Most people who know the aurochs know how it ended: the last cow died in a Polish forest in the 17th century, and people did it. The authors accept that for Poland. They do not accept that it explains Scandinavia, where the species was gone thousands of years before, and where the usual suspect, the Mesolithic hunter with a spear, turns out to have been busy at the wrong time and in the wrong places to carry the whole blame. A second belief sits underneath the first, namely that the animals were cut off from the continent by about 10,000 years ago and began their slow decline in isolation. The genomes do not quite say that either.

Start with the isolation, since the genetics speak most plainly there. All 11 sampled Scandinavian aurochs belong to the P1 mitochondrial haplogroup, and their common maternal ancestor lies somewhere between 18,200 and 12,500 years ago (cal BP, 95% highest posterior density). They cluster with western European sequences, which fits a colonization out of a Franco-Iberian refugium after the ice, across a land bridge that low seas and rebounding land had opened in the western Baltic. For a long time the working assumption was that a river and lake system draining across present-day Denmark cut the Scandinavian population off as early as 10,000 cal BP. The nuclear genomes agree that the animals had drifted apart from the mainland; in a multidimensional scaling of allele sharing, the Scanian, Zealand and south-central Swedish samples sit in a tight group far from the western European one, with the Zealand and Swedish animals the most drifted. But the mitochondrial tree, built from 34 mitogenomes of which 11 are Scandinavian, shows lineages that still join up with continental ones as recently as 10,300 to 8,500 cal BP, and the pattern suggests some contact with the mainland lasted until the Danish-Swedish straits finally flooded around 8500 to 8000 cal BP. The Jutland animal, Hjo1, diverges from the Scanian and Zealand sequences at 9,900 to 8,500 cal BP.

The cost of arriving was already written in the genomes by then. The long runs of homozygosity that mark recent matings between close relatives are almost absent in Ska1 and Ska3, and what fills their profiles instead is a heavy load of short and medium runs, which the authors read as the signature of a population that had been small for a long time, a bottleneck during colonization, rather than of parents who were cousins. Hjo1, the more recent animal from North Jutland, has the larger fraction of long runs, so inbreeding in the strict sense does show up, but later and farther north. The authors keep saying how thin this evidence is: the demographic inference rests on two individuals, and each genome, they note, carries the genealogies of hundreds of ancestors, which helps without settling anything. They also decline to say that inbreeding killed the species. The woolly mammoth went through a severe bottleneck and still lasted roughly 6,000 years on Wrangel Island (Dehasque and colleagues, 2024), so a poor genome can be a very long sentence.

Read more

The usual picture of an Inka coca ritual comes from Spanish chroniclers, who described leaves being burned or bundled as offerings to spirits and deities. Most readers who know anything about the Inka carry that picture without noticing its source. The Inka left no writing, so every account of what they did with coca in a ceremony was set down by someone else, in Spanish, after the fact, and the leaves in the llama wool are the first direct evidence of coca in ritual at Tambo Viejo. Lidio Valdez of the University of Calgary and his co-authors Katrina Bettcher (Trent University) and M. Soledad Gutiérrez report the find in Ñawpa Pacha,1 where it appeared online on September 7, 2026. Their abstract is blunt about the mismatch: the written sources say the coca was burned, and the archaeological evidence consists of the actual leaves.

Read more

Most readers who follow the field carry a tidy story about it. Clovis First, the idea that hunters of the Clovis culture, around 13,500 years ago, were the original colonists, collapsed when older sites turned up, and the story since has seemed to be one of dates sliding steadily backward: White Sands at about 23,000 years, Chiquihuite Cave at about 32,000, the Cerutti Mastodon site in southern California at 130,000. Genetics, in this version, is the slow discipline waiting to be persuaded. The review describes the relationship differently. The genomes that helped finish off Clovis First also put a limit on how far back the ancestors of living Native Americans can be pushed, because the diversification of their lineages, the pattern expected when a founding group first spreads into a new region, appears in the genomic record no earlier than about 21,000 years ago.

The authors make a point about the Americas that sets them apart from the rest of the Homo sapiens story, in which our species repeatedly moved into continents already held by Neanderthals, Denisovans and other so-called archaic hominins. Here the genomic data imply that a small group of people entered two continents with no other humans on them. That predicts a sparse early archaeological record, with few sites, made sparser by the extra thousands of years in which weather, rivers and soil could destroy what was left. The authors argue that the post-glacial record of both continents looks exactly like this.

The Cerutti claim runs into the sharpest version of the problem. If DNA were ever recovered from a site that old, the authors write, it should belong to Neanderthals, Denisovans or some unknown lineage, since Homo sapiens did not begin its persistent expansion beyond Africa until 60,000 to 70,000 years ago. The oldest Native American genomes, from Upward Sun River in Alaska and from Anzick, carry patterns of archaic admixture that do not match that scenario. Even a much later arrival shortly before the Last Glacial Maximum (LGM, about 25,000 to 20,000 years ago) fits the genetic evidence poorly, though the authors treat arrivals at 40,000 years or later as more feasible than the 130,000-year claim.

The Siberian background is better documented than most of what follows it. Humans were living in the Arctic of East Siberia by 32,000 years ago, at the Yana Rhinoceros Horn sites, and genomes from two people there define a group called the Ancient North Siberians. The Yana people did not contribute directly to the ancestry of the First Peoples, the review’s term for the direct ancestors of Native Americans, but a group largely descended from related ancestry did, and it is represented by a 24,000-year-old individual from Mal’ta in south-central Siberia. That Ancestral North Eurasian ancestry makes up as much as 40 percent of the First Peoples lineage. Between 25,000 and 20,000 years ago, a group carrying it met ancient East Asians, and the resulting metapopulation then sat in isolation through the cold and dry peak of the LGM, a span called the Beringian Standstill. Where it sat is unresolved; researchers have proposed a refugium east of Lake Baikal, the East Siberian Arctic Shelf, and central Beringia.

Read more

That figure sits almost exactly on top of what clinicians report today. Studies of infants seven to twelve weeks old put modern prevalence between 37.8%and 46.6%. The comparison is imperfect, since the Matelica sample is adults and the clinical figures are babies, and I will come back to why that matters, but the coincidence is hard to ignore.

The belief it disturbs is a tidy one. In 1992 the American Academy of Pediatrics began advising parents to lay infants on their backs to reduce sudden infant death syndrome, and in the years that followed, pediatric clinics saw more flattened heads. The sequence is easy to read as cause and effect, and a good deal of popular and some clinical writing treats flat head syndrome as the price of a lifesaving policy, or more broadly as a problem of strollers, car seats and bouncers. The paper’s own introduction is more careful than that: the post-campaign rise coincided with growing clinical attention, and the pre-campaign estimates, drawn mostly from British clinical observation with varying case definitions, ran from 0.3% to 48%. Nobody had a clean baseline.

Read more

The usual telling of this period treats Homo erectus as the species that arrived with the bigger brain and the smaller molars and simply took over, an ancestor with an obvious future. Drimolen gives a different count. Martin says that 2 million years ago Paranthropus robustus outnumbered Homo erectus at the site by more than 10 to 1, and that our ancestors “looked like long shots to survive.” That ten-to-one figure comes from the university’s summary of the study. The paper itself reports fourteen P. robustus to one H. erectus within its batch of eighteen fossils, and calls that ratio broadly compatible with Swartkrans (Brain, 1993) and with earlier dental samples from Drimolen.

The paper, in Annals of Human Biology1 (Martin, A. B. Leece, Stephanie Baker, Andy I. R. Herries and David S. Strait, 2026), describes eighteen cranial and mandibular specimens excavated between 1992 and 2019, most of them dug by Andre Keyser’s team in the 1990s. Fourteen went to Paranthropus robustus, one to Homo erectus sensu lato, and three could not be placed beyond Hominini. The Paranthropus material also supports the authors’ argument that the Drimolen population, named the subspecies P. robustus ukusa, differs from the later Swartkrans and Kromdraai B samples through gradual change over time, with less bite force at Drimolen. They reject the older alternative, that Drimolen simply held mostly females and Swartkrans mostly males.

Read more

Most people who have met Meretneith at all have met her as a woman standing next to a man. Flinders Petrie, digging tomb Y at Abydos, first took her for a king, “Merneit,” and noted that the name on her stela lacked the customary serekh; once other inscriptions turned up with a feminine ending on the name, she became a queen consort, and later a regent holding the throne for her young son Den. In every version she keeps the seat warm for someone else. A serekh is the frame, modelled on a palace façade and topped by a god, inside which a ruler’s name was written, and it belongs to the person who actually rules. If the second and third impressions at Saqqara name Meretneith, then a woman wore it at the very start of the dynasty, and the regent is the wrong category for at least part of her career.

The trouble is that nobody can pick the stopper up. Sue Kelly (Czech Institute of Egyptology, Charles University) and Crystal Miller (University of Bucharest) write in Prague Egyptological Studies1 that neither the sealing nor Emery’s glass plate has been located, so the only surviving evidence is a black-and-white photograph in the excavation archive plus his published line drawing. That was enough for some scholars to accept Emery’s reading and for Peter Kaplony, in 1963, to call the publication unconvincing. Rudolf Anthes doubted that the two repeated signs in the name compartment were the hoe sign at all, and Silke Roth rejected the identification in 1997 on the grounds that Meretneith’s name is spelled differently elsewhere.

Read more

Two assumptions sit underneath how most readers picture religious education in the ancient world, and a study of first-millennium BCE Babylonian schools by Shana Zaia of Vrije Universiteit Amsterdam, published this year in Iraq,1 cuts against both. The first is that teaching children about the gods requires a religion class, with a scripture at the center and clergy running the school. Babylonia had no separate unit on religion or theology, no text functioning as holy scripture, and no seminary equivalent, and even the schools located in temple complexes were not primarily religious institutions. The second assumption is that ancient religion, being ancient, was handed down intact. Zaia agrees that Mesopotamian religious conventions were by and large conservative, but her cases show trained scribes and priests breaking from them whenever the break could be presented as a continuation of something older.

The schooling was private and household-based, with adult professionals as instructors, often the father of one or more of the boys in the class, and pupils drawn from elite or upwardly mobile families (Zaia notes that evidence for women’s scribal education is limited). Elementary students began with basic signs, lexical lists, proverbs and a few literary works; the second stage added harder lists, bilingual incantations, prayers, hymns and, in the north, a text called Tintir that catalogs the names, shrines, gates and streets of Babylon. Those headed for exorcism, divination, cultic singing or healing then entered a specialist phase resembling an apprenticeship.

Read more

There is a common intuition about newcomers and old hands, and most of us would apply it here without thinking. People arriving in a rich place spend freely, because they have not yet learned what things cost, and their descendants, taught by generations of experience, learn to waste nothing. Under that sits a second intuition, that people are thrifty with a material because the material is scarce. A study of 489 fluted spear points from the valley, published online on August 27 in the Journal of Archaeological Method and Theory,1 runs against both. Its authors, Jesse Tune, an associate professor of anthropology at the University of Mississippi who directs its Center for Archaeological Research, and D. Shane Miller, an assistant professor of anthropology at the University of Alabama, found that Clovis hunters resharpened their points close to the practical limit before letting them go, while the Cumberland people who followed them threw away broken points with plenty of working edge still on them.

Clovis points are broad, with a concave base and a long channel flake, the flute, removed from the base partway up each face. They are found across much of North America, and the best current dating places their manufacture between about 13,050 and 12,750 years ago. Cumberland points came after them and stayed closer to home in the Midsouth. They are narrower and thicker, and their flutes run much further toward the tip. Their age is less secure: Tune’s earlier work puts them somewhere between roughly 12,800 and 12,000 years ago, so on paper the two ranges overlap slightly at the edge. No site older than Clovis is known anywhere in the valley, and the authors treat Clovis groups as the first people to walk it and Cumberland groups as their successors, several centuries on, who inherited a country already mapped in memory.

What the pair set out to test is whether knowledge leaves a mark in stone. The paper leans on a 2003 framework by the archaeologist M. Rockman, which splits what colonists must learn into three kinds. Locational knowledge, that an outcrop exists and where, can be had in a single encounter. Limitational knowledge, how the stone’s quality varies across the outcrop, how far the source extends, whether it can be reached in every season and how it compares with others in the region, builds up through repeated use. Social knowledge, the version passed between generations and folded into how a group moves and makes its tools, takes longest of all. The valley’s abundance is what makes it a fair test, because where stone lies in nearly every creek, a difference in how carefully people used it is hard to put down to the stone supply, and knowledge is what remains.

Read more

That insistence had respectable sources. Joyce Marcus, in her 1992 survey of Mesoamerican writing systems, and Esther Pasztory, in her 1997 book on the city, both characterised Teotihuacan as a state without a true script, and the vocabulary that grew up around its signs did the same work more quietly. James Langley, who catalogued the signs in the 1980s, spoke of a “notational system” and of “emblematic glyphs,” terms that placed Teotihuacan a rung below the Classic Maya and the Aztecs, whose hieroglyphs everyone agreed recorded language. The German epigrapher Thomas Barthel saw the danger in 1987, warning that a purely visual reading tempts scholars to treat the signs like airport pictograms. The suspicion was not unreasonable, because Teotihuacan texts are short, usually a single glyphic compound, and the city left no long carved narratives of kings and wars of the kind that made Maya decipherment possible.

The new paper on Cave 19, published in Antiquity1 by Christophe Helmke and Jesper Nielsen of the Institute of Cross-cultural and Regional Studies at the University of Copenhagen, with Carolina Meza Rodríguez and Mario Córdova Tello of Mexico’s Instituto Nacional de Antropología e Historia (INAH), takes the opposite position as settled. Drawing on a run of papers by the authors and by Karl Taube, they describe a logo-phonetic script comparable to Maya and Aztec writing, with a repertoire they now estimate at almost 120 signs, in which researchers have identified calendar notations, place names, personal names, titles and even verbs.

Which is why a painting not much bigger than a dinner plate, on a cliff far from the city, matters more than its size suggests. A script shows itself in the way its signs behave: the same compound should mean the same thing wherever it turns up, whether on an apartment-compound wall in the city or in a shelter on a hill in Morelos. Terse texts give the skeptics little to test, so every example found outside the metropolis is a chance to check whether the signs travel with their meanings intact. Cave 19 offers something close to that test, because one of its glyphs appears to name a place that is also named in a mural inside Teotihuacan.

Read more

Göbeklitepe (the older two-word spelling, Göbekli Tepe, is still the more common one in English) lies 15 kilometers northeast of the city of Şanlıurfa in southeastern Türkiye, and for the first two decades after excavation began in 1995 it was described as a sanctuary: a place where hunter-gatherers came together to raise T-shaped pillars carved with wild animals, and where hardly anyone lived. Klaus Schmidt of the German Archaeological Institute (DAI), who led the excavation from its first season, called his 2006 book on the site Sie bauten die ersten Tempel, “They Built the First Temples.” In that picture, the human bone that kept turning up in the fill of the great pillared rooms was read, reasonably enough, as part of the ritual.

The count now stands at 928 fragments, not including the two graves, recovered from inside those rooms (the current team calls them special buildings) and from the ground between them. They come from men and women of every age, and some bear marks of deliberate handling. In June 2017, Julia Gresky, an osteologist at the DAI’s Division of Natural Sciences in Berlin and Georg-August University Göttingen who has run the analysis of the site’s human remains since 2009, published three skulls with deep grooves cut along the midline in Science Advances, reading them as evidence of a previously unknown form of Neolithic skull cult; human burials, that paper noted, were “still absent from the site.” Now she is lead author on a paper, published September 22 in PLOS One,1 arguing that much of the scattered bone reached the special buildings by accident. Gresky and her co-author, Lee Clare of the DAI’s Istanbul Department, propose that people living on the higher parts of the mound buried their dead beneath their own floors, as households did across Neolithic Anatolia and the Levant, and that the slopes later gave way and carried floors, rubble and bones down into the hollow where the special buildings stand.

Human bone was one of the few places where Göbeklitepe’s imagery, which includes a headless human figure on Pillar 43 in Special Building D, touched actual bodies. If most of the scatter is landslide debris, the physical case for ritual treatment of the dead shrinks toward a handful of carved skulls, and the settlement around the monuments starts to look like what recent seasons have suggested on other grounds, a place where people lived, died and were put under the floor. Telling a ritual deposit from a slide depends on taphonomy, the study of what happens to a body after death, and taphonomy needs a baseline. Until 2017 there was not a single excavated grave at the site to provide one.

Read more

The assumption most readers bring to Bronze Age Greece is that an age of palaces, of exchange networks reaching across the Aegean and beyond, of newcomers carrying steppe-related ancestry into the region from around 2200 BCE, must also have been an age when large numbers of people left home and settled among strangers. Pottery and metal certainly moved along those networks, but whether many people moved with them and stayed is a separate question, one that a study published September 23 in PLOS One1 tried to settle by counting. Among 199 individuals from six sites in central and southern Greece, tested through tooth enamel or skull bone, the researchers found 17 whose chemistry says they grew up somewhere else.

The six sites span about 2,300 years. The oldest is Kalyvia in Elis, a cemetery of rock-cut chamber tombs dated to roughly 3400 to 2800 BCE, straddling the end of the Neolithic and the start of the Early Bronze Age. Next are the rock shelters at Kavos on the Cycladic island of Keros, about 2500 to 2200 BCE, where most bones were systematically carried off after the bodies decomposed, leaving small bones and teeth behind, and the teeth are what this kind of analysis needs. The three graves on Mount Kynortion, filed in the study under the site name Apollo Maleatas, follow in the 19th century BCE. Kirrha, a low mound near the Corinthian Gulf by modern Itea, and the North Cemetery at Ayios Vasileios, 12 kilometers south of Sparta beside a recently identified Mycenaean palace, both cover the passage from the Middle to the Late Bronze Age around 1700 BCE. The youngest is Voudeni, a Mycenaean cemetery of more than 80 chamber tombs northeast of Patras, used from about 1450 to 1050 BCE, and at 102 individuals it supplies half the dataset.

The count matters because archaeology, left to itself, cannot tell a traveler from a settler: an imported pot in a Mycenaean tomb could have come with a trader who sailed home the next week or with someone who married into the village and was buried beside it forty years later. The ratio of strontium-87 to strontium-86 in soil and water depends on the age and type of the underlying rock, passes into crops and then into the people who eat them without meaningful change, and is locked into tooth enamel as the enamel forms. A first molar begins at about birth and is finished by about age three; the petrous bone, the dense wedge of the skull around the inner ear, stops remodeling after about age two. The team led by Anja Frank, who did the work at the National Museum of Denmark and has since moved to Germany’s Federal Institute for Geosciences and Natural Resources in Hannover, sampled one or the other from each person (107 enamel samples, 92 petrous bones) and compared each result with two expected ranges: a surrounding-area range, built from published measurements within 25 kilometers of the site, and a burial-population range, built from the site’s own dead once obvious outliers were set aside.

Seventeen of 199 is about 8.5 percent, roughly one person in twelve. Broken out by period and pooled with earlier strontium studies from other Greek cemeteries, the share barely moves: between 5.9 and 8.2 percent for the Neolithic and Early Bronze Age, 5.4 percent across the Middle-to-Late Bronze Age transition, 9.8 percent for the Late Bronze Age, the era of the Mycenaean palaces. Single sites swing wildly when the samples are tiny (two of the three people on Mount Kynortion were non-local, a rate of 66.7 percent that means very little), but the larger samples sit close together, with Voudeni at 7.8 percent, Kirrha at 9.4, and Ayios Vasileios at 5.9. The authors point out that strontium studies of Bronze Age Italy and Denmark have found larger shares of outsiders, while conceding that overlapping ranges within Greece may explain part of the gap.

Nor did the newcomers come from one direction. At Voudeni and Ayios Vasileios some fall above the burial-population range and some below it, so each site drew on at least two source regions, and every one of the 17 could have grown up elsewhere in Greece, though their values would also fit parts of Italy, Cyprus, Turkey, and Spain.

Read more

The colored version of the face shows a man with brown skin, a heavy brow, a full dark beard and hair to his shoulders, and most people who see it will take it for an estimate of how one individual looked some 315,000 years ago, although no single individual lies under it. The Leipzig skull is a composite, and the new paper lists its parts: the braincase and upper face of Irhoud 1, the fossil a mine worker turned up at the site in 1961; the lower jaw of Irhoud 11, a different adult; the upper tooth row, possibly, from a specimen catalogued as Irhoud 10; and two dental fragments, Irhoud 21 and Irhoud 22. By the paper’s account all of them come from one deposit, Layer 7, which fire-heated flint tools date to about 315,000 years ago, give or take 34,000.

Irhoud 1 was first identified as an African variant of the Neanderthals, about 40,000 years old, and its estimated age climbed for decades, to between 100,000 and 200,000 years in 1991 and about 160,000 in 2007, until two papers in Nature in June 2017 reclassified the Jebel Irhoud fossils as Homo sapiens and put them at roughly 315,000 years. One, led by Jean-Jacques Hublin, then director of the Leipzig institute’s Department of Human Evolution, described the fossils; the other, led by Daniel Richter, a geochronologist also at the institute, supplied the flint dates along with a recalculated age of about 286,000 years for a tooth from Irhoud 3, a juvenile’s jaw that is not part of the composite.

Facial approximation of this kind begins with somebody else’s head. The team took a CT scan of an anonymous modern adult, collected, the paper notes, with the donor’s written consent, and deformed that skull in the free 3D program Blender until its outer surface and its inner cavity matched the Jebel Irhoud composite, carrying the donor’s soft tissue along with the bone. Soft-tissue thickness values came from reference data on modern people of African ancestry, compiled in the forensic artist Karen T. Taylor’s 2000 textbook Forensic Art and Illustration and chosen, the authors write, for their geographic and evolutionary closeness to the site. Skin color, lighting and hair are, by the authors’ own description, artistic conventions without grounding in the fossil. A second set of renderings, in grayscale and without hair, strips those away and leaves only the relation of face to bone.

Hublin’s team had described the Jebel Irhoud fossils as a mosaic. The face, jaw and teeth fell in line with modern humans while the braincase and the shape of the brain cavity stayed more primitive, and those modern features were the core of the case for placing a 315,000-year-old population in our species, a placement the 2017 paper framed with some care, as documenting early stages of the Homo sapiens clade. A reconstruction lays flesh over exactly that part of the head, and the flesh has no direct evidence behind it, because no soft tissue of any kind is known from hominins of the Middle Pleistocene (roughly 774,000 to 129,000 years ago) and every millimeter of skin on this model is borrowed from people alive today. The result in the new paper that holds up best against outside measurement concerns a part of the head that no one looking at the portrait will ever see, the empty space where the brain was.

Read more

The explanation most of us carry for failing joints is mechanical, and it is old. Wilton Krogman, a physical anthropologist at the University of Pennsylvania, gave it its most durable name in a 1951 Scientific American essay, “The Scars of Human Evolution,” whose summary line noted that although man stands on two legs, “his skeleton was originally designed for four.” In the popular version, hips, knees and lower backs carry weight that four limbs once shared, they wear down under it, and osteoarthritis is the bill that comes due late in life. The cartilage from her hands and from the 48 other bodies fits that story poorly. Human cartilage held roughly a third as much of the water-binding sugar chains called glycosaminoglycans, or GAGs, as ape cartilage did, and the shortfall was plain in the elbow, a joint that carries less weight in a person than in a chimpanzee walking on its knuckles.

Earlier measurements cited in the paper put the GAG loss in osteoarthritic human cartilage at 25 to 38 percent below healthy human cartilage, and the gap the team found between human and ape cartilage is about twice that. The authors suggest the shift may have pushed human cartilage “towards a threshold of vulnerability,” such that “all humans carry a baseline GAG deficit that primes the joint for degeneration.”

A joint whose chemistry tracked only its workload would not look like that elbow, and the likelier culprit is heredity, in particular the stretches of DNA that set how strongly genes are switched on. Those switches are what the study’s leaders, Fumitaka Inoue, an associate professor at Kyoto University’s Institute for the Advanced Study of Human Biology (WPI-ASHBi), and David Gokhman, a principal investigator at the Weizmann Institute of Science in Rehovot, Israel, had set out to map across the human skeleton. Their strongest pathway signal pointed at the machinery for making GAGs. The paper cannot close the question an anthropologist asks first: why a lineage would turn down its supply of the molecules that keep joints hydrated and springy, and whether that was a price paid for something else or a loss that selection simply stopped guarding against.

Read more

At Oreškovica-Selište, a fortified Vinča settlement elsewhere in Serbia, the enclosing ditch measured seven meters across and 3.6 deep. Faced with one half again as wide and nearly a meter deeper, most of us would assume a chain of command behind it, someone with the standing to put a few hundred people to work, and we tend to assume more broadly that a village which grows past a couple of hundred people starts sorting itself into ranks. The archaeologist Gregory Johnson gave that intuition a mechanism in 1982, arguing that decision-making comes under what he called scalar stress as groups grow, and that hierarchy is one of the usual ways the strain gets relieved. A new study of Opovo, published this year in Archaeologia Austriaca,1 describes a village far larger than anyone had thought, enclosed by five ditches, whose houses show little sign that any household rose far above the rest.

The earlier picture came mostly out of one trench, sixteen meters by twenty, dug between 1983 and 1989 by teams from the universities of Novi Sad and Berkeley and the National Museum in Pančevo, on a site that a local priest and amateur archaeologist named Andreas Majer had found in 1965. Ruth Tringham, the University of California, Berkeley, archaeologist who led that project, put the settlement at about five hectares with no more than ten households standing at any one time, and in 1992 called it a “small hamlet.” The population estimate that went with that picture was about 200. It was an odd hamlet. Between 62 and 71 percent of the animal bones from the old excavation came from wild animals, red deer above all, a higher share than at other Vinča sites and closer to the pattern at sites of the Tisza culture, its Late Neolithic neighbor on the Great Hungarian Plain, and the hill stood in marsh ringed by old loops of the Timiș River, ground Tringham later called agriculturally marginal. One early idea had Opovo as a seasonal camp for hunting and gathering expeditions, until Nerissa Russell, working through the bones for her 1993 doctoral dissertation at Berkeley, ruled out seasonal use and made it a permanent village with a mixed economy that leaned hard on wild animals.

Tringham is the last of eleven authors on the new paper, whose lead author, Fynn Wilkes, is a postdoctoral researcher in the ROOTS Cluster of Excellence at Kiel University, and it takes her hamlet apart with her name on the byline. The new team could not even say exactly where her trench lay; of two blank patches near the middle of their magnetic map, which they read as possible open plazas, one may be the old excavation.

Opovo matters to people who will never visit the Banat because the Vinča culture belongs to the centuries when people in southeastern Europe first lived in large settlements for generations at a stretch, and whether inequality came bundled with that way of living is an old argument that has not been settled. The village sits near the start of the Eurasian story that Timothy Kohler of Washington State University and his colleagues told in a 2017 paper in Nature, which compared house sizes from sites on several continents and found that wealth disparities after the Neolithic grew larger in Eurasia than in North America or Mesoamerica.

Read more

Cucuteni-Trypillia (Cucuteni in Romanian usage, Trypillia in Ukrainian) is a Copper Age farming culture, the Chalcolithic in technical terms, that spread across what is now eastern Romania, the Republic of Moldova and Ukraine. Its best-known settlements are the megasites of central Ukraine, among them Maidanetske, Talianki and Nebelivka, each covering well over a hundred hectares and ranking among the largest settlements in Europe in the early fourth millennium BCE. The intuitive assumption is that production scales with population, so the biggest kilns ought to turn up in the biggest towns. Where published estimates exist, kilns at those megasites had working surfaces of 2 to 4 square meters, and the largest reported before now, at Stolniceni I, a megasite in Moldova, reached 4.6 to 4.8. Ursoaia covers almost 8 hectares, and its kiln is more than half again the size of the one at Stolniceni I.

The kiln was excavated in 2025 and published this year in Antiquity1 by eleven researchers from Iași, with Diana Măriuca Vornicu of the Romanian Academy’s Institute of Archaeology as first author. One of her coauthors, Felix-Adrian Tencariu of the Institute of Interdisciplinary Research at Alexandru Ioan Cuza University of Iași, wrote the 2015 monograph (in Romanian) on prehistoric and protohistoric pottery-firing installations across Romania, so the verdict that this one is exceptional comes partly from someone who knows the ordinary ones well.

A household firing its own cooking pots and storage jars would have little obvious use for an upper chamber of around 12 cubic meters, and less reason to keep patching such a structure through firing after firing. The size bears on an old argument in archaeology about craft specialization, the point at which making a thing becomes a particular person’s or group’s job, and about whether that shift waits on cities and states or can begin among village farmers. The authors conclude that production at Ursoaia “likely moved beyond domestic requirements” toward a more intensive model. They stop there, and say plainly that calling the site a community-level workshop would take evidence they do not yet have.

A magnetometer survey in 2024 traced a ditch enclosing almost 8 hectares and, inside it, about 60 burnt houses in small clusters, numerous pits and at least eight kilns. Fired clay carries a magnetic signal the instrument can read from the surface, so the figure counts houses that burned, which may not be every house that stood, and the map on its own cannot say how many houses or kilns were in use at the same time. One reading, a dipolar anomaly of up to 22 nanotesla right against the ditch, looked like a firing installation. A trench cut to test it exposed a well-preserved two-chamber updraft kiln (bicameral, in the specialist term), and the decorated pottery around it, together with radiocarbon dates, places the settlement in the phase Romanian archaeologists call Cucuteni B1.

Read more

Anyone who has seen the leather-wrapped bundles of Gran Canaria’s pre-Hispanic dead is likely to carry two assumptions about the caves they came from: that a burial cave in the Canary Islands was made for burial, and that the bodies inside had been dried before anyone carried them up the rock. The second assumption was recently given a specific form. In a 2025 chapter on the first human settlement of Gran Canaria, the archaeologist J. Velasco-Vázquez, working from historical accounts and from museum insect specimens with no record of where in a grave they had been found, proposed that corpses were left in the sun before burial. A study published this year in Antiquity,1 led by Pedro Henríquez-Valido of the University of Alcalá with colleagues in Bordeaux and Las Palmas, uses the insects of this one cave to argue that its dead went in soon after death, onto a platform built over what was left of somebody’s grain.

The name Cruz de la Esquina belongs both to that burial cave and to the complex of about twenty artificial caves around it, and the bones alone could not settle how the dead were handled. When a survey found the cave in 2001, most of the skeletal remains lay loose on the floor, apart from a few vertebrae and ribs from one person that disease had fused, so that they held together after everything around them had come apart. Archaeothanatology, which reconstructs a funeral from the order in which a body’s joints give way, had very little still joined to work with. Scraps of rope, rush fabric and hide survived, and better-preserved burials elsewhere on the island suggest the dead were wrapped in bundles. But the time between death and wrapping decides something basic about how these people treated a corpse, whether it was dried first or put into the rock while still decaying. Study of the skeletons is still under way; the paper gives no count of the dead, though three individuals have been radiocarbon dated.

Read more

Anyone who learned human origins through Lucy probably holds a tidy picture of her genus. In that picture, Australopithecus walked upright on the ground, and its apelike shoulders, curved fingers and long arms were inheritance from tree-dwelling ancestors, no longer doing much. Lucy is the 3.2-million-year-old Australopithecus afarensis skeleton from Hadar, Ethiopia, catalogued as A.L. 288-1. She complicated that picture in 2016,1 when computed tomography (CT) scans suggested her arms had borne a good deal of her weight. But she was tiny, and among living primates the small ones tend to climb more. A study published September 16 in Science Advances 2removes that excuse.

Little Foot was a larger animal, and relative to its thigh, it loaded its upper arm like a gorilla that spends about a quarter of its time in trees. Kristian Carlson, a professor of clinical medical education at the Keck School of Medicine of the University of Southern California, and 15 co-authors compared thigh-shaft strength with upper-arm-shaft strength in more than 1,400 modern humans and apes. They fit a line through each group and measured each fossil’s distance from each line in standardized units. The number that matters is small enough to pass for a rounding error. StW 573 sat 3.2 units below the human line and 0.057 units from the line for female western lowland and Grauer’s gorillas, practically on it.

Researchers have argued for decades over how much australopiths (the informal name for members of Australopithecus) climbed. Zeresenay Alemseged, a paleoanthropologist at the University of Chicago, called the debate deadlocked in a 2023 review in Nature.3 A long arm or a curved finger bone could be climbing gear or a leftover that selection never cleared away. Limb shafts are harder to wave off, because they keep remodeling through life, partly in response to the loads they carry.

Read more

The older thing was a grape seed from Valencia, recovered below the water table in the moat of the city’s Islamic wall and dated by its deposits to the tenth century CE. Ancient DNA in grape seeds comes mostly from the seed coat, which is maternal tissue, so what the team sequenced was in effect the genome of the vine that bore the seed. According to the study, published in Current Biology1 by Guillem Pérez-Jordà of the University of Valencia, Carolina Royo of the ICVV and eleven colleagues, that vine was a child of Heben and a full sibling of Viura, the main white grape of Rioja, known elsewhere in Spain as Macabeo. Most people picture a wine grape the way they would picture a strain of wheat, as the latest in a long line of parents and offspring, so that three thousand years of Spanish winemaking would mean something like three thousand years of vine generations. Heben fits that picture badly. It was setting seed before the tenth-century vine sprouted, and it is alive in Logroño now, one genetic individual kept going by cuttings for at least 1,100 years.

The Valencia seed is one of 28 the team managed to sequence, out of 164 waterlogged seeds put through DNA extraction, from six sites in Spain and one in Sardinia dated between the ninth century BCE and the fourteenth century CE.

Grapevines (Vitis vinifera) were first domesticated in the South Caucasus and the Near East, and how they reached the far western Mediterranean has mostly been argued from the genomes of living vines, which keep traces of ancestry but no dates, whereas seeds from dated layers carry both. Testing the seeds for kinship, against each other and against 140 modern vines, also let the team count how many times Iberia’s cultivated grapes have reproduced sexually since they arrived, and for some lineages the count is two or three.

Read more

Surveys through the early 1970s turned up hundreds of Paleolithic sites in Lebanon, then the civil war that began in 1975 ended plans for further work, and instability has limited fieldwork since. In the southern Levant, meanwhile, a century of digging at Tabun, Skhul, Qafzeh, Amud, Misliya, Nesher Ramla and Qesem built the reference record, and its patterns have often been read as the whole region’s. So the corridor most readers picture, the land bridge that Homo sapiens and Neanderthals crossed between Africa and Eurasia, is a southern one, fringed by steppe and desert and hostage to every swing of the ice ages. The two studies, both in Communications Earth & Environment, argue that the central Lebanese coast was more wooded than that picture allows, and that it changed far less.

Zumoffen’s Ras el-Kelb material is the oldest of the three collections studied. He also gathered bones inside the cave at Nahr Ibrahim, farther up the coast, and a tooth from that lot dates to about 245,000 years, plus or minus 23,000, near the shift from the Lower to the Middle Paleolithic. The third collection comes from Naame, south of Beirut, which another Jesuit, Henri Fleisch, excavated in 1967; its main layers fall within Stage 5, and one tooth dates to 107,000 years, plus or minus 13,000. The three sites lie 15 to 30 kilometers apart, each near a river, on a coast that until now had no high-resolution climate record tied to human occupation.

Gabriele Russo, of the paleoanthropology group at the Senckenberg Centre for Human Evolution and Palaeoenvironment at the University of Tübingen, led both studies; Sireen El Zaatari, a professor in that group who conceived the broader project, supervised. If one stretch of the corridor stayed livable while the rest swung with the ice ages, it would have been open when the south was not, and the southern record would be a poor stand-in for it.

Read more

Mizyn, which older literature often spells Mezin, has yielded 18 of these figurines, the largest group of female figures from Eastern Europe after the Last Glacial Maximum, the coldest stretch of the last Ice Age. Thirteen have a figure engraved on the front, narrow cuts sketching a neck, shoulders, arms, and a triangle long read as pubic, and 11 carry meanders and rows of parallel zigzags on the back and sides. Anyone familiar with Paleolithic art would sort those two kinds of marking without much thought: the front is a picture of a person, and the back is the abstract part, pattern for its own sake or a lost symbol. Simon Radchenko of the Museum of Archaeology at the University of Stavanger in Norway, the study’s first author, and five colleagues argue that the back is the more literal of the two, because its zigzags and meanders copy structures inside the tusk.

Owen’s lines are the tusk’s growth layers, visible on a crosswise cut as concentric rings around the central nerve channel, much like the rings of a tree. Schreger lines cross those rings at an angle and form a crosshatch that ivory specialists also call engine turning or stacked chevrons. Both show best on a clean break or a polished face, and both run through the tusks of the woolly mammoth, Mammuthus primigenius, as they do through elephant tusks.

Schreger lines differ enough between species to matter for wildlife enforcement. Near a tusk’s outer surface they meet at acute angles, under 90 degrees, in mammoth ivory, and at obtuse angles, over 115 degrees, in modern elephant ivory, a difference used to police the elephant ivory trade. The U.S. National Fish and Wildlife Forensics Laboratory in Ashland, Oregon, whose research underpins the method, warns against judging from a single angle, so an inspector checking a carved bangle today measures several, reading the same crosshatch the Mizyn carvers chose to reproduce.

European figurines of women, made over more than 30,000 years, mark two thresholds for archaeologists: the first pictures of human bodies, and the later turn toward schematic ones reduced to an outline and a few cuts. That second turn is usually told as a western story ending in the pared-down Gönnersdorf style, which spread through central Europe between about 17,000 and 13,000 years ago. Dmytro Kiosak of the Leibniz-Zentrum für Archäologie in Mainz, a coauthor, has said the results undercut the old idea that Paleolithic art developed along a single track, simple first and elaborate later. If these are among the first schematic figurines made after the Last Glacial Maximum, and their most abstract-looking marks came out of the ivory, then an early step toward abstraction in Europe was taken by people looking hard at a tusk, on surfaces they had polished smooth.

Read more