the persistent holdout. The authors are appropriately cautious about the LDA result: LDA is designed to maximize between-group variance and can produce artificial separations, and the Middle Pleistocene African grouping lumps specimens with quite different degrees of relatedness to derived H. sapiens.

The overall picture is that Kabua 1 has a broadly H. sapiens neurocranium in terms of globularity but retains features that, taken together, pull it toward a more archaic morphological zone. It is not a Neanderthal. It is not Kabwe. But it is not a specimen that would dissolve without trace into a sample of recent Africans.

The date and what it means

Getting a reliable age for Kabua 1 has proved difficult, and the new uranium-series results don’t fully resolve that difficulty. They do establish a conservative minimum.

Uranium-series dating of fossil bone works because uranium migrates into bone from groundwater after burial. The ratio of uranium isotopes to decay products gives an indication of how long that process has been running. The complication is that uranium can also leach back out, or a second wave of uranium can overprint the first. Each of these events scrambles the signal. Three samples from Kabua 1 were dated: one from the biparietal-occipital fragment, one from a tooth in the mandible, and one from the temporal bone.

The tooth has undergone at least three distinct uranium mobilization events, including leaching, making any age derived from it unreliable. The temporal fragment gives an apparent minimum age of around 114,000 years, but there is evidence that a past leaching event may have affected this figure too. The biparietal-occipital sample gives a minimum age of 64.4 ± 5.4 ka. The team judges this the most conservative and defensible figure, because the pattern of uranium isotope ratios in that sample is most consistent with a simple accumulation history unaffected by leaching.

This is a minimum age. The actual date of death could be older, potentially considerably older. But at least 64,000 years old places Kabua 1 in the Late Pleistocene, broadly contemporaneous with the populations that would eventually spread beyond Africa.

What diversity looks like before the bottleneck

The reason Kabua 1 matters is not just that it is another Late Pleistocene African fossil. It is that it adds to a pattern.

Several other Late Pleistocene African fossils show a similar mismatch between what age would predict and what morphology delivers. Iwo Eleru from Nigeria, dated to around 14,000 years ago, has morphology that looks more archaic than it should given its age. Lukenya Hill in Kenya and Ishango in what is now the Democratic Republic of Congo, both around 20,000 to 25,000 years old, show comparable features. Hofmeyr from South Africa, dated to roughly 36,000 to 38,000 years ago, has a continuous supraorbital torus, a brow ridge, which is vanishingly rare in derived H. sapiens. Nazlet Khater 2 from Egypt, from around the same period, has a face and mandible with proportions that recall Middle Pleistocene hominins far more than they resemble recent Africans.

These fossils come from different times and different parts of the continent. They do not cluster into a single variant population. What they share is the preservation of morphological features that most straightforward models of H. sapiens origins would predict should have been largely gone by their respective dates. In each case, the overall assignment to H. sapiens is not seriously contested, but the details resist easy explanation.

Two broad frameworks exist for thinking about this. One holds that H. sapiens origins were pan-African: our species emerged as a continent-wide process of population fragmentation and coalescence, with derived traits arising in different regions and spreading through periodic gene flow. Under that model, morphological diversity across Late Pleistocene Africa is expected, because populations were partly isolated and some archaic-looking features represent survivals of deep population structure. The competing view holds that derived H. sapiens originated in a specific region and spread with varying degrees of replacement, and that individuals with more archaic morphology either belong to lineages that eventually went extinct without contributing to living people, or represent the tail ends of variation within a broadly modern species that our comparative frameworks don’t fully capture. Both positions currently have genetic and morphological evidence in their favor.

What Kabua 1 does, with its new date and new analysis, is take its place in the growing series of fossils that make the African Late Pleistocene look unexpectedly variable. The team places it alongside Hofmeyr and Nazlet Khater as examples of Late Pleistocene H. sapiens that combine a broadly modern neurocranial shape with features that, in isolation, would look considerably older.

There is a structural reason to expect this kind of diversity. Living populations outside Africa descend from a relatively small founding group, and multiple successive founder effects compress morphological variation. Living Africans are already more variable, craniometrically, than non-Africans. The Late Pleistocene fossil record suggests the full historical range was wider still. With a small fraction of that variation preserved in bone and most of it lost, we are probably only beginning to map what H. sapiens actually looked like during the period when the species was exclusively African.

How much of that variation is ancestral to anyone living today, and how much belongs to lineages that terminated without genetic descendants, is a question that morphology alone cannot answer. Ancient DNA from Kabua 1 would help, but extraction from a specimen of this age from an equatorial environment is unlikely to succeed. For now, Kabua 1 returns to the record with a more secure identity than it had before, which makes it more useful and, in some ways, more puzzling.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

The mound at Baraleti sits near the center of the Javakheti Plateau, and its name tells you something right away. Natsargora means “hill of ashes.” Not a metaphor. When excavations began in 2023, researchers from the Samtskhe-Javakheti Project found out why: layer after layer of burning events, each folded into the next across roughly three thousand years of occupation. People kept coming back to this place, and they kept burning things down, or finding it burned, and building again.

That kind of stratigraphy does not happen by accident. It takes a place that matters.

The Javakheti Plateau is a high-altitude grassland tucked into the southern Georgian highlands, between the Greater Caucasus range and the borders of Turkey and Armenia. It sits at elevations that make sustained agriculture difficult. It has been largely off the radar of systematic archaeological survey. Since 2017, a joint Georgian-Italian initiative called the Samtskhe-Javakheti Project has been changing that. Eight years of fieldwork, combining remote sensing, GPS mapping, and GIS-based analysis with targeted excavations, have produced a picture of the plateau as something quite different from the empty frontier it has usually been treated as. The team has documented more than 168 archaeological sites ranging from Bronze Age settlements to medieval monasteries.

The sheer number is striking. The more interesting question is what kind of occupation these sites actually represent.

Fortresses That May Not Have Been Fortresses

Much of the earlier scholarly attention on this region focused on the plateau’s most visible features: the “Cyclopean” fortifications, large-scale stone enclosures with walls built from blocks of almost absurd dimensions. The conventional assumption was that these were defensive citadels, military architecture serving the needs of settled populations under threat.

The SJP’s survey data complicates this. Several of the megalithic enclosures, when examined closely, appear to have operated not as permanent military installations but as temporary refuges. A 2022 study of the Abuli and Shaori complexes by Licheli and colleagues proposed that mobile pastoralist groups used such structures seasonally, as secure points during their movements through the highlands or during periods of instability. Cyclopean walls, on this reading, were less about holding territory than about having somewhere reliable to retreat to.

The difference matters. A landscape of defensive citadels implies a settled population defending fixed resources. A landscape of seasonal refuges implies something more fluid: herders moving between altitudes, their architecture a response to mobility rather than sedentism, the plateau functioning as an intermediate zone rather than a destination.

Sites like Abulis Gora and Saro-1 support a picture of long-term but episodic use. Occupation was not continuous; it was recurrent. People returned to the same ground across centuries, which implies deep familiarity with the landscape, knowledge of where to find water and shelter and stone, transmitted across generations. The necropolises near Bertakana and Lake Tabatskuri make this point in a different register. Funerary traditions do not persist at the same location for that long without a community that keeps returning.

What the Ash Preserves

At Baraleti Natsargora, the occupation sequence runs from the Early Bronze Age through the Iron Age, roughly 3500 to 500 BCE. The recurrent burning episodes are both literal and interpretively useful. Each marks a break. Each rebuilding marks a renewal. The defensive wall uncovered in excavation, along with traces of partition walls and clay installations consistent with domestic use, suggests the site functioned as a fortified settlement during its peak phases.

A bronze solar disk came from the vicinity of Baraleti during earlier surveys. It is a carefully made object: concentric bands of raised knobs, angular incised motifs, regularly spaced perforations around its rim. Objects like it have been found across southern Georgia, and the pattern of association is consistent. They appear in burials, and they appear disproportionately in female graves. The Baraleti disk now sits in the Akhalkalaki Museum, and while its precise findspot within the site remains uncertain, the weight of regional evidence suggests it accompanied someone into death.

What a solar disk meant to the people who made it and buried it is harder to say. The iconography connects to a wider tradition of solar imagery in the protohistoric South Caucasus and across the steppe. Whether it functioned as an identity marker, a symbol of status, a ritual object, or something that resists those categories is a question the object itself cannot answer. The pattern of female association is at least worth sitting with.

At Meghreki Fortress, a few kilometers to the east, the team found something more unexpected than bronze metalwork. Inside domestic structures provisionally dated to the Late Iron Age and Achaemenid horizon, roughly the sixth to fourth centuries BCE, were fired clay plaques carrying incised and painted geometric designs. Red, white, and dark blue pigments. The visual effect would have been striking in a room lit by firelight.

Painted clay plaques of this kind are not a standard feature of South Caucasian domestic assemblages. Parallels exist at Digasheni and Amiranis Gora, but the density and context at Meghreki stand apart. The interpretation being advanced is that decorated plaques marked ritualised or high-status domestic spaces. Not a temple. Not a public monument. A room in a house. The line between the symbolic and the everyday, at Meghreki, ran through the walls of ordinary buildings.

This is worth pausing on. Meghreki’s occupation runs from the Kura-Araxes culture in the Early Bronze Age through to the medieval period, a span of over four thousand years, with the Iron Age as its most intensive phase. Across that time, the architecture shifted, the material culture shifted, the political situation around the plateau shifted. But people kept building at the same spot, and at one particular moment they put painted clay panels on the walls. Whether that moment marked an intensification of ritual practice, a new form of social display, or an outside cultural influence that found local expression is open. The Achaemenid horizon connection is suggestive. Achaemenid influence reached the South Caucasus unevenly and was absorbed in different ways at different scales. A highland settlement in Georgia would not have taken it in straightforwardly.

What the SJP is building, across eight years and 168 sites, is a framework for a region that has been treated as peripheral for too long. The Javakheti Plateau turns out to have been a crossroads: a zone where highland and lowland traditions met and overlapped, where mobile groups and settled communities negotiated the same terrain, where symbolic objects traveled with people and sometimes outlasted them. The ash at Baraleti has been accumulating for five thousand years. There is still a lot to read in it.

Further Reading

* Gambashidze, I. 1999. Samtskhe in the middle of the 2nd millennium BC (according to archaeological materials from the Borjomi Gorge). Unpublished Ph.D. dissertation, Archaeological Research Center, Tbilisi.

* Orjonikidze, A. 1988. “The Early Bronze Age settlement of Digasheni I.” Dziebani 1(1): 15–22.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

Somewhere around 4,000 BC, people living at a promontory above the Sitna river in what is now northeastern Romania built something that didn’t fit. Their settlement had roughly 45 houses, each between 70 and 120 square meters, arranged in orderly rows behind a system of ditches and palisades. The houses were consistent in size, similar in construction, and distributed across a plateau of about 4 hectares. Then there was House 5/6. At approximately 350 square meters — nearly four times the median size of the surrounding structures — it sat between the ditches, positioned squarely near what was probably the settlement’s main entrance. It faces you as you arrive. You cannot miss it.

This building, at the site now called Stăuceni-’Holm’, is what Cucuteni-Trypillia researchers call a mega-structure. A team from Friedrich-Alexander-University Erlangen-Nürnberg, working alongside the Botoșani County Museum and the National Museum of Bukovina, partially excavated it over four-week campaigns in 2023 and 2024. The results, published in PLOS One in March 2026, add a new site to a very short list and raise questions that the excavation cannot yet resolve.

The Cucuteni-Trypillia complex is one of the more genuinely puzzling phenomena in European prehistory. Its settlements, distributed across what is now Romania, Moldova, and Ukraine, range from small villages to enormous concentrations of population. The largest of them, the so-called mega-sites, covered hundreds of hectares and housed thousands of contemporaneous dwellings — Maidanetske, in Ukraine’s Cherkasy Oblast, counts around 2,930 houses across 170 hectares. Some of these sites were occupied continuously for over three centuries. And yet they show, at least on the surface, almost no signs of hierarchy. No palaces. No obviously elite burials. No administrative script. The houses look alike. The material culture is distributed without obvious concentration of wealth.

The question that has driven much of the recent research is how communities of this size organized themselves. Some scholars have argued for a fundamentally egalitarian or non-hierarchical social structure. Others have pushed back on this, pointing to subtle differences in building size and position that the older literature sometimes smoothed over. The mega-structures sit at the center of this debate, because they are the most visible departure from the pattern of uniformity — and yet their function remains genuinely unclear.

What a Mega-Structure Is, and What It Isn’t

A working definition has emerged from prior research, particularly a 2019 synthesis by Hofmann and colleagues that catalogued 140 mega-structures across 19 sites. They are rectangular, substantially larger than the surrounding buildings, positioned in visible or open locations, and often architecturally distinct from domestic dwellings. The relative scale matters more than absolute dimensions: a 350-square-meter building in a settlement where the median house is 91 square meters is an outlier in a way that the same structure wouldn’t be in a different cultural context.

Before the excavation at Stăuceni-’Holm’, the primary evidence for mega-structures came from geomagnetic surveys. When clay-rich structures burn and collapse, the fired material leaves strong magnetic anomalies detectable from the surface. Researchers have used the patterning of these anomalies — their density, distribution, and configuration inside a structure’s footprint — to infer internal features: walls, hearths, platforms, room divisions.

The excavation at Stăuceni-’Holm’ delivered a sharp correction to this practice. When the team dug into the mega-structure and mapped its geomagnetic signal at successive levels, they found that nearly all of the anomalies they had expected to correspond to interior installations disappeared without trace. No hearths. No raised platforms. No subdividing walls where the magnetogram had suggested them. What they were seeing, it turned out, was simply the varying density of collapsed daub — the fired clay that had once formed the building’s floor and walls. Remove the daub layer, and the anomalies vanish. Only the foundation ditch, cut deeper into the subsoil, remained clearly visible.

The authors frame this directly: no information on the interior layout of a Cucuteni-Trypillia mega-structure can be reliably derived from the magnetogram alone. Given how much of the existing comparative analysis rests on geomagnetic data, this is a meaningful caution.

What the excavation did find was a foundation ditch tracing the building’s perimeter, roughly U-shaped in section, with postholes cut into it at intervals of 70 to 90 centimeters. The posts had been wedged in place using fragments of fired clay, apparently debris from earlier structures elsewhere on the site. Beneath the clay floor, small tree trunks had been split in half and laid flat-side-down, forming a wooden substructure over which a layer of clay was then applied and smoothed. Two substantial postholes ran along the building’s central axis, deep enough (more than 80 centimeters below the floor level) to have supported a roof. A third central post, if confirmed in the unexcavated northern portion of the structure, would divide the interior lengthwise into two halves.

Noticeably absent: ovens, storage features, grinding stones. The amount of burnt clay associated with the structure is far lower than what appears in the domestic buildings at the site, which typically burn thoroughly and leave substantial daub deposits. The mega-structure’s walls may have been largely wooden, burning without leaving much material trace.

The Dating Problem

The team’s working hypothesis, based on the pottery collected from the surface and during excavation, was that Stăuceni-’Holm’ dates to the Cucuteni A3 phase — a period generally assigned to roughly 4350 to 4050 cal. BC. If correct, this would make the mega-structure among the earliest known examples of the type, younger only than the Pre-Cucuteni mega-structure at Baia, dated to around 5000 BC.

Then the radiocarbon results came back.

Two short-lived samples from the building’s floor — a fragment of Prunus sp. fruit stone and a seed of Polygonum convolvulus, both drawn from protected locations between the floor timbers — returned dates placing the structure in the 40th to 39th centuries BC. That is, around 4000 to 3800 cal. BC. For Cucuteni AB material, not Cucuteni A3.

The team considers and rejects several ways out of this contradiction. The samples came from annually growing plants, so an old-wood effect can be excluded. They were recovered from stratigraphically secure positions connected directly to the floor’s construction. A pot found smashed in one of the foundation postholes — a small globular cup in fine-tempered clay, decorated with red ochre and incised designs in what the team identifies as Cucuteni A3 style with Precucuteni technical traditions — links the pottery type to the same depositional event as the radiocarbon dates.

The honest conclusion they reach is that something is wrong with the existing chronology, or at least with its application to this region. The absolute dating framework for Cucuteni A3 in the Upper Siret-Prut area rests on very few radiocarbon measurements, several of which carry high standard deviations and uncertain connections to diagnostic material. For that specific region, the published literature relied on four dates before 1995, three of them from charcoal at a single site. The Stăuceni dates, derived from short-lived plant material in an unambiguous context, may be the most methodologically sound anchor the region has.

The implication, if the dates hold, is that Cucuteni A3 may persist later in the western distribution area than current models allow. The researchers note that the absolute dates from Stăuceni-’Holm’ would be broadly contemporaneous with the mega-structures at Nebelivka and Dobrovody in Ukraine, and with the radiocarbon-dated mega-structure at nearby Ripiceni-’Holm’, dated to 4041-3975 cal. BC. What looked like a chronological outlier may instead reflect real overlap across the Cucuteni-Trypillia cultural zone that existing phase divisions obscure.

What Was Happening Inside

The finds from the building’s interior are sparse in the way that characterizes most excavated mega-structures. Concentrated pottery deposits, mostly as sherds, appear especially toward the building’s eastern interior. Among the more distinctive objects: a clay bowl bearing a zoomorphic protome in the form of a bull’s head with once-massive horns; three ladles, one still bearing traces of painted decoration on both faces; a single clay cone of the type sometimes labeled “conical idols” in the literature; and 87 flint artifacts, about half of them burned. The flint count is notably higher than the handful recovered from the Ukrainian mega-sites, though this likely reflects proximity to local raw material sources rather than functional difference.

Archaeobotanical sampling from the floor produced carbonized cereal grains (unidentifiable below the level of Cerealia), associated weed taxa including Polygonum convolvulus, Galium cf. aparine, Chenopodium album, and Echinochloa crus-galli, along with gathered fruits from cornelian cherry (Cornus mas), plum (Prunus sp.), elder (Sambucus sp.), and hawthorn (Crataegus sp.). Carbonized fruit stones are a marker of in-situ food consumption. Feather grass (Stipa sp.) awns appeared in several samples, consistent with insulation, bedding, or floor covering. One sample contained mineralized remains of henbane (Hyoscyamus niger), a psychoactive and medicinal plant.

None of this points unambiguously toward any single function. The food remains suggest something was consumed here, but not at the scale of a communal storehouse or dedicated feasting hall. The absence of ovens and grinding stones argues against sustained domestic food processing. The henbane is intriguing but insufficient. The researchers decline to commit to a function: storage building, unlikely; dedicated ritual space, no clear evidence; assembly hall for collective decision-making, possible; residence for a high-status household, not ruled out.

The structure occupies, in the settlement layout, a position analogous to what Hofmann and colleagues designate P5 in the larger Ukrainian ring-settlements: outside the main residential zone, well-visible to anyone approaching the site from the south, the only accessible direction given the steep slopes to the west, north, and east. That positioning has meaning. Something was meant to be seen there, or encountered upon entry.

Only about 25 percent of the building has been excavated. The northern half, unexcavated, may yet contain the third central posthole that would confirm the building’s internal division, an oven or hearth that would reframe the food residue evidence, or the absence of these things. The archaeology of Cucuteni-Trypillia settlements tends to reward patience. The building waited six thousand years to be partly reopened. Whatever it was, it was built to be noticed.

Further Reading

* Hofmann R, Müller J, Shatilo L, Videiko M, Ohlrau R, Burdo N (2019). Governing Tripolye: Integrative architecture in Tripolye settlements. PLOS One. https://doi.org/10.1371/journal.pone.0222243

* Gaydarska B, Nebbia M, Chapman J (2019). Trypillia Megasites in Context: Independent Urban Development in Chalcolithic Eastern Europe. Cambridge Archaeological Journal 30(1):97–121. https://doi.org/10.1017/s0959774319000301

* Hofmann R, Müller J, Müller-Scheeßel N (2024). Trypillia mega-sites: a social levelling concept? Antiquity 98(398):1–21. https://doi.org/10.15184/aqy.2024.18

* Diachenko A, Harper TK, Chernovol DK, et al. (2024). Testing scale-dependent temporal and spatial biases in relative chronology using AMS 14C dating: A case study of Early–Middle Cucuteni-Tripolye sites in Southeastern Europe. Journal of Archaeological Science: Reports 55:104495. https://doi.org/10.1016/j.jasrep.2024.104495

* Asăndulesei A, Tencariu FA, Nicu IC (2020). Pars pro toto — Remote Sensing Data for the Reconstruction of a Rounded Chalcolithic Site from NE Romania: The Case of Ripiceni–Holm Settlement (Cucuteni Culture). Remote Sensing12(5):887. https://doi.org/10.3390/rs12050887

* Ohlrau R (2020). Maidanets’ke. Development and decline of a Trypillia mega-site in Central Ukraine. Sidestone Press, Leiden.

* Chapman J, Videiko M, Gaydarska B, Burdo N, Hale D (2014). Architectural differentiation on a Trypillia mega-site: preliminary report on the excavation of a mega-structure at Nebelivka, Ukraine. Journal of Neolithic Archaeology16:135–6.

* Kirleis W, Corso MD, Pashkevych G, Schlütz F, Hofmann R, Terna A, et al. (2023). A complex subsistence regime revealed for Cucuteni–Trypillia sites in Chalcolithic eastern Europe based on new and old macrobotanical data. Vegetation History and Archaeobotany 33(1):75–90. https://doi.org/10.1007/s00334-023-00936-y

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

The petrous bone is not beautiful. It is a dense, pyramidal wedge buried in the base of the skull, housing the cochlea and semicircular canals, its name taken from the Latin petrosus, meaning rock-like. It weighs almost nothing. You could balance one on your fingertip. Under most circumstances it would register as unremarkable.

In paleogenomics, it has become the most fought-over fragment of bone on earth.

The cochlea within the petrous is the densest, most highly mineralized structure in the human skeleton. It finishes forming by the eighth week of fetal development and then, essentially, stops changing. Unlike most bone, it doesn’t remodel after birth. It sits there, stable, shielded, preserving its chemistry for centuries or millennia after death. When researchers began testing which skeletal structures yield the highest-quality ancient DNA, the cochlea came out on top, beating out teeth, femora, and everything else. The petrous became the default target for nearly every major paleogenomics lab in the world.

According to metadata compiled in the Allen Ancient DNA Resource (AADR v9), a curated database of ancient human genomes, ancient DNA has been retrieved from the petrous bone in 47% of all sequenced individuals. In 96% of those cases, it was the only bone sampled.

That is a lot of irreversible drilling.

Petrous Fever

The phrase “petrous fever” appears in a 2024 paper in Current Anthropology, describing the gap between what ancient DNA labs claim to prioritize and what they actually do in practice. It describes a pattern of extensive, often redundant sampling from skeletal collections, frequently without clearly defined research questions, driven by the logic of data generation rather than specific inquiry. More than 10,000 ancient human genomes have been generated over the past decade, the majority relying on petrous bones. The momentum is real. So is the problem it creates.

The cochlea is not just a DNA vault. The bony labyrinth inside the petrous bone is also a morphological record. Its shape reflects population history and has been used to trace migration and evolutionary divergence across time. The cranial base, of which the petrous is a component, carries a strong phylogenetic signal less susceptible to environmental noise than other cranial regions. And because the petrous forms so early in development and remains so stable, its fine structural details survive taphonomic processes that destroy nearly everything else. Studies of inner ear anatomy, basicranial morphology, stable isotopes, and radiocarbon dates have all increasingly relied on the same small piece of bone.

When you drill into the cochlea to extract DNA, you destroy it. The morphological data, the histological information, the stable isotope signal, the inner ear geometry that might have told you something about population dispersal or evolutionary change: all of it is gone. Researchers visiting skeletal collections have encountered cases where both petrous bones from a single individual were removed entirely, one for genetic analysis and one for radiocarbon dating, sometimes both for separate DNA extraction attempts by the same lab, even when a single sample would likely have been sufficient.

The question that Lumila Paula Menéndez and colleagues at the University of Bonn, along with a large international team, set out to address is both practical and ethical: can you digitally preserve a petrous bone before destroying it for molecular analysis, and if so, does the imaging process itself damage the DNA you’re trying to extract?

The Worry About X-Rays

Micro-computed tomography, or µCT, works by passing X-rays through a sample from multiple angles and using algorithms to reconstruct high-resolution three-dimensional cross-sections. For the inner ear, it produces extraordinarily detailed models of the cochlea and semicircular canals, images that can be archived, shared, and analyzed without any physical contact with the original bone. Its advocates have been pushing for it to become a standard step in any research workflow involving the petrous.

The hesitation came from a reasonable concern. CT scanning uses ionizing radiation, and ionizing radiation can damage DNA. If scanning compromises the very molecules researchers are trying to recover, the approach defeats its own purpose.

Prior work on animal samples and other human bones had suggested that standard µCT parameters fall below radiation doses likely to cause meaningful DNA degradation. But the petrous bone is not a generic skeletal element. Its microstructure and diagenetic history are distinctive. Its density means it interacts with radiation differently. And because it yields such unusually high proportions of endogenous ancient DNA, even minor damage could have disproportionate effects. No previous study had tested µCT effects directly on archaeological human petrous bones.

Menéndez and her team found themselves in a position to do exactly that, not by design but by circumstance. In 2021, Menéndez contacted a lead geneticist who was already sampling petrous bones from South American archaeological individuals for a population genomics project. She proposed adding µCT scans before any destructive procedures. Some samples were scanned. Others, already in the pipeline, were not. The mixed dataset created a rare opportunity for comparison.

The team analyzed 93 petrous bone samples from Middle and Late Holocene archaeological contexts across the Andes, Central Argentina, and the Cuyo region, spanning roughly 6,000 to 200 years before present. Fifty had been µCT-scanned before molecular analysis; 43 had not. They compared six standard parameters used to evaluate ancient DNA quality and authenticity: endogenous DNA content, read length, cytosine deamination patterns, which are a chemical hallmark that confirms the ancient origin of DNA sequences, and contamination estimates based on both mitochondrial and nuclear DNA. The scanning itself was carried out on a Skyscan 1176 Bruker scanner at the Faculty of Medicine, Paris Diderot University, at 35 µm voxel size, 65 kV, and with a 0.5 mm aluminum filter to reduce unnecessary radiation exposure.

No statistically significant differences emerged between the scanned and unscanned groups across any of the six parameters. Endogenous content was similar. Read length was similar. Deamination patterns indicating authentic ancient origin were nearly identical between groups. Nuclear contamination estimates showed no difference at all.

One parameter came close to significance. Mitochondrial contamination estimates were marginally higher in the scanned samples, with a p-value of 0.051. But 48 of 49 scanned samples remained below the standard 5% threshold for genomic analysis, and the team’s most parsimonious explanation is selection bias: researchers tend to choose visually well-preserved, more complete bones for scanning, which may incidentally have had more handling and thus slightly more exposure to environmental contaminants. No corresponding effect appeared in nuclear contamination, which would be expected if radiation-induced damage were the mechanism. And it is worth noting that the highest endogenous DNA content in the entire dataset was found in the scanned group, an outcome that makes no sense if scanning is degrading the material.

The absence of a detectable effect under these specific conditions is not proof that no effect exists under all conditions. The authors are careful here. Synchrotron scanners, which can deliver far more intense radiation, are a separate matter. The study’s design, because it wasn’t a matched-pair experiment, limits the ability to attribute every observed pattern solely to the scanning variable. Subtle effects at the margins of detectability might still exist. Future controlled experiments using matched samples from the same individuals would strengthen the picture considerably.

There is also one caveat that sits slightly outside the scope of this study but matters for field practice: some evidence suggests that µCT scanning can reduce collagen yield, which could affect the reliability of radiocarbon dates from bone. Researchers building multi-method workflows that include 14C analysis should factor this in.

What Happens When You Slow Down

The paper’s practical argument follows from its result. If µCT scanning under standard parameters doesn’t compromise ancient DNA, there’s no technical reason to skip it before destructive sampling. Which means the barrier that many labs cited for not scanning first is substantially weaker than it seemed. What remains are the logistical and financial barriers, and the cultural inertia of a field that has organized itself around rapid data production.

The workflow Menéndez and her team propose is sequential and deliberately unhurried. Before any bone powder is collected, before any drilling begins, the petrous goes through macroscopic assessment, biological profiling and osteobiographical analysis, digital preservation via CT and surface scanning, and compositional pre-screening using techniques like X-ray fluorescence and near-infrared spectroscopy to evaluate collagen preservation and guide sample selection. Molecular sampling, with a clear research question in hand, comes last.

This is not a radical reorganization. It is a sequencing change. It asks researchers to wait a few steps before drilling, to ask what other questions might be answered by this bone before it’s gone.

The obstacles are real. CT scanners are not uniformly accessible. The global average cost of a scan is around $100, and many research groups operating with limited budgets have not planned for imaging. Some institutions house their own µCT equipment with trained operators. Others operate in regions where even medical CT scanners are unavailable within thousands of kilometers. The paper addresses this and suggests collaborations with public medical facilities, universities, and private medical centers as partial solutions. Portable surface scanning and photogrammetry present fewer logistical barriers and can often be done in situ.

The geopolitical dimension here is not incidental. Most of the ancient human samples driving the aDNA boom originate in the Global South. Most of the large, well-resourced laboratories generating genomes from them are in the Global North. Menéndez and her colleagues are explicit about this: labs in core research countries have both the capacity and a particular responsibility to lead the transition toward more sustainable practices, precisely because they are the primary beneficiaries of destructive sampling from collections held in institutions that often lack the infrastructure to implement better practices on their own.

The proposed workflow is also an argument about what counts as evidence. The petrous bone that tells you about genome-wide ancestry might also have told you something about inner ear geometry and population dispersal. About diet and mobility through stable isotopes. About developmental biology. Once it’s powder, none of those conversations are possible. The field has been systematically foreclosing its own future questions.

Researchers at collections across multiple continents have encountered individuals from whom both petrous bones were removed entirely, nothing left. It’s difficult to know how many of those samples were digitally preserved before the drilling started, the paper notes. Probably only a small fraction.

The proposed five-step workflow won’t be adopted uniformly or quickly. There are labs that will keep sampling as they always have, and there are collections that will continue to be depleted without the kind of careful sequencing this paper advocates. But the technical argument for delay has weakened. The concern about X-ray damage to DNA, under standard µCT conditions, appears to be manageable. What’s left is a choice about what kind of science gets done with what remains.

Further Reading

* Mallick S, Micco A, Mah M, et al. The Allen Ancient DNA Resource (AADR): a curated compendium of ancient human genomes. Scientific Data 11(1):182 (2024). https://doi.org/10.1038/s41597-024-03031-7

* Pinhasi R, Fernandes D, Sirak K, et al. Optimal ancient DNA yields from the inner ear part of the human petrous bone. PLoS One 10(6):e0129102 (2015). https://doi.org/10.1371/journal.pone.0129102

* Immel A, Le Cabec A, Bonazzi M, et al. Effect of X-ray irradiation on ancient DNA in sub-fossil bones: guidelines for safe X-ray imaging. Scientific Reports 6:32969 (2016). https://doi.org/10.1038/srep32969

* Duval M, Martín-Francés L, Wood R. On the impact of micro-CT scanning on radiocarbon dating of fossil material: a cautionary note for the palaeoanthropological community and beyond. Radiocarbon 67(4):709–718 (2025). https://doi.org/10.1017/rdc.2025.23

* Sirak K, Fernandes DM, Cheronet O, et al. A minimally-invasive method for sampling human petrous bones from the cranial base for ancient DNA analysis. BioTechniques 62(6):283–289 (2017). https://doi.org/10.2144/000114558

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

Somewhere between 1665 and 1680, a man was buried at the hip with a clay pipe. Traces of soot were still lodged in the bowl. The pipe maker’s stamp on the heel showed a crowned rose and the monogram RW, later traced to Reichard West, a pipe maker recorded in council minutes from Mannheim, Germany, who appears in the historical record between 1673 and 1675 and died around 1675 to 1676. Clay pipes of this period were fragile things, rarely lasting more than three years. The pipe had probably been purchased not long before the burial.

That small object, found at the hip of individual 229.2 in a cemetery beneath what is now the Stadtcasino concert hall in Basel, Switzerland, helped anchor an entire group of burials to one of the most significant mortality events in Swiss history: the last recorded plague outbreak in the country, which struck Basel between 1665 and 1670.

The site has a layered past. The former Barfüsserkloster, a Franciscan monastery founded in 1250, was abandoned after the Reformation in 1529 and its cloister and outbuildings repurposed as a kind of early institution adjacent to the city’s Bürgerspital, the municipal hospital. The cloister garden became a cemetery. Excavations in 2016 and 2017 uncovered the remains of 279 individuals buried in the post-Reformation period, including a cluster of four graves in the northern section of the site containing 15 individuals interred in coffins and earthen pits, some buried simultaneously.

A new study published in Antiquity, led by osteoarchaeologist Laura Rindlisbacher of the University of Basel, brings together ancient DNA analysis, stable isotope data, historical archival research, and detailed osteological observation to reconstruct who these people were before they died. The answer is not flattering to the city that produced them.

Fifteen Bodies, One Bad Life

The 15 individuals in graves 229 through 232 ranged in age from a child of 3 or 4 to an older adult. The average age at death was 17.7 years. Only four individuals survived beyond 20. Because so many were subadults, sex assessment was impossible for eight of them; of those who could be assessed, five were likely male and two likely female.

The demographic shape of this burial group is, by itself, unusual. In a cemetery that reflects ordinary background mortality, young children under six typically dominate the subadult count, because childhood disease in pre-modern populations was front-loaded. Here the picture is different: the largest spike falls in the 10-to-19 age range, a period that typically shows low mortality in non-catastrophic samples. The team considered and ruled out warfare as an explanation. There are no perimortem injuries. The mortality profile points to epidemic.

Ancient DNA analysis confirmed it. Teeth from ten individuals were sampled, and five returned positive results for Yersinia pestis. The best-preserved sample, from individual 229.2, produced sufficient genomic coverage to place the Basel strain on a maximum likelihood phylogenetic tree. It clusters most closely with Y. pestis strains recovered from plague contexts in London (1560-1635) and Marseille (1720-1722). All three carry a depletion in the pla gene region on the pPCP1 plasmid, a characteristic of post-Black Death strains first identified in Late Second Pandemic contexts after 1353. The depletion may have reduced transmission rates relative to earlier pandemic strains. It did not make the disease survivable for the five individuals who carried it.

What is striking is not just that these people died of plague, but what their bodies record of the years before they did.

Despite averaging under 18 years of age, the group shows a level of skeletal wear more consistent with a much older and more stressed population. The team applied a modified Index of Skeletal Frailty (ISF), a metric that accounts for accumulated pathological conditions across the skeleton. The average ISF for the Basel group was 3.3. For comparison, 18-to-25-year-olds in medieval London monastic and non-monastic burial contexts yielded an ISF of around 2.09. The Basel subadults were considerably more worn.

Every individual for whom teeth survived showed at least one carious lesion. All scored individuals displayed enamel hypoplasia, transverse bands of disrupted enamel development that form during periods of physiological stress in early childhood. Sixty-one percent of available teeth were affected. Degenerative spinal changes consistent with spondyloarthrosis deformans were found in all 15 individuals. Among those under 15, the changes were described as light but present; those aged 15 to 25 showed medium-stage degeneration. Two individuals had worn their front teeth down from habitual use of the mouth as a tool, a pattern associated with trades that require holding or manipulating objects by the teeth.

Seven healed traumatic fractures appear across four individuals. Two people had congenital conditions: a cleft palate and an idiopathic scoliosis of the thoracic spine.

Who the Hospital Would and Would Not Take

Stable isotope analysis of rib collagen from 11 of the 15 individuals shows a diet dominated by C3 plants and animals that consumed them. The isotopic values cluster closely together, with no meaningful differences between subadults and adults, or between likely males and females. The group ate the same food. It was not much of a diet. Compared with other Basel skeletal assemblages, the δ¹⁵N values sit at the lower end of the distribution, suggesting limited animal protein intake relative to populations with access to better nutrition.

The consistency of the isotopic data matters because it points toward shared social circumstances. These were not individuals from different rungs of Basel’s hierarchy who happened to die together. The bones suggest they came from the same corner of the city’s social world.

That world was structured in ways that made survival during an epidemic a matter of social capital, not just biology. The Bürgerspital, the municipal hospital right next door to the cemetery site, admitted patients with full citizenship status from the city or its surrounding Landschaft, or those with the financial backing of an employer. Domestic servants, even those from the Landschaft, could be expelled from the city if they fell sick. The hospital records confirm that 76 burials took place at the Bürgerspital in 1667 alone, during the last outbreak, compared with a typical annual count of 10 to 60. But those 76 represent only those who qualified for entry.

City physician Felix Platter left a detailed report on the 1610/1611 epidemic in the Basel suburbs. The data, analyzed in the early 1980s by historian V. Lötscher, shows that roughly 77 percent of those who fell ill and died in the suburbs were under 20 years old. Young people in the 15-to-20 age range showed a lethality rate of around 72 percent, compared with 54 percent among adults. Women survived infection at higher rates than men.

These numbers map closely onto the demographic structure of grave group 229-232.

The explanation Rindlisbacher and her colleagues propose is not simply that plague preferred the young. It’s that precarious work conditions prior to and during the epidemic shaped who was already immunologically depleted when Y. pestis arrived. Enamel hypoplasia forming between the ages of three and six reflects physiological stress during precisely the years when, in Early Modern contexts, children from poorer households were being integrated into family labor. The degenerative spinal changes in adolescents point to physically demanding work that left skeletal traces. If you could not stop working, you could not recover from illness, and you could not avoid the crowded conditions that spread the pathogen.

The flip side of this is less visible in the bones but present in the archival record: access to social networks during a crisis determined who received care at all. Citizenship status, respectability, and ties to an employer were not just social niceties. They were the mechanism by which a person gained entry to institutions with shelter and medical attention. Those outside that network died in a cloister garden that wasn’t even recorded in the official list of Basel’s plague cemeteries.

The team is careful to note that the Stadtcasino site is not a simple lower-class cemetery. The diversity of burial practices across the larger assemblage, including coffin burials, personal items, and coins, suggests a range of social positions among the hundreds interred there over the decades. The 15 individuals in graves 229 to 232 represent one moment in that longer story, likely the epidemic of 1667 to 1668, and probably the most socially marginal of the hospital’s associated dead. A broader comparative analysis of the full assemblage is ongoing.

What the study adds, precisely, is the bridge between osteological evidence of hard short lives and the documentary evidence of a city that structured access to survival along lines of citizenship, employment, and social standing. The bones do not lie about the labor. The archives do not lie about the gates.

Further Reading

* Spyrou, M.A. et al. (2019). Phylogeography of the second plague pandemic revealed through analysis of historical Yersinia pestis genomes. Nature Communications, 10. https://doi.org/10.1038/s41467-019-12154-0

* Susat, J. et al. (2020). Yersinia pestis strains from Latvia show depletion of the pla virulence gene at the end of the second plague pandemic. Scientific Reports, 10. https://doi.org/10.1038/s41598-020-71530-9

* Castex, D. & Kacki, S. (2016). Demographic patterns distinctive of epidemic cemeteries in archaeological samples. Microbiology Spectrum, 4. https://doi.org/10.1128/microbiolspec.poh-0015-2015

* Marklein, K., Leahy, R., & Crews, D. (2016). In sickness and in death: assessing frailty in human skeletal remains. American Journal of Physical Anthropology, 161, 208-225. https://doi.org/10.1002/ajpa.23019

* Lewis, M. (2016). Work and the adolescent in medieval England AD 900-1550: the osteological evidence. Medieval Archaeology, 60, 138-171. https://doi.org/10.1080/00766097.2016.1147787

* Lötscher, V. (1987). Felix Platter. Beschreibung der Stadt Basel 1610 und Pestbericht 1610/11. Basel: Schwabe. [VERIFY — cited indirectly through the Rindlisbacher et al. paper; confirm bibliographic details before listing]

* Alfani, G. & Murphy, T.E. (2017). Plague and lethal epidemics in the pre-industrial world. Journal of Economic History, 77, 314-343. https://doi.org/10.1017/S0022050717000092

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe

Somewhere around 3600 BC, a man was laid to rest in a stone-walled tomb beside Loch Calder, in what is now the far north of mainland Scotland. Later, his son was placed there too. Then, some years after that, his son’s son. The bones of all three eventually ended up gathered together on stone benches inside the northern chamber of a cairn called Tulloch of Assery A, the remains partly disarticulated by the time of their final arrangement, coated in clay, handled and repositioned as each successive death brought another man to the same threshold.

That three-generation male line now shows up in ancient DNA. A new study published in Antiquity, led by Vicki Cummings, Chris Fowler, Iñigo Olalde, Sarah Cuthbert, and David Reich, analyzed 40 new aDNA samples from 22 individuals whose remains had been placed across five chambered cairns in Caithness and Orkney, dating from roughly 3800 to 3200 cal BC. The results map a network of biological relationships that reaches across community boundaries, across decades, and across the Pentland Firth.

The five tombs are Tulloch of Assery A and B, Tulach an t-Sionnaich, and Rattar East on the Scottish mainland, and Holm of Papa Westray North on the Orkney island of Papa Westray. Three of the mainland sites cluster within 200 meters of one another around Loch Calder. All three are intervisible. Rattar East sits on the coast, looking out toward the islands. Holm of Papa Westray North is on the other side of that stretch of water.

Each tomb belongs to what archaeologists call the stalled cairn tradition: drystone-walled chambers divided by upright slabs into internal compartments, set within a surrounding cairn. They look, from the outside, broadly similar. But they are not identical. Tulach an t-Sionnaich has a single box-like chamber. Tulloch of Assery A has two opposing chambers. Tulloch of Assery B is tripartite. The shared form announces membership in a common architectural tradition; the differences, the team argues, signal something equally deliberate.

What the Bones Say, and What They Don’t

The osteological picture at each site is difficult. Excavations recovered a combined minimum number of 37 individuals across the five tombs, but the assemblages are fragmentary, commingled, and in most cases heavily disturbed. At Tulach an t-Sionnaich, the human remains were sandwiched between a stone layer and a layer of animal bone — cattle, deer, bird, fish, mollusc — and intermixed with limpet shells, flint, and pottery. At Tulloch of Assery B, an adult skeleton had a leaf-shaped arrowhead embedded in a thoracic vertebra. At Holm of Papa Westray North, at least 14 individuals are represented, including four children, but much of the assemblage had been disturbed well before scientific excavation began, likely by antiquarian digging in the nineteenth century.

aDNA analysis was possible for only 22 of those 37 individuals. From that subset, the team identified nine pairs of close genetic relatives at the third degree or closer. All nine were male-to-male pairings. Beyond those close relationships, shared identity by descent analysis detected around 100 more distant relative pairs, including 11 pairs sharing enough DNA to indicate they were likely fourth- or fifth-degree relatives.

The specific relationships matter. At Tulloch of Assery A: the father-son-grandson sequence just described. At Rattar East: two brothers. At Holm of Papa Westray North: a father and son, plus a man who was the father’s maternal uncle or half-brother. And connecting the two areas of Loch Calder: the male buried at Tulach an t-Sionnaich was likely the paternal uncle, half-brother, or grandfather of one of the males at Tulloch of Assery A. The man at Tulach an t-Sionnaich and the senior male at Tulloch of Assery A may have been close enough kin that their communities recognized a shared patriline, even as they built separate monuments.

None of that, the authors are careful to note, means patrilineal descent was the only thing going on. Kinship is more complex than any single axis of relatedness. The closest relationship between any two women in the study was fifth degree. No mother-daughter pairing was detected. But the data do show a consistent pattern: in Caithness especially, it was fathers and sons, brothers, uncles and nephews, who ended up sharing the same stone chambers across multiple generations.

What makes the Holm of Papa Westray North picture stranger is this: two of the women buried there were more closely related, genetically, to men from the Loch Calder tombs than to any of the men buried alongside them on Orkney. One woman was a fourth-degree relative of the man at Tulach an t-Sionnaich. Another was a fourth-to-fifth-degree relative of the same man. Neither was closely related to the three principal males at Holm of Papa Westray North. Whatever brought them to that island tomb, it was not proximity to their nearest male kin.

Stone, Distance, and the Work of Memory

The team frames the tombs not merely as repositories for the dead but as what they call “technologies of descent and affinity.” Building a tomb in the style of existing tombs was itself an act of social positioning. Each time a new cairn went up near Loch Calder in the 3600s to early 3500s cal BC, those building it were placing themselves in explicit visual and material relationship with what was already there. The proximity was not accidental. The Tullochs of Assery are 30 meters apart.

There is something specific and deliberate in the arrangement at Tulloch of Assery A that speaks to how memory worked in these communities. The three generations of males had presumably been placed in the tomb at different times, each man carried in upon death. Yet the bones ended up gathered together on the same stone benches, some coated in clay, suggesting the disarticulated remains had been handled and intentionally repositioned. Relatedness was being actively maintained as a physical fact, remembered and reenacted in the arrangement of bone. The building and use of the tomb was not a single event but an ongoing practice.

The chronology, worked out through radiocarbon dating in combination with the genetic relationships, suggests the Loch Calder tombs were built and used in fairly rapid succession during the thirty-seventh and thirty-sixth centuries BC. The main phase of use at Holm of Papa Westray North comes somewhat later, with a father-son pair likely interred somewhere in the mid-3300s. The primary small single-chamber phase at Holm of Papa Westray North probably predates even those individuals. The tomb on the island began as a single box-like chamber within a round cairn, later extended into a four-stalled monument. It was rebuilt to accommodate more dead, or more categories of dead.

In Orkney, something different was happening with space. Orcadian stalled cairns were not typically built in tight clusters near one another. Even the concentration of cairns on the island of Rousay are not intervisible. And in Orkney, dry-stone walled houses appear alongside the cairns in the landscape, drawing on the same construction techniques and materials as the tombs themselves. The living and the dead occupied architecturally similar structures, but those structures were distributed differently across the terrain. In Caithness, the logic seems to have been aggregation around the dead. In Orkney, dispersal was the pattern for both.

The team interprets this as an emerging difference in how kin groups related to one another. Both regions shared a common architectural tradition and, the DNA confirms, maintained real biological connections across the Pentland Firth. But within a few centuries, communities on either side of that strait were expressing kinship differently through stone. Caithness communities built tombs near existing tombs. Orcadian communities moved away from existing sites, building new monuments and houses in new locations. The shared tradition was real. The divergence was also real.

Whether those two women at Holm of Papa Westray North had crossed the water themselves, or whether their ancestry simply reflected older patterns of movement between the regions, the DNA cannot say. Their presence in an Orcadian tomb, more closely tied genetically to men in mainland cairns than to the men beside them, suggests that the relationships between these communities were neither simple nor static. The tombs were built to last. The people in them had lives that ran in multiple directions at once.

Further Reading

* Fowler, C. et al. (2022). A high-resolution picture of kinship practices in an Early Neolithic tomb. Nature 601: 584–87. https://doi.org/10.1038/s41586-021-04241-4

* Olalde, I. et al. (2018). The Beaker phenomenon and the genomic transformation of north-west Europe. Nature 555: 190–96. https://doi.org/10.1038/nature25738

* Corcoran, J.X.W.P. (1967). The excavation of three chambered cairns at Loch Calder, Caithness. Proceedings of the Society of Antiquaries of Scotland 98: 1–75.

* Cummings, V. & Fowler, C. (2023). Materialising descent: lineage formation in early Neolithic southern Britain. Proceedings of the Prehistoric Society 89: 1–21. https://doi.org/10.1017/ppr.2023.2

* Fowler, C. (2022). Social arrangements. Kinship, descent and affinity in the mortuary architecture of Early Neolithic Britain and Ireland. Archaeological Dialogues 29: 67–88. https://doi.org/10.1017/S1380203821000210

* Ritchie, A. (2009). On the fringe of Neolithic Europe: excavation of a chambered cairn on the Holm of Papa Westray, Orkney. Edinburgh: Society of Antiquaries of Scotland.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.anthropology.net/subscribe