Evolutionary Insights by Anthropology.net

On the Danish islands, the local ancestry that had been widespread between about 2200 and 1600 BP had essentially disappeared by 1250 BP, roughly a century before the Viking Age. (BP means calibrated years before present, with the present fixed at 1950 CE.) In its place, the Scandinavian-related ancestry of most individuals is best modelled by an Iron Age source from northern Germany or southern Jutland. That is the central finding on Viking ancestry in a new ancient DNA study of how the Germanic languages spread. It was published with a companion paper that applies the same data to the origin of the Celtic languages.

The usual picture has Viking Age Danes and Swedes descending from people settled in southern Scandinavia since the Bronze Age, speaking a language that developed in place. Earlier genetic studies detected southern ancestry arriving but, the authors note, generally assumed local continuity. The new work says that continuity broke in Denmark and southern Sweden during the Migration Period (1575 to 1375 BP), when Old Norse, the language of the Vikings, took shape.

Both papers appear in Nature. Hugh McColl is first author on both, sharing that position with the linguist Guus Kroonen on the Celtic paper, and Eske Willerslev is last author. The team sequenced 712 new ancient genomes and combined them with published data into a working set of 4,587 individuals spanning roughly 5,000 years. The Germanic paper proposes that a population it calls Eastern Scandinavian, linked to the northeastern Baltic, reached Sweden by about 4000 BP and carried the earliest form of Germanic. That is some 800 years after the Corded Ware groups traditionally credited with bringing it. The paper also traces most of the Scandinavian-related ancestry of the Anglo-Saxons, Frisians and Langobards to Jutland and northern Germany.

The companion paper tests four competing models for how Celtic reached an Iron Age range that ran from Iberia to the Balkans. It detects three large movements of people between 4400 and 2800 BP. It judges the last of them to be the most parsimonious vector for Celtic, meaning the one that needs the fewest extra assumptions: a westward spread from eastern central Europe between 3200 and 2800 BP, associated with the cremation-dominated Urnfield Culture. That puts the paper behind the longstanding Celtic-from-the-East model. In both papers the most important conclusion sits beside the largest gap in the sampling, and that shapes how far I would trust each one.

How does ancient DNA trace the spread of a language?

DNA cannot show what anyone spoke, as both papers acknowledge, but it does show who descended from whom. The method uses identity-by-descent (IBD) segments, which are stretches of chromosome that two people share because they inherited them from a common ancestor. Recombination shortens these stretches every generation, so long segments mark recent relatives and short ones deeper links. Many genomes here are thin, down to 0.1× coverage, meaning only about one position in ten was read even once. The team filled in the missing genotypes by imputation from a reference panel of present-day genomes.

In IBD mixture modelling, each individual gets a palette recording how much DNA they share with each of 386 genetic clusters. A target’s palette is then fitted as a blend of the palettes of chosen source groups. Because every palette draws on the whole dataset, a Viking Age Dane can be linked to an Iron Age source population even when the two share few segments directly.

Applied to Scandinavians older than 2800 BP, the method finds three clusters. The Eastern Scandinavian one carries extra hunter-gatherer ancestry best matched by hunter-gatherers from Latvia and Lithuania. The authors attribute it to an unsampled population between Finland and northeastern Sweden; no genomes yet exist from that region and period. Between about 3200 and 2450 BP, early Germanic took loanwords from Celtic to the south and lent words to Finno-Saamic languages to the east. Neither of the other clusters reached both contact zones, so the authors call Eastern Scandinavian ancestry the most parsimonious vector.

Tracking later movements is harder, because everyone who postdates the arrival of Steppe ancestry in Europe is so closely related. The Celtic paper’s solution, which I think is the smartest move in either study, is to follow which regional variety of older Neolithic farmer ancestry people carried. Bell Beaker groups absorbed different local farmers in different places, so a source of confusion becomes a regional tag. In England, ancestry tied to Neolithic farmers of Iberia and France overtakes local farmer ancestry between 3800 and 3200 BP. That matches an earlier report of large-scale migration into Britain (Patterson and colleagues, 2022). Between 3200 and 2800 BP, ancestry tied to Bronze Age Hungary and Serbia rises in England, becoming widespread by at least 2600 BP.

Iberian individuals from Urnfield contexts carry that eastern ancestry by 3150 BP. The authors use ancestry related to the Knovíz Culture of the Czech Republic as a stand-in for Urnfield populations generally. By 2800 BP it is common across former Urnfield regions, though its effect weakens westward and the older Bell Beaker-related ancestry still dominates in Iron Age England. The authors then checked each movement against published linguistic dates. These covered the formation of Proto-Celtic, the undocumented ancestor of all Celtic languages, and the later split between its Brittonic branch (including Welsh) and Goidelic branch (including Irish). Only the third movement fit most estimates on every count.

Why does cremation leave gaps in the ancient DNA record?

What I most want readers to notice is where the missing genomes fall. Both papers place their most consequential change in centuries when most people were cremated, and cremation destroys DNA. For the Celtic paper this is the Urnfield period of 3200 to 2800 BP. The authors say the timing of the transition in France cannot be measured directly for lack of Late Bronze Age samples, so they infer it from Britain and Iberia.

For the Germanic paper, the gap runs from 1575 to 1200 BP, when Old Norse formed. I counted the individuals in the paper’s ancestry plot for Denmark and Sweden (Extended Data Fig. 13). The panel for 1575 to 1230 BP holds 15 people, against roughly 200 on each side, and only four are Danish genomes older than 1250 BP. Because hundreds of genomes anchor each end, I accept that the population of the Danish islands changed, but the genomes cannot say when within the window. The paper’s date, probably 1550 to 1450 BP, comes from non-genetic evidence, including tree rings from 654 Danish oak samples.

Norway is the result I think historical linguists should weigh most. Old Norse took shape there too, yet the paper finds genetic continuity in Norway from the Iron Age to the Viking Age. The authors conclude that demographic and cultural mechanisms combined to drive the language. They also observe that the later boundary between East Norse and West Norse dialects roughly follows the boundary between the newcomers’ ancestry and Norwegian Iron Age ancestry, even in Viking areas of Britain. That is a firmer link between genes and language, because it concerns a split inside a language rather than its birth. The authors also suggest the expansion through Denmark may partly explain why Old Norse speakers called their language dǫnsk tunga, the Danish tongue.

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 last centimeter of hair from a girl of about eight, buried on the volcano Chuscha in Argentina during the Inca Empire, holds a chemical missing from the five centimeters grown before it. That chemical is cocaethylene, which the liver makes when cocaine and alcohol are in the blood together. Its arrival suggests that in roughly the final month of her life this child was drinking heavily while taking coca. Earlier tests on the same mummy found no cocaine at all.

The result comes from new drug testing on Inca mummies, and it complicates a picture many readers carry of the capacocha, the rite in which the Incas sacrificed children and young women to serve as messengers to the gods. The colonial chronicles read almost like a recipe, with chicha (maize beer) to calm the child and crushed coca placed in the mouth before the killing. From there it is easy to assume one state ritual with one drug routine. This paper supports part of that picture and raises doubts about the rest, though less firmly than its press release suggests.

What drugs were found in the Inca mummies?

Dagmara Socha of the University of Warsaw, Marzena Sykutera of Nicolaus Copernicus University in Toruń and six colleagues report in Antiquity that coca markers turned up in all six hair samples they tested and the alcohol marker in four of those six. Harmine, an alkaloid linked to ayahuasca, appeared in one nail from one person. They searched for ten compounds in eight individuals from five volcanoes in Peru and Argentina. None of the eight showed harmaline, mescaline (from San Pedro cactus), N,N-DMT or bufotenine (from Anadenanthera and Virola), or scopolamine or atropine (from Brugmansia and Datura).

Two of the eight came from Ampato in Peru: Ampato #1, a girl of 14 known as the Ampato Maiden (the press release calls her the Lady of Ampato, or Juanita), and Ampato #4, a girl of about 10. Sara Sara, also in Peru, gave a girl of about 14. In Argentina, Chuscha gave the girl of about eight, Quehuar a girl of about 10, and Llullaillaco three children: the Doncella, about 14; the Younger Girl, about six; and the Boy, seven. The main text gives no date for any of them, only the span of the Inca Empire, from the late thirteenth century to 1572 CE.

The method depends on keratin, the protein in hair and nails, which traps compounds circulating in the body. Hair grows outward from the scalp, with each centimeter standing for about a month, so a strand can be read as a rough timeline. Nails can’t, because compounds enter them by more than one route. The team washed each sample, then ran an extract through liquid chromatography-tandem mass spectrometry, which separates the compounds in a sample and identifies each by its mass and the fragments it breaks into. Results come in picograms per milligram (pg/mg), or trillionths of a gram in each thousandth of a gram of sample.

The results span about four orders of magnitude. Cocaine measured 16,740 pg/mg in the Doncella, who was also found with a coca quid (a ball of dried leaves) in her mouth, 1,410 in the Boy and 960 in the Younger Girl. It measured 31 in the last centimeter of the Chuscha hair, 2.5 in Ampato #4 and 1.5 in Sara Sara. Benzoylecgonine, the main product of the body’s breakdown of cocaine, accompanied it each time. Cocaethylene appeared only in the Llullaillaco children and Chuscha. The two nails gave harmine at 0.6 pg/mg in Ampato #1 and no target compound in Quehuar.

I find the coca result the strongest in this paper, because the drug and its breakdown product appear together in every hair sample, and quids and cheek-bulged figurines agree. It also reverses earlier negative tests on Chuscha and Sara Sara. The authors credit improved sensitivity, which seems likeliest, although Sara Sara’s 1.5 pg/mg is a small number to carry a reversal.

Chuscha is the only sample read segment by segment, across six 10 mm sections covering about six months. Cocaine ranged from 135 to 342 pg/mg in the five older segments, then fell to 31 nearest the scalp, where cocaethylene appeared at 57. The authors estimate cocaethylene at about 166 percent of cocaine in the blood and suggest the shift marks entry into the final stage of the ritual. They concede that the variation could be an artifact of preservation and metabolism. I trust the appearance of cocaethylene more than the fall in cocaine, because the cocaine series already jumps from 135 to 342 between neighboring months, while cocaethylene is absent from five segments and then present. This is still one strand from one child, and grave robbers took her body in the 1920s and sold it into private collections before it reached the museum in Salta. The paper doesn’t discuss what that history may have done to the hair.

Did Inca children take ayahuasca before sacrifice?

The ayahuasca claim is the one most likely to travel, and it rests on the smallest number in the paper. Ayahuasca refers to the Banisteriopsis caapi vine and to drinks made from it, often combined with a second plant that supplies N,N-DMT. Harmine blocks the breakdown of psychedelic compounds, so it can prolong the effects of that second plant. The authors argue that Banisteriopsis or a related Amazonian plant is the most probable source of the Ampato #1 harmine, which would imply long-distance trade. They dismiss Passiflora as a source but don’t separately rule out tobacco.

I don’t think this measurement can yet carry an argument about one mountain’s ceremonies, because the comparison between mountains is confounded by tissue. Harmine appeared in no hair sample from anyone, only two nails were tested, and Chuscha, Sara Sara and Llullaillaco contributed none. The authors report contamination in the nails, probably environmental, that made results hard to interpret, and they raise the possibility that nails and hair take up these compounds differently. Then they read the absence of harmine in the hair of Ampato #4 as a possible sign of different treatment or a separate ceremony. CT scans do suggest she was a secondary burial (Socha et al. 2026), but her hair can’t say much about harmine when no hair tested positive.

What keeps me from dismissing the finding is earlier work by the same group, which reported harmine and harmaline in two other children from the mountain, Ampato #2 and #3 (Socha et al. 2022b). That makes three individuals from one mountain across two studies, though this paper doesn’t say which tissue the earlier results came from. Dr Dagmara Socha, lead author of the research, from the University of Warsaw, said in a statement from Antiquity that the harmine,

“may indicate that a harmine-containing beverage had a specific role in ceremonies connected with this particular sacred mountain.”

The authors’ next step is where I part with them. They suggest the vine may have been taken alone for its medicinal value, possibly to calm the children, and say the absence of other hallucinogens supports this. The same discussion warns that tryptamines such as N,N-DMT enter keratin poorly, so a negative test for them is weak evidence that they were absent. A modern study of ayahuasca (Osório et al. 2015) also can’t show what Inca ritual specialists intended for a 14-year-old, and a 0.6 pg/mg trace with no timeline says nothing about dose.

Was the capacocha one standardized ritual?

The paper’s largest claim is that the differences may reflect distinct mountain deities and local variation, and the design can’t separate that from simpler explanations. The mountains differ in tissue tested, length of hair analyzed and burial history. The Llullaillaco and Chuscha figures cover only the 10 mm nearest the scalp, while Ampato #4 and Sara Sara were analyzed whole, and Chu…

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

At position 996 of a collagen chain called COL1A2, the protein extracted from a right parietal bone from Xujiayao, a Middle Pleistocene site in the Nihewan Basin of northern China, carries lysine in 57 distinct peptides and arginine in one. The second Xujiayao specimen sampled, a mandibular ramus, gives 61 lysine-bearing peptides against a single arginine one. Every Denisovan with protein or genome data carries lysine at that position, while Neanderthals and modern humans carry arginine, and those two bones are the main reason the Xujiayao hominins are now being called Denisovan.

Since the fossils came out of the ground between 1976 and 1979, Xujiayao has been filed as Homo sapiens neanderthalensis, as early H. sapiens, as archaic H. sapiens, and, to avoid the argument, as archaic Homo. Most recently it became the holotype of Homo juluensis, a species proposed partly on the 1,700 ml endocranial capacity of the Xujiayao 6 cranium (Wu et al., Journal of Human Evolution, 2022). If you follow East Asian Middle Pleistocene hominins, you probably carry a picture of Dali, Jinniushan and Hualongdong as archaic humans with unsettled names, and of Denisovans as a lineage known from a genome plus a few molecularly confirmed jaws and one cranium.

The paper appeared in Nature on 21 October 2026. Song Xing, Zhongyou Wu, Qingfeng Shao and Junyi Ge are the equal-contribution first authors, and Qiaomei Fu of the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing is the corresponding author. The team sampled two of the 20 Xujiayao cranial, mandibular and dental fossils, recovered 18 endogenous ancient proteins, and concluded that the assemblage belongs to a late Middle Pleistocene East Asian Denisovan population. They then compared skull and jaw anatomy across East Asian archaic Homo and argued that most of those fossils, Dali, Jinniushan and Hualongdong included, are Denisovan as well, and that the size gap behind H. juluensis is variation within one population. This post is about how much of that sweeping reclassification two bones and one amino acid can carry.

How do researchers know the proteins are ancient?

The sampling was minimal. The team CT-scanned both specimens first (37.56 μm resolution), then drilled a small hole in each in a dedicated clean room, shielding the surrounding bone with aluminium foil. They dissolved the powder, split the proteins into fractions, and digested them with trypsin, an enzyme that cuts proteins into short peptides. A mass spectrometer then reads the mass of each peptide fragment, which allows the amino acid sequence to be inferred. The runs were done in two independent laboratories, one at Capital Medical University and one at Fudan University, and the data were searched with two software packages, PEAKS Online and MaxQuant. Extraction blanks were processed alongside the samples, and a modern camel bone served as a positive control.

The hard part with any ancient protein is separating the real thing from modern contamination, and the authors lean on deamidation. Over long burial, the amino acids asparagine and glutamine slowly lose an amine group, so ancient proteins show high deamidation and fresh contaminants show little. In specimen PA1490 (the parietal), the main bone collagens had asparagine deamidation of 61.4 to 94.7% and glutamine deamidation of 44.9 to 92.6%, while keratins and trypsin, the usual contaminants, sat at 0.0 to 5.3% and 0.0 to 7.7%. PA1497 (the mandible) looks similar. After that filter, plus removal of peptides that could not be tied to a single gene, anything seen in the blanks, and anything MaxQuant could not confirm, 18 endogenous proteins remained: 13 for PA1490 and 15 for PA1497. Figure 2b gives the endogenous share of peptides as 91.4% for PA1490 and 94.1% for PA1497, and the extraction blank scores 0.0.

The taxonomic call comes in two steps. Searches against a mammalian type I collagen database placed both samples in the genus Homo, and the team then looked at single amino acid polymorphisms (SAPs), positions where sequences differ between lineages. The informative one is COL1A2-996. The paper’s Figure 3 shows the supporting peptides, 207 peptide-spectrum matches (individual spectra assigned to a peptide) for PA1490 and 244 for PA1497. The lysine peptides also show elevated deamidation at nearby residues (Asn977 at 69.2% and 69.4%, Gln1002 at 100% in both), which argues that they are genuinely ancient. The authors also built a Bayesian tree from 2,074 amino acid positions in eight proteins, with Denisova3, the Altai and Vindija 33.19 Neanderthals, 50 present-day humans from the Human Genome Diversity Project, and a chimpanzee (Pan troglodytes) as outgroup. Xujiayao and the Denisovans form a clade with posterior probability above 99%.

Within that clade the tree gives nothing: individuals fall into unstable arrangements with posterior probabilities below 50%, because no variant other than COL1A2-996K is shared. That is why the authors say only that Xujiayao is Denisovan in a general sense, without placing it relative to Xiahe, Penghu, Harbin or Denisova Cave. The dating, for what it is worth, is 199 to 138 ka from combined uranium-series and electron-spin-resonance methods (Shao et al., Quaternary Science Reviews, 2026), which makes Xujiayao about contemporary with the earliest confirmed Denisovans from northeastern China and southern Siberia.

Does skull shape make Dali and Jinniushan Denisovan too?

This is where I slow down, because the molecular result covers two bones and the morphological claim covers most of East Asia. With Xujiayao and Harbin both classed as Denisovan, the authors characterize a Denisovan skull. It has erectus-like traits (a low, wide neurocranium; a vault 7.7 mm thick at bregma and 12.0 mm at the parietal tubercle on Xujiayao 6), derived traits shared with later Homo (a strongly arched temporal squama), and a combination found in neither H. erectus nor modern humans: a low profile with near-parallel lateral walls, a nearly vertical but non-rounded occipital, and a mandible with a depression in the planum triangulare. They then check whether Dali, Jinniushan, Hualongdong and Xuchang show the same suite, and conclude that they do.

I find the Xujiayao assignment far stronger than that extension, for a reason the paper concedes itself: the suite “does not appear simultaneously in individual specimens but manifests in different forms or combinations across various specimens.” A definition that tolerates different combinations in different fossils is hard to falsify, and the definition was built from Xujiayao an…

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

Seven and a half meters down in Shanidar Cave, in the Zagros Mountains of Iraqi Kurdistan, the left shoulder blade, or scapula, of a Neanderthal had lain for some 75,000 years with a hole through it. The bone was too soft to pick up. The excavators soaked it where it lay with an acrylic consolidant, Paraloid B72 dissolved in acetone, backed it with cotton gauze, and cut it out of the ground inside a block of its own sediment; it had come through nearly whole, which is rare for a bone so thin in places that it is almost translucent. The hole sits low on the back of the blade, in the broad hollow anatomists call the infraspinous fossa. It is about 15 millimeters across and very nearly a perfect circle, and at least three hairline cracks run out from its rim.

“It’s so round, it’s like, what is it?”

That was Emma Pomeroy’s first thought, she told Science News. Pomeroy, a biological anthropologist at the University of Cambridge, is lead author of a study published October 15 in Scientific Reports that describes these bones, dug up in 2019 and 2022, and the person they belonged to, a Neanderthal the team calls Shanidar A. Anyone raised on the phrase Stone Age pictures a killing in this period done with stone, a flake of flint knapped to a point and bound to a shaft. Neanderthals did use weapons like that, but a hafted stone point leaves a different mark. When archaeologists drive replica stone points into carcasses, the wounds come out oval or lozenge-shaped, sometimes a narrow slot, following the flattened section of the blade, whereas a sharpened wooden shaft, round in section, leaves a circle.

Shanidar has encouraged the stone picture before. A man excavated there in the 1950s, known as Shanidar 3, has a narrow puncture in his left ninth rib, and in 2009 Steven Churchill, an evolutionary anthropologist at Duke University, and his colleagues tried to reproduce it by driving replica Mousterian and Levallois points into the chests of pig carcasses, using a calibrated crossbow to control the energy of each strike. They judged the wound most consistent with a light, stone-tipped projectile and, because early modern humans were in western Asia at around that time and may already have had throwing weapons, raised the possibility that one of them threw it. The rib had started to heal, so Shanidar 3, who died roughly 50,000 years ago, some 25,000 years after Shanidar A, seems to have lived for weeks after he was hit. The older bone points to a different weapon. The University of Cambridge’s announcement presents the new hole as possibly the first evidence that sharpened wooden spears were used anywhere in the prehistoric Middle East, and possibly the earliest example in the hominin fossil record of lethal violence between individuals with a wooden spear, a second claim the paper itself stops short of.

On October 7, in Science Advances, Palmira Saladié and colleagues at the Catalan Institute of Human Paleoecology and Social Evolution described 597 pieces of wood from Abric Romaní, a rock shelter near Barcelona that Neanderthals used between about 62,000 and 40,000 years ago. Much of that wood survives only as the shape of where it was: water heavy with calcium carbonate wrapped the pieces in travertine as they rotted, and the excavators recorded the molds. Thirteen had been worked or used, among them two spears, the better preserved 2.45 meters long. Josep Vallverdú, an archaeologist at the institute, told Gizmodo that the simplest lesson of the site was that the Paleolithic was “not a world made only of stone,” a strange thing to need to say about a period whose name means Old Stone Age. It needs saying because stone is what lasts, so stone is what archaeologists mostly have; wooden shafts, handles and spears turn up only where some accident of chemistry has kept them.

Shanidar has no travertine and no waterlogged mud. Apart from pollen and charred wood, the paper notes, the cave has kept almost nothing organic, so the spear that went through Shanidar A is gone, and the only record of it is the space it took up on its way out of the body. That is why the hole has been worth the trouble the Cambridge team has taken over it. Read closely enough, it is a cast of a weapon nobody will ever hold, and perhaps a little more than that, a record of the direction the weapon came from, which is as close as the evidence gets to whoever was at the other end of it.

A root can bore through buried bone, and so can a burrowing insect or a careless trowel. Because the scapula came out with the sand beneath it still in place, the team could scan the whole block by micro-computed tomography (micro-CT) and look down through the hole into the sediment, where there was no root channel and no burrow of the right width. The rim carries the same crust of manganese oxide as the rest of the bone, so the hole is old, and its edges show no remodeling and none of the disorganized woven bone a living body lays down when it starts to repair a wound. The cracks running from it, the kind that form when fresh bone gives way under a sudden load, put it at the time of death; perimortem, in the paper’s term.

A video of a 3D rendering of the hole in the scapula and fragments embedded in the sediment underneath the scapula in situ prior to conservation. Credit: Dr. Emma Pomeroy

Then the scans showed something better than an absence. In the sand directly under the hole lay at least four slivers of bone, each at least 4 millimeters long. One of them can be fitted back into the edge of the hole with confidence, because a faint ridge on the inner face of the scapula, a linea muscularis, where muscle once anchored, runs from the blade across the break and onto the fragment, the way a road lines up again across the two halves of a torn map. The others probably came from the hole as well, but they are small and brittle, and some broke further during recovery and conservation, so where each belongs in the rim can no longer be worked out. Whatever went through the bone punched them out and carried them a short way into the soft tissue on the back of the shoulder blade, and there they stopped. As the body decayed, the fine sand it lay in probably trickled into the spaces the flesh left behind and held them in place, and when floodwater later tore through and carried off most of the skeleton, the slivers stayed where they were, sheltered by the ridge of the scapular spine, which lay upstream of them.

On the back of the blade the rim of the hole flares outward in a shallow bevel. Bone fails rather like glass struck by a pellet, chipping away in a cone on the side the force leaves by, so a bevel marks an exit, and this one says the weapon came through the chest from the front and left through the shoulder blade. The slivers add the angle. They were pushed slightly outward, toward the arm, and slightly down, toward the feet, and they lie tilted away from the center of the hole, from which the authors infer that the weapon entered from above and from nearer the midline of the body, heading down and outward to the left shoulder blade; if it was then pulled back out through the front, tha…

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

At the Ōzuka tomb, five horses parade across the stone gateway leading into the burial chamber. Their eyes are large and deliberate. Their harnesses are picked out in careful color. You can tell, from the way the artist rendered their bodies, which animals are male and which are female. Riding them are six human figures, and every one of them is a stick. No faces. No sex. One rider is drawn at roughly a quarter the size of his own horse.

This wasn’t a failure of skill. The same hands that gave the horses anatomical precision chose, apparently on purpose, to leave the people blank. A new study by Claudia Zancan, published in Antiquity, argues that this wasn’t an isolated quirk of one tomb but a consistent regional pattern across the decorated tombs of Kyushu, Japan’s southernmost island, and that the pattern tells us something specific about how these communities thought about death.

A Regional Style That Refused to Individualize

Kyushu sits closest to the Korean Peninsula of any part of Japan, and during the sixth and seventh centuries it absorbed plenty of continental influence. Across the water, Koguryŏ tombs painted the dead as dominant, detailed figures towering over their servants and animals, exactly the kind of imagery you’d expect from an elite burial tradition that wanted to broadcast the status of the person inside. Kyushu had access to that visual vocabulary. It didn’t use it.

Zancan’s study systematically catalogued anthropomorphic figures across 38 decorated tombs, 20 painted and 18 rock-cut, and found the same thing again and again: human figures rendered smaller than the objects around them, stripped of individualizing detail, and, in every single case in the sample, impossible to sex. Not one figure across the entire corpus shows primary or secondary sexual characteristics, even in tombs where the artists were clearly capable of that level of detail. Compare that to contemporary Kyushu funerary sculpture like sekijin stone figures or the haniwa of the Kanto region, both of which do occasionally mark sex and social role. The anonymity in the tomb paintings looks like a choice, not a limitation.

This matters because the period itself was a moment of real transformation in Japanese mortuary practice. Older Kofun-period burials were sealed pits, built once and closed for good. The horizontal stone chambers that appear in the Late Kofun period could be reopened, over and over, for successive interments. A tomb stopped being a monument to one death and became a space for ongoing ritual. Zancan’s argument is that the imagery inside these chambers reflects that shift. If a tomb is going to be reopened for generations, its decoration doesn’t need to commemorate the specific person laid to rest last. It needs to do something else.

Borrowing a classification system originally built for Iberian rock art, Zancan mapped where the human figures actually sit within the architecture of each tomb, and three locations kept recurring. On the exterior façades of rock-cut tombs, figures appear holding weapons, facing outward, and these are the only figures in the entire corpus with anything resembling faces. They read as guardians, warding off whatever might approach from outside. Inside, at the entrance to the burial chamber, a second set of figures marks the threshold itself, the physical and symbolic boundary between the world of the living and the world of the dead. This is architecturally and mythologically loaded ground in Japanese tradition. It’s the same threshold where the god Izanagi, in the origin myth, searches for his dead wife Izanami and finds her transformed by the underworld. And at the very back of the chamber, on the rear wall closest to where the body would be laid, the human figures shrink to their smallest scale of all.

None of these three positions functions as a portrait spot. Read together, they sketch a route: guarded threshold, liminal crossing, final resting place. The figures aren’t depicting who died. They’re depicting the passage every dead person takes to become an ancestor.

What the Human Figure Was Actually For

There’s a reflex, when looking at a burial, to go looking for the individual. Sex, rank, personal effects, some marker of who this specific person was in life. Zancan’s study is essentially an argument against applying that reflex here. The tombs of Kyushu weren’t organized around commemorating a particular Homo sapiens life. They were organized around a process, a transition from person to ancestor, and the human figures in the paintings were symbolic participants in that process rather than likenesses of anyone in particular.

That reframes what the small size and blank faces are actually doing. They aren’t evidence of an underdeveloped visual tradition sitting in the shadow of more sophisticated continental art. They’re evidence of a community using abstraction on purpose, choosing anonymity as a way of expressing something collective rather than personal. Scale, in these tombs, seems to track ritual importance rather than realism. The horses at Ōzuka are large because horses carried real symbolic weight in the funerary procession being depicted. The riders are small because whoever they represent doesn’t matter as an individual. What matters is that someone is making the journey.

Zancan notes that this pattern has never been systematically tested outside Kyushu, and much of the physical evidence needed to test it is disappearing. Pigment degrades. Stone erodes. Some of these tombs have already lost the imagery that would let anyone verify whether the same rules of anonymity and placement hold elsewhere in Japan. Whatever answer that question eventually gets, it’s going to depend on how much of this material survives long enough to be looked at again.

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 drained lakebed in the mountains of Jotunheimen doesn’t look like much: gray silt, exposed stone, the odd meltwater channel cutting through it. But in 2022, when the water level at Lake Tesse dropped for its usual spring maintenance, someone walking the exposed mud spotted wood poking out of the ground. Not driftwood. Poles. Dozens of them, still standing upright where they’d been driven into the lakebed, arranged in a tight circular chamber with a fence of angled supports leading into it. A fish trap.

That alone wouldn’t be remarkable. People have been building lattice-wall fish traps in Scandinavia for millennia, and the type is well documented from wetland sites across northwestern Europe going back to around 4000 BC. What made this one strange was the radiocarbon date.

The trap was closer to 4950 BC. And it sat in a lake that, by every reconstruction of the region’s glacial and hydrological history, should not have had fish in it at all.

A landscape with a locked door

Lake Tesse sits at 854 meters above sea level in the central Norwegian mountains, fed and drained by a river system that eventually reaches the Oslofjord. The ice sheet that covered this terrain for most of the last glacial period didn’t finish retreating from the area until around 8400 BC. Once it did, the landscape opened up fast, both to colonizing plants and animals and to the hunter-fisher-gatherer groups moving in from the coasts. Fish followed too, but only where they could. In the lowlands, trout, pike, and a couple dozen other species swam upstream from the sea and from the shrinking freshwater and brackish remnants of the Baltic as relative sea levels dropped through the Early Holocene.

Above a certain elevation, that migration hit a wall. Literally: waterfalls and steep gradients along the Tessa River made it physically impossible for fish to swim up into Lake Tesse from the watercourses below. Historically, the lower stretches of that same river system supported around twenty-five species. The upper stretches, cut off by the falls, supported exactly one: brown trout. Not because brown trout were better adapted to alpine conditions than everything else. Because brown trout were the only species that ever got up there, and they didn’t get up there on their own.

That’s the setup the new paper, published in Antiquity by Axel Mjærum and colleagues at the University of Oslo, is working with. If fish couldn’t climb the falls, and Lake Tesse had fish anyway, something else put them there. The obvious candidate, in a lake that’s been renowned since the twentieth century as one of Norway’s most productive trout fisheries (it once yielded close to nine tonnes of fish a year before it was dammed for hydro-power in the 1940s), is people. The trap dating just pushes back how far that “something else” goes. Considerably.

What the poles actually show

The excavation recovered four traps in total, with 73 individual poles subjected to dendrological and radiocarbon analysis. The construction itself is unglamorous and entirely functional: small stems of Scots pine, with some birch and willow mixed in, averaging about 37 millimeters in diameter, driven vertically into the lakebed to form a circular chamber roughly 3.5 meters across, braced by poles set at 30 to 45 degree angles. Cut marks and traces of lashing show the wood was deliberately worked, not just accumulated debris. It’s the kind of trap ethnographically documented across the Nordic countries well into the twentieth century, built in spring, fished through summer, sometimes maintained and rebuilt over years.

The dating is where it gets interesting. Trap 5 comes back at roughly 4950 BC via radiocarbon on its outer tree rings, placing it squarely in the Late Mesolithic. Trap 2 dates to around 5000 BC as well. Trap 1, dated both by radiocarbon and by cross-matching against a regional pine tree-ring chronology from Jämtland, Sweden, comes out at 3850 BC, with two of its posts felled about twelve years apart, suggesting the structure was periodically rebuilt rather than used once and abandoned. Trap 3 is later still, around 2700 BC. Altogether, the traps at Tesse span more than two thousand years of use, starting in the Mesolithic and continuing well into the Neolithic, long after farming had arrived in southern Scandinavia but before it had displaced hunting, fishing, and gathering as the dominant subsistence base in this particular region.

The Tesse traps aren’t isolated evidence, either. Burned fish bone from nine other sites across the Jotunheimen and Hardangervidda mountains has produced radiocarbon dates that predate 1700 BC, five of them clustering in the fifth millennium BC, roughly contemporary with the earliest Tesse traps. Where the bones have been identified to species, they’re brown trout, and only brown trout. Given that lowland fishing assemblages from the same period are far more taxonomically diverse, that pattern is hard to explain as fish being caught elsewhere and hauled up into the mountains as preserved food. It looks instead like the same species being deliberately introduced, over and over, at multiple lakes.

Ruling out the alternative explanation matters here, because there is one on the table: birds. Waterfowl are known to carry fish eggs on their feet or feathers over short distances, and it’s been proposed as a mechanism for colonizing isolated water bodies. But the case against it in the Norwegian mountains is fairly clean. If birds were doing the work, you’d expect a scatter of species turning up unpredictably across different lakes, not the same single species appearing again and again with what looks like reasonable synchrony across a two-thousand-year window. Genetic work on modern mountain trout populations backs this up too: their population structure points to repeated, stepwise upstream translocations within river systems rather than one-off natural colonization events.

None of this is unprecedented for later periods. Fishing rights over mountain lakes show up in European texts from the fourteenth century onward, already assuming fish had been introduced. A twelfth-century Norwegian runestone references the practice explicitly. Sedimentary ancient DNA from a Spanish Pyrenean lake recently pushed the confirmed record of human fish introduction back to the seventh century AD. What the Tesse dates do is detonate that timeline. Not centuries. Millennia. And crucially, they place the origin of the practice not with farmers, but with people who had no farming at all.

Why bother moving fish into the mountains

There’s an economic logic buried in this that’s worth sitting with, because it isn’t really about fishing for its own sake. Norwegian Mesolithic hunter-fisher-gatherers, judging from rock art, faunal assemblages dominated by ungulates, and sites clustered along known elk and reindeer migration corridors, put serious cultural weight on big-game hunting. The Langfonne ice patch, just eleven kilometers from Lake Tesse, has produced evidence of reindeer hunting going back to at least the fifth millennium BC. But big-game hunting in mountain terrain is unpredictable and risky. Success rates are low, and a bad trip into the high country can mean a long stretch with nothing to show for it.

Introducing a stable population of brown trout into a mountain lake changes that calculation. Trout are calorically dense, comparatively easy to catch relative to elk or reindeer, and once established, they don’t run away or migrate elsewhere. A fish population effectively domesticates the risk profile of the mountains, turning them into something closer to the resource-rich lowland environments these groups already knew how to exploit. That, in turn, makes it feas…

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

Grave 311 held an eighteen-year-old whose teeth didn’t match their bones.

That mismatch is the whole story, so it’s worth sitting with it before anything else. A second molar forms in early childhood, between roughly three and eight years old, and once it’s mineralized it stays chemically frozen at that age forever. Bone collagen is different. It keeps remodeling throughout life, so it mostly records the years just before death. In this one skeleton, the two tissues told opposite stories. The dentine carried a strong signal of C4 plants, almost certainly millet, the kind of signature you’d expect from a diet common across parts of southeastern Europe. The bone collagen showed nothing of the sort, just the C3-dominated cereal diet, rye and buckwheat, typical of everyone else buried in medieval Vilnius. Strontium and oxygen values from the enamel pointed the same direction as the childhood carbon signal, away from the Vilnius region entirely, toward baselines more consistent with present-day Ukraine or southern Poland.

Put plainly: this person grew up somewhere far to the south, moved to Vilnius sometime in adolescence, and adjusted their diet to match. They died there, buried according to strict Christian custom, with no grave goods at all.

That’s a striking amount of individual biography to pull out of a skeleton. But the reason it matters goes beyond one traveler’s life story. Lithuania was, famously, the last pagan state in Europe. Grand Duke Jogaila didn’t formally adopt Catholicism until 1387. Before that, the standard picture is of a country still cremating its dead, still worshipping outside the Christian framework that had already swallowed the rest of the continent. Grave 311 belonged to a cemetery that shouldn’t, by that picture, have existed yet.

The cemetery is on Bokšto Street in Vilnius’s old town, part of a district medieval chroniclers called the Civitas Ruthenica, the Ruthenian city. Archaeologists excavated more than 500 intact graves there over eight years, and the burial customs are unambiguously Christian: bodies laid supine, oriented east to west, cross pendants and enkolpions turning up in some graves, cremation entirely absent. Radiocarbon dating places the cemetery’s use across roughly 150 years, spanning the second half of the thirteenth century through the fourteenth, meaning this Orthodox Christian community was burying its dead in a formally pagan city for a century and a half before Catholicism ever arrived.

That much was already known from the archaeology. Byzantine-style jewelry, the chaplets, the decorative headpieces made from ornaments stitched to leather and cloth, closely resembled fashions from Halych-Volhynia, a region of the former Kievan Rus’ now split between western Ukraine and southern Poland. The material culture argued for a connection. What it couldn’t do was prove that any given skeleton had actually made that journey rather than simply adopted the style secondhand. That’s the gap isotope analysis is built to close, and it’s why a research team led by Giedrė Motuzaitė Matuzevičiūtė sampled bone, dentine, and enamel from 15 individuals, layering carbon and nitrogen data against strontium and oxygen ratios and comparing everything to a local baseline built from modern plants collected around Vilnius.

A city sorted by sex

Once the isotope data came back, a pattern emerged that the grave goods alone hadn’t shown: mobility in this population was heavily gendered.

Every sampled female returned isotope values consistent with growing up locally. Not one showed the kind of childhood signature that would suggest an origin outside the Vilnius region. Several of these women were buried with grave goods, cowrie shells, glass pendants, the kinds of items that don’t fit neatly with strict Christian burial custom and read more as holdovers from older, non-Christian traditions. The team’s interpretation is that these were local women who converted, plausibly through marriage into Orthodox households, carrying older material habits into new religious ones without fully shedding them.

The men look different. Two individuals, from graves 139 and 226, had strontium ratios that fell outside the Vilnius plant baseline but matched values from Kernave, roughly 34 kilometers northwest and Lithuania’s earlier capital, hinting at movement between the two settlements rather than international migration. A third individual, from grave 214, sat close to grave 311’s outlier values, potentially another migrant from the south, though with a weaker signal. And grave 311 itself is the clearest case in the dataset: distinct in carbon, nitrogen, oxygen, and strontium all at once, an eighteen-year-old who almost certainly wasn’t born anywhere near the Baltic.

There’s a curious postscript. A woman buried in grave 312, sharing the same burial pit as grave 311, turned out to be a dietary outlier herself, not as extreme, but her carbon values also pointed toward some C4 plant consumption, unusual for someone otherwise reading as local. Shared burial pits in Christian cemeteries often indicate family relationships, one interment placed above another. The authors flag this as a possible kin connection, something only ancient DNA work could confirm, but if true it would mean an immigrant household, not just an immigrant individual, took root in Vilnius and stayed.

None of this settles why people were moving. The paper lays out several plausible drivers without picking one: dynastic ties between Lithuanian dukes and Halych-Volhynia rulers that would have opened trade and diplomatic channels, Orthodox Ruthenians becoming vassals as the Grand Duchy expanded southeast and absorbed their home territories, merchants drawn by the same open invitations Grand Duke Gediminas sent across Europe in 1323, or displacement from the Black Death, which swept through the region between 1346 and 1353 and, notably, hit Eastern and Northern Europe less severely than the rest of the continent. Vilnius may simply have been a place people fled toward.

What the isotopes can’t do, and the authors are careful about this, is pin an origin more precisely than “somewhere south, plausibly Ukraine or southern Poland.” Strontium baselines overlap heavily across Eastern Europe, and with only 15 individuals sampled out of more than 500 graves, grave 311 reads less like an anomaly and more like a fluke of discovery. If one immigrant surfaced by chance in a sample this small, the real number moving through the Civitas Ruthenica over a century and a half was almost certainly larger.

Vilnius spent the thirteenth and fourteenth centuries as Europe’s last openly pagan capital. It was also, this data suggests, already running an Orthodox Christian neighborhood populated in part by people who’d walked there from hundreds of kilometers away, and by local women who married into that world and adjusted their faith without entirely letting go of their own. Two things can be true about a city at once. This one just took eight centuries and a mass spectrometer to prove it.

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 glyphs sit at the bottom of an inverted L, tucked into the corner of a room that nobody was ever meant to admire. Nine hieroglyphs of mathematics, then two more that don’t belong to any calendar system at all. So says, the text reads, followed by a name: Sak Tahn Waax. White-chested Fox.

That’s it. That’s the whole announcement. No royal titles, no cosmic parentage, no k’atun-ending fanfare. Just a scribe, somewhere around 781 CE, writing down a formula and then putting his name on it, the way anyone might sign a worksheet.

Franco D. Rossi, David Stuart, and Heather Hurst, the team that deciphered the text, are calling it the first known instance of a Classic Maya mathematician-astronomer claiming credit for intellectual work. Not an artist. Not a sculptor. Someone who did the math.

That distinction matters more than it might sound like it should.

A room full of scratch paper

Structure 10K-2 at Xultun, Guatemala, is a small masonry building that archaeologists opened in 2010 after it had spent roughly a millennium buried under fill. Its interior walls were painted with portraits of seated men bearing the title taaj, “obsidian,” a designation for ranked specialists. One wall shows the Xultun ruler Yax We’n Chan K’inich dressed as a maize deity with a scorpion’s tail. It’s a formal room, in other words, or it started as one.

What happened next is the interesting part. Someone kept coming back and writing on it. Painting numbers over the portraits. Reusing patched plaster surfaces for fresh calculations. By the time the room went out of use, it held more than 50 discrete texts crowded into the northeast corner and along the east wall: a lunar table, columns of “super numbers” divisible by both the Mars and Venus cycles, a ring number, and dense little clusters of calendrical arithmetic that have almost nothing to do with the room’s original decorative program.

Rossi, Stuart, and Hurst read this as a workspace, not a monument. The kind of place where someone was training in calendrical calculation, possibly while making bark-paper codex books, of which only four pre-Columbian examples survive anywhere, all centuries younger than this room and none of them archaeologically excavated. If that reading holds, 10K-2 is the closest thing we have to a Classic-period scriptorium caught mid-use, walls covered in the equivalent of scratch paper.

Scratch paper, notably, tends not to get signed. Most of it doesn’t say who did the arithmetic. Text 19 does.

Reading a formula that was never meant to be permanent

Reconstructing it took more than a decade. The east wall’s plaster had suffered from root growth and water infiltration, and much of Text 19 survives only as faint pigment traces. The team layered flatbed scans, infrared photography, and dStretch-processed imaging to pull 11 glyph blocks out of a surface that, to the naked eye, mostly looks like eroded stone.

What emerged is a sequence of five dates connected by four intervals, expressed in the conventional Maya vocabulary of Distance Numbers, followed by that two-glyph attribution. The whole formula spans exactly 2,920 days, which is not an arbitrary span. It’s the lowest common multiple of the 584-day Venus synodic cycle and the 365-day solar year, meaning Venus’s position relative to the sun repeats almost exactly every 2,920 days, or eight solar years, or five Venus cycles. This interval shows up constantly in Maya inscriptions, most famously in the Dresden Codex Venus tables.

What’s unusual about Text 19 isn’t the total. It’s what the scribe did inside it. Rather than dividing the 2,920 days into the five Venus periods that generated the number, he parceled it into a completely different sequence of nested intervals: 20 days, then 260, then 1,560 (two Mars cycles), then 1,080 (three 360-day Tun cycles), each interval larger than the last. Read in order, those intervals happen to follow a Fibonacci-like progression, though whether the Maya mathematician who built this recognized that pattern as such, or simply arrived at it through some other logic entirely, is not something the epigraphic record can tell us. The team is careful not to claim more than the text supports.

Nine glyphs to state the math. Then two more.

The name at the end

Cheheen, the first of those closing glyphs reads, a quotative particle that shows up elsewhere in Maya script meaning roughly “so says” or “thus spoken.” It’s the kind of marker used to attribute a statement to someone. What follows is unambiguous: SAK-TAHN-wa-xi, Sak Tahn Waax, White-chested Fox.

The authors are appropriately cautious about what this attribution actually means. It’s possible White-chested Fox calculated and painted the formula himself. It’s also possible the phrase cites him as an authority, the way a modern text might say “according to so-and-so,” crediting someone else’s method rather than one’s own hand. Either reading still makes this the only known case of a Classic Maya mathematical text naming an individual responsible for the intellectual content, as opposed to naming a king whose reign the calculations happen to serve.

That’s a strange gap in the record, when you sit with it. Maya scribes and sculptors signed things. Painted ceramics carry artists’ signatures. Carved monuments sometimes name the sculptor. Rossi, Stuart, and Hurst note that individual scribes and artists were increasingly visible in the eighth century, claiming credit for craft and style in a way that represented a real cultural shift. But the people doing the underlying astronomical and calendrical calculations, the ones whose work actually determined when kings were inaugurated or monuments erected, stayed invisible. Their results got carved in stone. Their names didn’t.

Part of the explanation may simply be genre. Formal inscriptions were built to display outcomes: coordinated dates, celestial alignments, the appearance of destiny aligning with political timing. They weren’t built to show the mechanics behind those outcomes, and mechanics are exactly what tend to carry an individual’s fingerprints. Text 19, sitting in a back room where nobody but other specialists would ever read it, had no audience to perform for. Maybe that’s precisely why a name could attach to it. There was no larger narrative for the name to compete with.

Why this text and not one of the other 50-plus in the room remains an open question the authors don’t try to resolve. Several other inscriptions in 10K-2 look, by their handwriting, like they could be the work of the same scribe. If so, White-chested Fox may have produced much of what’s on that wall, and chose to sign only the one formula that reads as a complete, self-contained composition rather than a fragment of some larger calculational tool. A finished piece, maybe, rather than a working note. It’s a guess grounded in the text’s structure, not a certainty.

Contemporaries in India, Mesopotamia, China, and Greece who worked out solar and planetary cycles have long had names attached to their achievements. Astronomy in the Classic Maya world produced calculations of comparable sophistication and, until now, no comparable names. Somewhere in northeastern Guatemala, in the final years before Xultun’s central precinct empt…

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 tibia fragment from a Belgian cave had been sitting in a museum collection, confidently refitted to a particular leg bone, for years. Then researchers sequenced its DNA. It didn’t match. The fragment belonged to a different individual entirely — and only by redoing the refitting and pulling a new sample from the correct bone did the picture come clear. This is the kind of thing ancient DNA analysis does now: not just fill in population histories, but correct the physical reconstruction of bodies that died forty thousand years ago.

That correction is one small moment in a much larger study, published today in Nature, that represents the most genetically detailed look yet at the final Homo neanderthalensis populations of northwestern Europe. Researchers from the Max Planck Institute for Evolutionary Anthropology and collaborating institutions recovered genetic data from 27 individuals who lived less than approximately 52,500 years ago, drawn from ten archaeological sites in Belgium and France. The findings speak directly to some of the most contested questions in Neanderthal research: how connected were these populations, how much did they interbreed with close relatives, and did genetic deterioration contribute to their extinction?

The short answers, as far as this dataset allows: more connected than previously suspected, less inbred than their eastern counterparts, and probably not dying from a genomic death spiral.

A region, not a scattered handful

Most of the sampled sites cluster in the Meuse Basin of Belgium — Goyet, Spy, Couvin, Trou Magrite, Fonds-de-Forêt, Engis, Walou — with two French sites, Saint-Césaire and Arcy-sur-Cure. The density of Middle Palaeolithic sites in this region is unusual, and that density is precisely what makes it possible to study something approaching a regional population rather than isolated individuals from distant locations.

The team generated a high-coverage genome (22.4-fold) from a Goyet individual designated GN1 — only the fifth high-coverage H. neanderthalensis genome sequenced to date. For specimens with poorer DNA preservation, they used a new hybridization capture panel called ArchaicPlus, designed to target around 2.3 million SNPs informative for Neanderthal and Denisovan variation. Combined, this approach yielded nuclear data from 19 skeletal remains across six sites.

The genetic picture that emerged is of a loosely knit regional population. Most of the Belgian and French individuals are more closely related to each other than to other contemporaneous late Neanderthals, including Vindija 33.19 from Croatia and the Mezmaiskaya individuals from the Caucasus. Their estimated population split from the Vindija lineage falls around 54,000 years ago — roughly 10,000 years before these individuals actually lived. A relatively shallow separation. Distinct, but not isolated.

What is perhaps more telling is what the genomes do not show. The H. neanderthalensis individuals from Chagyrskaya Cave in Siberia, studied a few years ago, displayed unmistakable signatures of close-relative mating: long homozygous-by-descent tracts in the genome, indicative of repeated inbreeding within a small and probably bounded group. The GN1 individual from Goyet shows nothing of the sort. Her proportion of genome contained in homozygous tracts is comparable to Vindija 33.19, not to the Chagyrskaya or Denisova 5 individuals. No excess of long tracts. No evidence of recent inbreeding. Whatever social dynamics shaped the Altai groups, they were not universal to the species.

The Goyet assemblage is worth pausing on. The remains recovered there — eventually assigned to at least six individuals — are highly fragmented, and many bear cut marks and percussion damage. An analysis published in 2016 interpreted these as evidence of cannibalism: defleshing, marrow extraction, bone use as tools. The individuals were processed like prey.

The new genetic data adds a strange dimension to this. Among the nine Goyet specimens sequenced, the researchers identified two clusters of genetically identical fragments, three bones each, representing two individuals (GN1 and GN2). One cluster had been wrongly refitted across two different tibias in the original skeletal analysis; only DNA comparison revealed the error. Beyond those within-individual matches, however, none of the Goyet Neanderthals appear to be close relatives of each other. No parent-offspring pairs. No siblings. A neonate is present among them, but his mother is not represented among the sampled females.

Whether the Goyet assemblage represents a single community or accumulated over time is genuinely unclear, and the researchers are appropriately cautious about inferring group structure from it. But the absence of close kin is notable. Isotopic evidence adds another layer: sulfur isotope values suggest local foraging at Spy but non-local origins at Goyet. Whoever those individuals were to each other, they were not a family, and at least some of them came from somewhere else.

Contrast that with Spy Cave, roughly thirty kilometers away, where the anatomical articulation of some skeletal elements has been interpreted as evidence of deliberate burial — though that reading remains contested. Genetically, the Spy individual, whose upper molar and femur turned out to belong to the same adult male (confirmed through nuclear DNA comparison), belongs to the same broad regional population as the Goyet individuals. Two sites, similar genetics, radically different treatment of the dead.

A ghost in the mitochondrial tree

The simplest story about late H. neanderthalensis in western Europe is that they formed a single, broadly connected population in their final millennia. The new data largely support that story, but with notable exceptions.

A molar recovered from Couvin, Belgium carries a mitochondrial genome that falls not within the main late Neanderthal clade but in a deeply divergent lineage shared with a Neanderthal called “Thorin” from Grotte Mandrin in the Rhône Valley of France. The divergence between this lineage and the mainstream late Neanderthal mtDNA is estimated tens of thousands of years deeper than the splits within the main clade. This is a lineage that predates the main radiation of late Neanderthals and apparently persisted alongside them in western Europe — possibly including other individuals known from Gibraltar and Poland.

The radiocarbon dating for Couvin places the individual at roughly 49,000 to 44,000 years ago, and the molecular dating shifts accordingly when anchored to those dates. But when the molecular clock runs unconstrained, it pushes the estimate past 100,000 years — a discrepancy similar to what has been reported for Thorin. The archaeological context argues against the Couvin tooth being an intrusion from an older deposit, and the researchers consider this unlikely. What seems more plausible is that a divergent maternal lineage, ancient in origin, was more widespread in western Europe than previously recognized and coexisted with other late Neanderthal lineages until close to the end.

The Arcy-sur-Cure individual (AR-30) from northeastern France adds a separate complication. Her nuclear genome suggests a population split from Vindija 33.19 around 131,000 years ago — much earlier than the other Belgian and French individuals, whose splits cluster around 54,000 years. The team explored possible explanations: faunal contamination (ruled out by metagenomic screening), modern human contamination (too low to account for the signal), or gene flow from a deeper Neanderthal lineage into AR-30’s ancestry. No firm conclusion is possible from the current data. More sequence coverage would help.

No accumulation of genetic damage

One influential hypothesis for H. neanderthalensis extinction holds that small, isolated populations were gradually worn down by the accumulation of mildly deleterious mutations — genetic load accumulation, a process documented in …

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

In a 2025 study published in PNAS, researchers led by Christine Ou and Steven Heine asked roughly 250 people in each of 20 countries, spanning six continents, how long they’d slept the night before. France came out on top at 7 hours and 52 minutes. Japan came in last at 6 hours and 18 minutes. The gap between them, an hour and thirty-four minutes, is roughly the difference between a full sleep cycle.

If sleep science worked the way most public health messaging implies it does, that gap should show up in the data as a gap in wellbeing. Japan should look, on paper, like a nation quietly grinding itself down. It doesn’t. Diabetes rates, heart disease, obesity, life expectancy: none of it tracks with how long a country sleeps. In a separate analysis of 353 national sleep averages pulled from 14 different datasets covering 71 countries, the team found no relationship between a country’s average sleep duration and its rates of heart disease, diabetes, or life expectancy. Stranger still, countries where people slept longer had higher obesity rates, the opposite of the pattern researchers have repeatedly found when they study individuals within a single country.

This is the kind of result that should make you sit up. Not because it’s surprising in isolation. Findings about sleep and individuals are everywhere, and so are claims that whichever number a press release happens to favor is the one that will fix your life. What’s surprising is what happens when you zoom out from individuals to populations. The relationship between sleep and health doesn’t vanish at the national level. It gets reorganized.

What the Curve Actually Looks Like

Inside any given country, the study replicates the familiar shape: a curve, not a line. Sleep too little, your health composite score (built from depression, chronic conditions, subjective health, and overall wellbeing) drops. Sleep too much, it drops again. There’s a sweet spot in the middle, and it’s a real, statistically robust feature of the data within every single country in the sample.

But the position of that sweet spot moves. The researchers found a turning point, the amount of sleep associated with peak health, for each of the 20 countries individually, and those turning points differed significantly from one another. The optimal number isn’t one number. It’s twenty numbers, one per culture, and they don’t converge.

What did converge, across every country studied, was something else entirely: the gap between how much people actually slept and where their personal curve peaked. In every country, average sleep duration fell short of the turning point. Everyone, everywhere, is sleeping somewhat less than their own culture’s apparent optimum. The shortfall is universal. The optimum is not.

There’s a second result buried in here that’s arguably more interesting than the headline finding, and it has nothing to do with hours. The researchers asked each participant not just how long they’d slept, but what they believed their cultureconsidered an ideal amount of sleep. Then they measured the gap between a person’s actual sleep and their own estimate of that cultural ideal. People whose sleep was closer to what they believed their culture expected, regardless of whether that was six hours or eight, reported better health. The effect held up even after controlling for the raw number of hours slept.

In other words: it’s not just that six hours works for some places and eight works for others. It’s that matching the local norm, whatever that norm happens to be, carries its own independent health signal. The researchers floated a few explanations. Maybe people feel subjectively healthier when their habits feel normal. Maybe there’s friction, low-grade and cumulative, in being out of step with everyone around you (worrying about missing the early train, structuring your evening around a schedule nobody else keeps). Maybe it’s something more biological: people whose sleep architecture doesn’t fit their environment may be, for reasons unrelated to the hours themselves, less healthy to begin with, and the correlation runs the other direction. The data can’t distinguish between these, and the authors are upfront that it’s correlational throughout. But the fact that “fit” predicts health independently of “amount” is the kind of finding that quietly reframes the whole question.

The View From Three Million Years Back

None of this happens in a vacuum, obviously, and one of the more useful frames for thinking about why a six-hour average and an eight-hour average can both sit at the top of their respective curves comes from outside the PNAS paper entirely.

David Samson, an evolutionary anthropologist at the University of Toronto, has spent years studying sleep across the primate order, lemurs, orangutans, chimpanzees, and eventually humans themselves, including extended fieldwork living alongside the Hadza in Tanzania and the BaYaka in the Republic of the Congo. His phylogenetic models, built from sleep data across more than 30 primate species and controlling for body size, brain size, social structure, and terrestriality, generate a prediction for how much Homo sapiens “should” sleep given our biology: about 11.5 hours per 24-hour period.

We sleep, on average, about seven.

That’s not a small discrepancy. Owl monkeys sleep up to 17 hours. Tarsiers manage 15. Lemurs sit around 13 to 14. Great apes, our closest relatives, average somewhere between 9.5 and 10. Humans are the outlier of the entire order, the primates who sleep the least, by a wide margin, relative to what their biology would predict.

Samson’s argument, laid out in his book The Sleepless Ape, is that this isn’t a deficit. It’s the result of a real evolutionary shift, one he dates to roughly 1.8 million years ago, when Homo erectus began building shelters. Once you have a controlled sleeping environment, natural selection has room to start trimming non-REM sleep, gradually, over hundreds of thousands of years. At the same time, the advent of cooking with fire collapsed the daily chewing budget. Chimpanzees spend five to six hours a day chewing raw food. Gorillas spend up to eleven. Cooked food cut that down to roughly an hour for humans. Between the sleep reduction and the chewing reduction, Samson estimates our ancestors freed up something on the order of four extra hours a day, time that could go toward toolmaking, social bonding, teaching, and the kind of cumulative culture that no other primate manages at scale.

The “human sleep paradox” is that we’re the short-sleeping primate who also lives the longest and thinks the hardest. Samson’s framing is that the short sleep isn’t despite our cognitive advantages. It’s bound up with how we got them.

Where the Stories Meet, and Where They Don’t

Here’s where it gets genuinely interesting, because the two pictures, the PNAS cross-cultural data and Samson’s evolutionary one, line up in one place and pull apart in another.

They agree on the headline: humans sleep less than the textbooks have generally assumed they should, and it’s not a crisis. The widespread narrative of a modern “sleep deprivation epidemic,” driven by phones and stress and artificial light, gets undercut from two completely independent directions. The PNAS data shows no health penalty for nations that sleep less. And Samson’s fieldwork found something that should be more alarming to that narrative than it usually gets credit for: small-scale societies like the Hadza and the BaYaka sleep less than industrialized populations, not more, averaging around 6.4 hours, with sleep efficiency around 70%, well below the 85% the National Sleep Foundation considers high quality. If anyone should be sleeping “naturally,” free of screens and shift work and 11pm emails, it’s hunter-gatherers. And they’re sleeping worse, by the conventional metrics, than people in Tokyo or Toronto.

So why aren…

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