레이블이 Academic인 게시물을 표시합니다. 모든 게시물 표시
레이블이 Academic인 게시물을 표시합니다. 모든 게시물 표시

2026년 11월 22일 일요일

Four Letters We Inherited From the Neanderthals: A Deletion That Shrinks the Back of the Brain — and Lowers Schizophrenia Risk [네안데르탈인 — 후두엽, 조현병 낮춘 유전 변이]

LEIPZIG, 16 NOVEMBER 2026 — The case against the Neanderthals was built out of skulls, and for most of the last century the skulls were read as evidence of a mind that had taken a different road and stopped. Longer front to back than ours. Lower in the vault. And at the back of the head, a shelf of bone — the occipital bun — that seemed to say something specific: a large visual cortex, a great deal of brain spent on seeing, less left over for the faculties we had decided to be proud of.

On 16 November, in the Proceedings of the National Academy of Sciences, Hugo Zeberg and Svante Pääbo published a result that makes that century of reading harder rather than easier. A common variant in the human genome that shrinks part of the occipital lobe — the cuneus, a fold on the inner surface at the back of the brain — was not invented by us. It came from Neanderthals.

Each copy a living person carries is worth about 0.85 cubic centimetres less cuneus, an effect on that structure larger than any other variant known; by one measurement, carriers’ occipital lobes are roughly a tenth smaller in volume. And the same variant is associated with something nobody was looking for in a story about brain size: in people who carry two copies, the risk of schizophrenia is about 10 per cent lower.

The variant that shrinks the visual brain, in other words, is also the variant that appears to protect it from one of the most heritable disorders of the mind. Then comes the turn that keeps the paper in the conversation. Neanderthals had larger occipital lobes than we do. The variant they gave us makes ours smaller. The debt runs against the direction of the loan.

La Ferrassie 1 Neanderthal skull in left lateral view
The skull that started the argument: La Ferrassie 1, one of the most complete Neanderthal skeletons ever recovered, excavated in the Dordogne in 1909 and photographed here in left lateral view. Low, long braincase, receding forehead, and the shelf of bone at the back — the occipital bun — that made anatomists think the visual brain had grown at the expense of everything else. (Image: Wikimedia Commons user “120”, CC BY 2.5)

The bone deserves its own paragraph, because everything rests on how much of a brain a skull can be made to confess. A cranium is a mould of the brain taken from the outside, and the inside of that mould — the endocast, in the older literature — reproduces the cortex the way a pillow reproduces a face: the gross divisions, the big sulci, the relative proportions of the lobes, and nothing of the tissue.

On that coarse scale the difference between the two species is real and repeatable. Neanderthal braincases are elongated, flattened front to back, widest toward the rear; modern human braincases are globular, with a high rounded vault and a bulging parietal region, a shape that appears late and quickly in our own fossil record. The Neanderthal occipital region is, cast after cast, relatively larger. Averaged across the catalogue the Neanderthal brain is not smaller than ours — if anything slightly larger — which makes the shape the interesting part.

From shape, generations of palaeoanthropologists inferred a shift in the internal budget: more cortex assigned to the back of the head, to vision, to the machinery of a body and a landscape that may have demanded it. Marcellin Boule’s reconstruction of the old man of La Chapelle-aux-Saints, published in 1911, turned that inference into a hunched, ape-like brute; much of the posture was an artefact of the specimen’s severe arthritis, and the correction took decades. La Ferrassie 1, lifted from a rock shelter in 1909, carries the same low long vault and the same bun. The skull is not a blank, in other words. It says, in a voice nobody now disputes, that the two brains were built on different plans. What it cannot say is why, or by which letters.

Zeberg and Pääbo went looking for the letters. Their method is a filter with a strong prior. A variant in living people that came from Neanderthals should satisfy three conditions at once: high frequency in Neanderthal genomes, presence in some living populations, and absence from people of entirely sub-Saharan ancestry, whose ancestors by and large never met a Neanderthal.

The pair worked with the five Neanderthal genomes detailed enough to be searched base by base, and with the two Denisovan genomes sequenced so far. The variant cleared every gate. It sits in all five Neanderthals. It is absent from both Denisovans, a cousin species — so it is not simply a property of archaic hominins in general. And in living people it is confined to populations with non-African ancestry.

The last step of the argument is arithmetic that has become standard in the field and is still worth spelling out. The variant lies inside a stretch of DNA 59.7 kilobases long that carries the Neanderthal pattern throughout. A segment inherited from a shared ancestor would have been broken up by recombination over hundreds of thousands of years into a fragment far too small to recognise.

A block that size entered the human gene pool whole, when the two species met in the last 50,000 to 60,000 years, and has not yet had time to dissolve. The variant itself is a deletion: four nucleotides missing on chromosome 14. It was first described in 2019, in a genome-wide study of the volume of the brain’s lobes, as a locus for occipital volume — and nothing more; its archaic origin is a much later reading.

It lies not in a gene but in the predicted recognition motifs for transcription factors, beside DAAM1 — a gene whose protein helps build and organise the actin cytoskeleton, and which has been tied to the planar cell polarity pathways that shape a developing nervous system. Carriers, the authors report, show altered DAAM1 expression. The four missing letters are not a broken part. They are a change in the settings.

Model of the DNA double helix
The letters, not the skull: four nucleotides missing on chromosome 14, inside a 59.7-kilobase block of DNA that still carries its Neanderthal pattern — a block far too large to be an inheritance from the common ancestor of the two species, and therefore a signature of interbreeding roughly 50,000 years ago. (Image: National Human Genome Research Institute, public domain)

Then the part that turns a genetic curiosity into a number. The authors brought the variant to the UK Biobank, which holds genotypes for half a million volunteers and brain images for a subset of them, and to the 1000 Genomes Project for the population picture. In the imaging data, each copy of the Neanderthal allele is associated with a cuneus about 0.85 cubic centimetres smaller.

Two copies, twice the subtraction. Reported as a proportion, carriers’ occipital lobes come out about ten per cent smaller in volume than those of people carrying the ordinary, non-Neanderthal version of the same stretch of chromosome. The cuneus is a wedge-shaped fold on the medial face of the occipital lobe, between the calcarine and the parieto-occipital sulci, part of the earliest cortical handling of what the eye brings in; it is not the whole of vision, and a smaller cuneus is not the same as seeing less well.

Zeberg’s own gloss on the magnitude is the line that will be quoted: “If you ordered a shot at a bar, that’s about how much brain volume it is.” About 20 per cent of people of European descent carry at least one copy — a figure that should stop anyone inclined to read this as a defect. A variant at that frequency, moving a structure of that size, spread across a continent, is not a rare accident. It is part of the ordinary equipment.

The finding that made the paper news is the one with nothing to do with seeing. In the same data, people carrying two copies of the Neanderthal variant were about 10 per cent less likely to have been diagnosed with schizophrenia — “almost 10%” in the paper’s own wording. Two qualifications keep this from becoming a headline about a Neanderthal gene against madness.

The first is that the association belongs to this one variant and not to Neanderthal ancestry as a whole: the authors checked, and the genome-wide proportion of Neanderthal DNA shows no such tendency, which makes the effect a property of a particular regulatory change rather than of being, in some overall sense, more Neanderthal. The second is that the mechanism is unknown. “We don’t know the details behind that interaction,” Pääbo said. “The gene is involved in many aspects of brain development.”

What the variant does not do is make people cleverer or duller. The paper examined cognition and schooling and found at most about two weeks of additional education per allele, against an average of roughly 15 years — a difference small enough to be an artefact of very large samples.

The temptation, which the authors raise and then decline to endorse, is that the two findings fit an old speculation: that the human cognitive package, with its unusual appetite for invention, arrived with a raised risk of psychosis, and that the Neanderthal version of this switch keeps some of the protection our own version traded away. “That would all be speculative,” Pääbo said.

“I think what we can say is that it was at least not very bad to carry it.” One loose thread is worth naming here: in 2021, Karen Berman and colleagues at the National Institute of Mental Health reported that people with more Neanderthal ancestry overall had lower rates of schizophrenia — a different framing from this paper’s careful “not ancestry in general,” and a reminder that “more Neanderthal” is not one quantity but a sum of thousands of separately behaving pieces.

Lobes of the human brain, lateral view, with the occipital lobe labelled
The back of the head, from outside: the lobes of the human brain in lateral view, with the occipital lobe in brown at the left. The cuneus, where the variant’s effect concentrates, is on the inner surface and hidden in this view, between the calcarine and parieto-occipital sulci. It is there that each copy of the Neanderthal allele takes away about 0.85 cubic centimetres. (Image: Sebastian023, CC BY-SA 3.0)

Which brings the argument back to the bone, and to the objection the paper cannot answer. Thomas Schoenemann, an anthropologist and cognitive scientist at Indiana University Bloomington, put it plainly: if this variant did in Neanderthals what it appears to do in us, their occipital lobes ought to be relatively small. The casts say they were large.

Either the variant behaved differently in the genome that evolved it — plausible, since its effect depends on the company it keeps, and a Neanderthal genome is not a human genome with a few archaic inserts — or the cuneus and the occipital lobe are not the same story, or the skull’s message has been over-read after all. “It calls into question what exactly we can say about its effects in Neanderthals at all,” Schoenemann said.

The authors do not claim otherwise: the paper’s finding is about present-day humans, and it is careful to say so. What it adds is a piece of population history that sits oddly with any simple story of benefit. Tracking the variant across ancient and modern genomes, Zeberg and Pääbo found that its frequency has stayed roughly constant for tens of thousands of years. Evolution has neither swept it to fixation nor purged it — the signature of something close to neutral in the currency selection actually trades in.

That makes the 20 per cent frequency in Europeans, against the 1 to 2 per cent average of Neanderthal ancestry, a matter of drift, of the founder populations that carried it out of Africa and into Europe, and possibly of small local advantages that never added up to a sweep.

Why the cuneus shrank, what altered DAAM1 expression does to a developing cortex, whether the association with schizophrenia survives in a cohort that is not overwhelmingly European — these are the questions the paper hands on. Organoids and animal models are the next instrument, and the honest position remains the one Zeberg stated: the mechanics are not understood.

Amud 1 Neanderthal cranium, frontal view
The people who left the letters: Amud 1, a Neanderthal cranium from Amud Cave in Israel, held at the National Museum of Natural History. The variant appears in all five detailed Neanderthal genomes, in neither of the Denisovan genomes sequenced so far, and in living people only among populations of non-African ancestry. (Image: Jonathan Chen, CC BY-SA 4.0)

Step back from the specifics and the paper becomes a lesson in what a genome is. We tend to speak of Neanderthal ancestry as a stain or a residue — one or two per cent of our DNA, a souvenir of an encounter that ended badly for one of the parties. That framing gets the direction of the relationship wrong. The archaic contribution is not a smudge on a finished text; it is a set of working parts, some of them running inside the most consequential organ we have.

A regulatory switch carried by one European in five, sitting in the predicted binding sites of transcription factors, adjusting the expression of a gene that organises the cytoskeleton of developing neurons, and leaving its signature in the volume of a fold at the back of the brain — that is not a relic. It is equipment in service. That it also shifts the odds of a psychiatric diagnosis is the kind of coupling genetics keeps turning up and our categories keep failing to expect.

Forty thousand years after the last Neanderthals died, four missing letters on chromosome 14 are still tuning the back of the human head, in a direction that runs opposite to the skulls which first told us the two species were different. The old man of La Chapelle, reconstructed as a brute, was a lesson in how easily bone is made to serve an argument. The lesson of this paper is subtler and harder to be smug about: the brain we call ours is a compromise reached in a hurry between at least two lineages, and parts of the compromise are still being audited.

 KEY FACT-CHECK
• The paper: Hugo Zeberg & Svante Pääbo, “A Neandertal-derived genetic variant reduces occipital brain volume in present-day humans,” PNAS 123 (48), e2624837123, published online 16 November 2026, print issue 1 December 2026, DOI 10.1073/pnas.2624837123, PMID 43038442. Zeberg: Max Planck Institute for Evolutionary Anthropology, Leipzig, and Karolinska Institutet, Stockholm. Pääbo: Max Planck Institute, Leipzig, and the Okinawa Institute of Science and Technology. Both are corresponding authors.
• The variant: a deletion of four nucleotides on chromosome 14, inside a 59.7-kilobase block that carries the Neanderthal pattern; first described in 2019 in a genome-wide association study of lobar brain volume (van der Lee et al., Communications Biology 2, 285) as a locus for occipital volume. It lies in predicted transcription-factor recognition motifs near DAAM1, and carriers show altered DAAM1 expression.
• Distribution: present in all five detailed Neanderthal genomes; absent from both Denisovan genomes sequenced so far; in living humans confined to non-African ancestry. About 20% of people of European descent carry at least one copy (as reported by Science).
• Effect on the brain: 0.85 cm³ less cuneus volume per copy — an effect larger than that of any other known variant on that structure. Two copies therefore about 1.7 cm³. Carriers’ occipital lobes measured in the UK Biobank imaging subset were about 10% smaller in volume than non-carriers’ (reported by Science); the per-copy figure is the paper’s own.
• Schizophrenia: risk reduced by “almost 10%” in homozygotes (paper). The paper explicitly states that Neanderthal ancestry as a whole shows no such tendency. Karen Berman’s 2021 study reported lower schizophrenia rates in people with more Neanderthal ancestry overall — a different measure, and not a contradiction to be resolved by either side alone.
• Cognition: at most two weeks of additional schooling per allele against an average of about 15 years — small or nonexistent (paper). The creative-trade-off interpretation is raised by the authors as speculation and left there.
• Not claimed: that Neanderthals had small occipital lobes (they did not), that the variant caused their anatomy, that it directly causes or prevents schizophrenia, or that carrying more Neanderthal DNA is protective in general. Mechanism unknown; organoids and animal models proposed as the next step.
• Critique on the record: Thomas Schoenemann (Indiana University Bloomington) notes that the occipital-volume link is not a clean result, since the same variant in Neanderthals should have produced small occipital lobes, and did not.
• Sources: paper (DOI above); Nature News, 16 Nov 2026; Science News, 16 Nov 2026; Scientific American, 16 Nov 2026; Smithsonian, 18 Nov 2026.

[심층 리포트] 뼈와 유전자가 서로 다른 말을 할 때 — 네안데르탈인이 남긴 네 글자

라이프치히, 2026년 11월 16일 — 네안데르탈인에 대한 판결은 뼈로 내려졌다. 지난 한 세기 동안 그 뼈는 ‘다른 길로 가다 멈춘 정신’의 증거로 읽혔다. 우리보다 앞뒤로 길고, 정수리는 낮았다. 그리고 뒤통수에는 뼈가 선반처럼 튀어나온 자리가 있었다. 후두융기(occipital bun)라 불리는 그 자리는 꽤 구체적인 말을 하는 듯했다.

Korean Version: [심층 리포트] 뼈와 유전자가 서로 다른 말을 할 때 — 네안데르탈인이 남긴 네 글자

Socko/Ghost

2026년 11월 21일 토요일

The Cousin Nobody Looked For: 180 Million People Carry the Closest Relative of Smallpox in Their Gut [아무도 찾지 않던 사촌 — 1억 8천만 명의 장 속에 사는 천연두 바이러스의 가장 가까운 친척]

PARIS, 18 NOVEMBER 2026 — For three decades the search for giant DNA viruses has been a story of abundance. They turned up in seawater and lake sediment, in soils and wastewater, in the plankton of the open ocean and in the amoebae that graze on all of it — one new lineage after another, some of them large enough to be seen under an ordinary light microscope.

Inside animals, the story was different. Surveys of the human gut kept returning stray signatures and little else: viruses passing through with food and water, present for a day or a week, never resident. That absence was more than a hole in a catalogue. Poxviruses — the family of variola, the smallpox virus estimated to have killed half a billion people in the twentieth century, of mpox, of vaccinia, the workhorse of molecular biology — sit oddly inside the giant-virus world, because they infect animals directly instead of single-celled organisms, and nobody could say when or where that jump happened.

The intermediate forms were missing, and so was the habitat they should have come from. On 18 November, a team coordinated from Genoscope, the French national sequencing centre outside Paris, and the CNRS reported in Nature that the emptiness was not biological. It was a problem of where the search had looked, and of what it had been counting.

The method was a fishing expedition through other people’s data. A metagenome is a library of DNA sequenced from a whole sample — a stool sample, a pond, a buffalo’s rumen — without separating the organisms first; assemble enough fragments and genomes of things nobody has ever cultured fall out of the noise.

The team screened 85,123 such assemblies, from the environment and from the digestive systems of humans, livestock and wild animals, using a statistical profile of one gene: RNA polymerase subunit A, a marker carried by nearly every virus in the phylum Nucleocytoviricota.

The screen returned 1,969,342 RNApolA genes. Sorting them by group produced the picture everyone expected — 240 asfuviruses, every one of them from the environment, mostly aquatic, and a single poxvirus, closest to Eptesipox, in the gut of an Australian marsupial carnivore — and then something that had no place in that picture at all: a clade of large, linear DNA viruses that kept appearing in guts. From it the team assembled 12 complete genomes, each less than 98 per cent identical to the rest, from humans, buffalo, goats, pigs, chickens and deer.

They run from 160 to 360 kilobases, longer than a small bacterial genome; they carry the 55-base-pair inverted terminal repeats that mark the class; they include all eight hallmark genes of the phylum; and in most of them the DNA polymerase B gene is split into two separate reading frames, a piece of genomic housekeeping that echoed, unexpectedly, a trans-splicing mechanism recently described in a completely different viral group. A second screen keyed to the major capsid protein widened the net to 864 positive assemblies and more than 200 genomes, the longest reaching 467 kilobases. They called the order Egovirales, the egoviruses.

Colorized scanning electron micrograph of monkeypox virus particles on the surface of an infected cell
The family egoviruses belong to: monkeypox virus particles (orange) on the surface of infected cells, colourised scanning electron micrograph, NIAID Integrated Research Facility. Egoviruses are the closest known relatives of this family — but no egovirus itself has ever been seen. Their shape is inferred from genes, not photographed. (Image: NIAID, CC BY 2.0, via Wikimedia Commons)

What those genomes describe is a virus that lives wherever its hosts live. In a focused survey of 11,788 metagenomes from 26 projects on the digestive tract — 17 of them covering humans — egovirus sequences appeared in people of every age and on every continent sampled: 193 human samples, 2.3 per cent of the positive set. The figure rose to 6.2 per cent among the Hadza, the hunter-gatherers of northern Tanzania, and three of the six human-specific clades turned up almost exclusively there, a hint that lifestyle shapes which varieties a gut carries.

Among livestock the signal was stronger: 11 per cent of pig samples, 14 per cent of goat, 26 per cent of buffalo, 7.8 per cent of cow, 11 per cent of chicken. Six clades belong to humans alone — five to the gut, and one, clade 03, known only from saliva, the first sign that egoviruses occupy more than one body site. The dominant one, clade 01, has now been characterised in 186 assemblies from 19 studies spanning Tanzania, Nepal, Fiji, the United States, England, Sweden, Denmark, Germany, Italy and Spain, with average amino-acid identity of 99.4 per cent across its RNApolA genes and 99.9 per cent across its capsid genes.

That is not a cloud of variants. It is one viral population, distributed across continents, and it accounts for 92 per cent of the egovirus-positive human gut samples. It appears in a three-year-old child and in a pregnant woman; in longitudinal studies it persisted in the same individuals for nearly a year. Extrapolating with United Nations population figures, the authors project roughly 180 million people carrying egoviruses, and possibly most of the world’s 200 million buffalo. Almost none of this exists outside animals: only nine egovirus-like sequences in the entire survey came from the environment — an aquifer, a cold seep, marine sediment — while more than 97 per cent of the RNApolA sequences and more than 99 per cent of the capsid sequences came from digestive systems.

Histological cross-section of small intestinal villi of the terminal ileum
The habitat, at the scale of tissue: a cross-section of the villi of the terminal ileum, H&E stain. Somewhere in the mucus and the anaerobic pockets between these folds live the single-celled protists that egoviruses are predicted to infect — and, by inference, the egoviruses themselves. (Image: Mikael Häggström, CC0, via Wikimedia Commons)

No egovirus has been caught in the act of infecting anything. The case for its host is built from borrowed genes. When the team compared the viral genomes with the genomes of eukaryotes and measured the flow of genes between them, two groups lit up and nothing else did: the phylum Parabasalia, which took 43 per cent of the egovirus-to-eukaryote signal, and the order Trichostomatia, with 42 per cent. Both are single-celled, both are anaerobic, both live in animal digestive tracts, and both feed by phagocytosis — swallowing particles whole, the same door that large DNA viruses use to get inside a cell.

Within Parabasalia the signal traced almost entirely to Trichomonas and Tritrichomonas, genera that include the most common non-viral sexually transmitted infection in humans, roughly 150 million cases a year, and a cattle parasite that also lives in the intestines of cats and the stomachs of pigs. Within Trichostomatia it was confined almost entirely to one family, Ophryoscolecidae, the ciliates that do much of the ruminant’s digestive work.

The negative control is as persuasive as the positive one: Blastocystis, another gut protist that coexists with egoviruses in human, pig and chicken samples, showed no gene flow at all. If the signal were an artefact of sharing a neighbourhood, Blastocystis would have shown it too.

It did not. And there is a fossil of the relationship: a gene called KilA, a core gene of both poxviruses and egoviruses, sits inside polintons — endogenous viruses integrated into the genomes of Trichomonas vaginalis — in two distinct lineages, 1,116 copies of one type and 718 of the other, scattered across all chromosomes of the 176-megabase genome. The most economical explanation is that egoviruses donated the gene, twice, to viruses that then multiplied inside the trichomonad, and that this happened before Trichomonas and its relatives diversified.

The evolutionary result is the headline. Phylogenomic trees built from four conserved proteins place the egoviruses as the sister group of the poxviruses, a distinct main branch inside the class Pokkesviricetes — the class that until now held two orders and explained neither of them. What the genes say about the particle is the strange part. Egoviruses carry homologues of pp220 and pp62, the structural polyproteins otherwise exclusive to asfuviruses, the ones that build a multilayered capsid in the African swine fever virus; their major capsid protein averages 806 amino acids, unusually large, with a highly variable outer face.

The predicted virion is therefore a multilayer icosahedron, like the asfuviruses and unlike the brick-shaped particles of poxviruses. They also carry a histone doublet, a gene pair proposed to pack the genome into nucleosome-like structures — known in other branches of the phylum, never before in this one. Put together, the tree and the capsid tell a sequence of events.

The common ancestor of egoviruses and poxviruses infected anaerobic, phagocytosing single-celled organisms inside the gut of ancestral animals, before the split of protostomes and deuterostomes roughly 700 million years ago. A lineage of that ancestor then acquired the ability to reproduce in animal gut epithelial cells. The capsid was rebuilt into a brick. The poxvirus family, on this reading, learned to be a virus of animals by first living inside the animals’ dinner.

What has not been done is as important as what has. No egovirus has been cultured, isolated or photographed; the capsid is a prediction from sequence, and every statement about what the virus does inside a cell is inference from gene flow. Whether egoviruses cause disease in humans or animals is not known, and the paper does not claim otherwise.

“We do not yet know whether egoviruses can cause disease in humans or animals,” said Subir Sarker, a molecular virologist at James Cook University in Townsville. “We need to isolate and culture egoviruses and test them in animal experiments to find out.” Anjali Ghoripalan, a postdoctoral researcher at the Australian National University in Canberra, put the same caution on the relationship itself: “Although egoviruses share some proteins with poxviruses, additional research is required to determine how closely related the two groups really are.

We are still largely guessing what they actually look like, so we need to culture them and observe them under a light microscope.” Fasséli Coulibaly, a structural biologist at Monash University, was more generous about the place of the find: poxviruses are an outlier in the giant-virus world, he told Nature, and their origin is a mystery; egoviruses could be the bridge between the microbe-infecting asfuviruses and the pathogenic poxviruses.

There are other limits. Metagenomics only sees what is abundant enough to assemble, so the survey is a floor, not a census; sampling is heavily skewed toward humans and livestock, and the authors say the diversity in wild animals is probably substantially undersampled. The 180 million is a projection from detection rates and population statistics, not a count of people.

And the numbers have already begun to drift in transmission: the paper projects “roughly 180 million” people, the French institutional release says “more than 160 million” and separately “more than 2 per cent of the world population,” and the most prevalent figure — more than 1 per cent, for clade 01 alone — measures a different thing from the total. One per cent of the world’s roughly 8.2 billion people would be about 82 million. The three numbers are not interchangeable, and treating them as one is how a carefully hedged estimate becomes a fact in a headline.

What the paper actually changes is smaller and more durable than the headlines. It adds a third order to a class that held two, and it plants the most consequential family of animal viruses inside the family tree of the gut microbiome — not as a curiosity, but as the probable ancestral state.

For human medicine the immediate consequence is a question rather than a treatment: a virus present in the intestines of more than a hundred million people, for up to a year at a stretch, in children and adults, is now something whose effects on immunity, on the microbiome, on the course of gut disease can be studied where before it was invisible.

For veterinary virology it is more concrete: buffalo, pigs and chickens carry egoviruses at rates high enough that any survey of livestock viromes will now have to account for them. The work that follows is unglamorous and slow — find a host cell line the virus will grow in, get a particle on an electron microscope, infect something and watch. Until that happens the discovery sits exactly where most metagenomics sits, at the edge of what can be claimed. But the shape of the claim is already clear enough.

Every poxvirus that has ever infected a person — variola, vaccinia, mpox, the molluscum virus that troubles children — belongs to a lineage whose closest known relative is a virus living quietly in the intestines of roughly 180 million people, and probably living there since before there were vertebrates to have intestines. It was not rare. It was not even hidden. It was simply in a place where, for thirty years, nobody had thought the family would be.

 KEY FACT-CHECK
• The paper: Morgan Gaïa, Ulysse Guyet, Hans-Joachim Ruscheweyh, A. Murat Eren, Shinichi Sunagawa, Eugene V. Koonin, Mart Krupovic & Tom O. Delmont, “Closest relatives of poxviruses prevail in human and animal guts worldwide,” Nature, published 18 November 2026, open access, DOI 10.1038/s41586-026-11283-z. Work coordinated from Genoscope (CEA) and the CNRS; a Research Briefing in the same issue is titled “Poxviruses probably evolved in the digestive system of animals.”
• Method: hidden Markov model of RNA polymerase subunit A screened against 85,123 metagenomic assemblies, yielding 1,969,342 RNApolA genes; a second survey used the double-jelly-roll major capsid protein (DJR-MCP). 864 egovirus-positive assemblies; 12 complete genomes (average nucleotide identity below 98%) plus more than 200 metagenome-assembled genomes, 81% complete on average, mean length 229 kb, maximum 467 kb. Prevalence survey: 11,788 metagenomes with at least 5 million reads from 26 gut-related projects.
• The finding: a proposed third order of the class Pokkesviricetes, Egovirales (“egoviruses”), with linear genomes of 160–360 kb in the complete set, 55-bp inverted terminal repeats, all eight phylum hallmark genes, and a DNA polymerase B gene often split across two reading frames. Sister group to the poxviruses.
• Distribution: six human-specific clades (five gut, one saliva only); 193 human samples, 2.3% of the positive set, rising to 6.2% among the Hadza; pigs 11%, goats 14%, buffalo 26%, cows 7.8%, chickens 11%. Clade 01 found in 186 assemblies from 19 studies on several continents, 99.4% amino-acid identity (RNApolA) and 99.9% (capsid), and in more than 500 additional archive metagenomes. Only nine environmental egovirus-like sequences were recovered.
• Predicted host: Parabasalia (43% of egovirus–eukaryote gene flow; mainly Trichomonas and Tritrichomonas) and Trichostomatia (42%; almost all Ophryoscolecidae). Blastocystis, which co-occurs with egoviruses, showed no gene flow — the study’s internal control. No infection of any host has been observed directly.
• Virion and evolution: multilayered icosahedral capsid predicted from asfuvirus-type structural polyproteins pp220 and pp62, unlike the brick-shaped poxvirus virion; DJR-MCP averages 806 amino acids; a histone doublet is present, a first for this class. Scenario: an egovirus-like ancestor infected anaerobic, phagocytosing protists in the gut of ancestral animals before the protostome–deuterostome split, about 700 million years ago; a later lineage moved into animal gut epithelial cells and its capsid was rebuilt.
• Not established: no isolate, no culture, no microscope image, no animal model; whether egoviruses cause disease in humans or animals is unknown. The authors state that wild-animal diversity is probably substantially undersampled, and metagenomics detects only genomes abundant enough to assemble.
• Reported discrepancies: the paper projects “roughly 180 million” human carriers from metagenomic signal and UN population statistics; the CEA and CNRS release of 19 November says “more than 160 million” and, in its body text, “more than 2 per cent of the world population.” The “more than 1 per cent” figure refers to the single most prevalent human clade (clade 01), not to all egoviruses; 1 per cent of about 8.2 billion people is roughly 82 million. Press accounts that equate the 1 per cent and the 180 million are merging two different calculations.
• Sources: Nature (DOI 10.1038/s41586-026-11283-z; News, 18 Nov 2026; Research Briefing); CNRS and CEA press releases (19 Nov 2026); Dong-A Science (19 Nov 2026); Phys.org (19 Nov 2026).

[심층 리포트] 아무도 찾지 않던 사촌 — 천연두 바이러스의 기원은 동물의 뱃속에 있었다

2026년 11월 18일, 프랑스 파리 근교의 국가 염기서열 해독 센터 제노스코프(Genoscope)와 프랑스 국립과학연구센터(CNRS)가 함께 이끈 국제 공동 연구팀이 Nature에 논문을 냈다. 30년 동안 거대 DNA 바이러스를 찾아다닌 사냥의 결말은 늘 풍요였다.

Korean Version: [심층 리포트] 아무도 찾지 않던 사촌 — 천연두 바이러스의 기원은 동물의 뱃속에 있었다

Socko/Ghost

2026년 11월 20일 금요일

The Word Before the Word: A Nickel-Sized Patch Above Your Left Ear Rewrites 150 Years of Brain Science [왼쪽 귀 위 동전, 150년 뇌과학을 다시 썼다]

BRESLAU, 1874 — A German neuroanatomist named Carl Wernicke sat with a group of patients who had one thing in common: a wound in the same strip of brain, just above and behind the left ear. They could hear perfectly well. They could also speak — fluently, at a normal pace, with the rhythm of ordinary conversation.

The problem was what came out. Words arrived in the wrong places. A patient meaning to say “shoe” said “foot.” Someone reaching for “leg” produced “peg.” Sentences dissolved into what clinicians still call, without much charity, word salad. Wernicke wrote up the cases that year in a short monograph, Der aphasische Symptomenkomplex, and proposed something that would sit in textbooks for a century and a half: that this patch of temporal cortex holds the “auditory image of words” — a template of how a word should sound, from which its meaning and its pronunciation could both be recovered.

Then the field simplified him. The patch became Wernicke’s area, the listening centre. Speech production was handed to Broca’s area, in the frontal lobe, and the two were drawn as separate rooms joined by a corridor of nerve fibres. The picture was clean enough to survive 150 years of neuroscience. On 19 November 2026, in Science, a team at the University of California, San Francisco published the experiment that finally breaks it — and the answer was hiding in a space about the size of a nickel.

The paradox Wernicke himself could not explain is embarrassingly simple. If this region is a listening device, why does destroying it also destroy speaking? Every neurology student learns the bedside signs: comprehension gone, repetition broken, output fluent but wrong.

Clinicians filed the production deficit under comprehension — the patient speaks badly, the reasoning went, because he cannot hear himself and correct his own errors. That is the “proofreader” hypothesis, and it is testable: if Wernicke’s area only monitors speech after the fact, its activity during speech should follow the sound, not precede it. For decades nobody could settle the timing.

Functional MRI resolves brain activity over one to two seconds; the events in question happen in a few hundred milliseconds, and the difference between planning a word and hearing yourself say it is smaller still. “The basic question has been, ‘Why would damage to a hearing area of the brain break the ability to speak?’” the study’s senior author, the neurosurgeon Edward Chang, said at a briefing ahead of publication. “That question has been open for 150 years.”

Wernicke's area shown in red on a 3D model of the left cerebral hemisphere
Wernicke’s area (red): the posterior section of the left superior temporal gyrus, roughly a hand’s width above the left ear. The study isolates a subzone of it, the left posterior STG, as the place where the next word is assembled. (Image: Polygon data from BodyParts3D, Database Center for Life Science, CC BY-SA 2.1 JP, via Wikimedia Commons)

Chang’s laboratory did not set out to overturn a textbook. It set out to use one. For fifteen years the lab has worked with epilepsy patients who come to UCSF for surgery and spend part of the operation awake, with a grid of electrodes laid directly on the exposed surface of the brain — high-density electrocorticography, the same clinical procedure surgeons use to map which tissue can be cut without costing the patient language.

The electrodes record neural populations at a resolution an MRI cannot approach, millisecond by millisecond. To that the team added a newer instrument: Neuropixels probes, recording the activity of single neurons. Across 28 participants, they ran tasks from naturalistic storytelling to simple word repetition, separating the moments of listening from the moments of speaking. Then they did the part that correlation alone cannot do.

In 13 patients they passed small currents of electricity through the electrodes, at carefully chosen timings, and watched what happened to speech. And they went back over the records of hundreds of patients who had had tumours removed from the temporal lobe, to see whose speech suffered when the surgeon’s scalpel passed through which millimetre. Three methods, one question: what is this tissue for?

The answer arrived about a quarter of a second early. In the left posterior superior temporal gyrus — lpSTG, a discrete subzone inside what everyone calls Wernicke’s area — populations of neurons increased their firing roughly 250 milliseconds before a word was spoken, and they did it in continuous speech, ahead of each word rather than in response to the sound of the last one.

The activity was separable from the auditory-feedback signals in neighbouring tissue, which behave the way a proofreader would: they answer the voice. The lpSTG was doing something else. It was already holding the word. Its cells were tuned not to individual sounds or to the movements of the tongue and jaw, but to properties of whole words — how long they are, how many syllables they carry, how often the language uses them.

“It’s more active when you’re about to say the word caterpillar than when you’re about to say the word cat,” said Matthew Leonard, a cognitive neuroscientist at UCSF and co-first author. The lead author, Alexander Silva, put it more sharply: what the region holds is an “auditory image” of the word you are about to say. Wernicke’s phrase, it turns out, was not a metaphor. It was a description of the file format.

Illustration of a subdural electrode grid placed over the surface of the brain during awake surgery
Illustration of a subdural electrode grid placed over the surface of the brain during awake surgery

Correlation is a description; causation is a verdict, and the verdict came from the operating room. When the researchers stimulated the lpSTG in the moment just before a patient began to speak, the patients made word errors that looked like Wernicke’s aphasia — substitutions and near-misses, the classic signature. Stimulation elsewhere, or at the wrong moment, did not produce the same effect.

The surgical records pointed the same way: patients whose tumour removal had encroached on this subzone were significantly more likely to come out of the operation with speech problems, even though the same operation spares tissue that merely listens. “We already knew that damage to Wernicke’s area can lead to speech production errors, but this publication answers why,” Adeen Flinker, a neuroscientist at NYU who wrote the accompanying commentary in Science and was not involved in the work, told Live Science.

The single-neuron recordings added the strangest detail of all. Some cells in the lpSTG fired selectively for particular words, and the same cells came alive both when the patient said the word and when the patient heard someone else say it. The machinery that builds a word and the machinery that recognises one are not merely neighbours. They overlap.

That overlap is where the clinical stakes are. Speech neuroprostheses — the implants that have let paralysed patients type and speak again in the past few years — have mostly read signals from the motor cortex, the last relay station before the muscles. That works for people whose damage sits below the cortex: ALS, brainstem injury.

It does not work for the far larger population with aphasia, whose language cortex itself is injured; the signal they need has already been scrambled by the time it reaches the motor areas. If speech is planned upstream, in a region that also listens, then decoding there could reach patients that motor-based devices cannot. The team tested the idea in volunteers with severe paralysis after brainstem strokes, who had electrodes implanted for a speech brain-computer interface.

Reading from the lpSTG, the researchers decoded words the participants were trying to say — and, in the same signal, words they were hearing. “What if you could target a single brain region and gain access to the full spectrum of speech, from listening all the way up to speech production?” Silva said. Chang went further at a press briefing, in terms that are as much about human origins as about medicine: this small subzone that turns the words we hear into words we can say may be part of what makes us human.

What the paper does not show is as important as what it does. The paralysis demonstration rests on a handful of participants — the university’s release describes one, the news accounts and the paper report two — and the causal evidence comes from stimulation that lasts seconds inside an operating theatre, not from a chronic implant.

Nobody has yet shown that lpSTG signals stay stable for months or years, which is the entire engineering problem of a neuroprosthesis; the decoding proves a principle, not a product. The lesion evidence is retrospective, drawn from patients who were already having brain surgery for tumours, and the frequency finding is already being described in opposite directions by careful reporters, one writing that activity rose for uncommon words, another for familiar ones — a reminder that press coverage of a 20-page paper is a summary of a summary.

More interesting is the question the paper opens in the frontal lobe. If the main action in both comprehension and production happens in the temporal cortex, what is Broca’s area for? “It is still clearly important,” Leonard said. “In some ways, it’s sort of the next mystery.” Evelina Fedorenko, a neuroscientist at MIT who was not involved, asked the question more bluntly: if both halves of language are happening in the temporal lobe, why do we have the frontal components at all?

Wernicke, it seems, was right about the junction and wrong about the single function — and the brain has one fewer clean room than the textbooks promised, which is the usual price of finally looking at it closely. Every word you said today began, about a quarter of a second before you said it, in a patch of tissue the size of a coin above your left ear. You simply were not in the habit of noticing.

 KEY FACT-CHECK
• The paper: Silva, Leonard, Hullett, Levy, Wang, Xu, Bhaya-Grossman & Chang, “A neural code for generative speech production in human Wernicke’s area,” Science 394(6826), p. 858, published 19 November 2026. DOI 10.1126/science.aef8963. Commentary: Adeen Flinker (NYU), “Pinpointing planned speech,” same issue, p. 841. Results were also presented at the Society for Neuroscience annual meeting on 15 November 2026.
• Methods: 28 participants with epilepsy or tumours, already implanted with intracranial electrodes for clinical monitoring; high-density electrocorticography plus single-neuron (Neuropixels) recordings; direct electrical stimulation in 13 patients; retrospective lesion analysis of hundreds of patients who had temporal-lobe tumour surgery.
• The finding: a discrete subzone of the left posterior superior temporal gyrus (lpSTG) becomes active about 250 ms before word onset in continuous speech, ahead of motor cortex activity, and encodes whole-word properties — length, syllables, frequency — rather than phonemes or articulation. It is distinct from adjacent auditory-feedback responses, and many of the same neurons respond to both speaking and hearing the word.
• Causality: stimulation just before speech produced errors resembling Wernicke’s aphasia; resections that damaged the subzone predicted post-operative speech deficits. Both are acute, intraoperative evidence.
• The BCI claim: in volunteers with speech paralysis from brainstem strokes, lpSTG activity decoded both intended and perceived speech. The UCSF release describes one participant; the paper and news accounts report two. Chronic stability and accuracy have not been demonstrated.
• Reported discrepancies: the direction of the frequency effect is described differently by press accounts (Scientific American: more active for uncommon words; National Geographic: stronger for more familiar ones). The paper’s abstract states that the region codes word length and frequency. The subzone’s size is reported as ranging from a bean to a nickel, varying between individuals.
• Still open: the role of Broca’s area; whether the same blueprint applies to reading, writing and typing; long-term implant viability; and whether the subzone explains why some people with damage in this area never develop aphasia.
• Sources: Science (DOI 10.1126/science.aef8963 and 10.1126/science.aem7976); UCSF release (19 Nov 2026); Live Science; Science News; Scientific American; National Geographic.

[심층 리포트] 말보다 0.25초 먼저 도착하는 단어 — 베르니케 영역은 ‘듣는 곳’이 아니라 ‘만드는 곳’이었다

1874년 독일 브레슬라우. 신경해부학자 카를 베르니케는 이상하게 닮은 환자들을 마주 앉혔다. 모두 왼쪽 귀 뒤쪽, 같은 띠 모양의 뇌 조직이 다쳤다. 청력은 멀쩡했다. 말도 했다. 문장은 유창했고 속도도 정상이었으며 억양도 살아 있었다. 문제는 내용이었다. 신발을 말하려던 입에서 ‘발’이 나왔고, 다리를 뜻했던 입에서 ‘못(peg)’이 튀어나왔다.

Korean Version: [심층 리포트] 말보다 0.25초 먼저 도착하는 단어 — 베르니케 영역은 ‘듣는 곳’이 아니라 ‘만드는 곳’이었다

Socko/Ghost

2026년 11월 19일 목요일

Nine Teeth in a Porcupine's Den: The Rainforest Was Never Ours Alone [열대우림 호저 굴에서 나온 이빨 아홉 개]

The official biography of Homo erectus is a story about open ground. For most of a century the species has been drawn striding across savannah and light woodland — a lean, long-legged walker with a handaxe, leaving Africa about 1.9 million years ago, reaching the Caucasus and northern China a couple of hundred thousand years later, and holding out in Java until roughly 108,000 years ago.

When the rainforest closed in, the story went, Homo erectus was finished: dense tropical forest was a habitat only Homo sapiens ever learned to crack. Two papers published in Nature on 18 November 2026 take that sentence apart.

Eight teeth from two caves in southern China, their authors argue, belonged to Homo erectus living under a closed canopy between 324,000 and 184,000 years ago — 100,000 to 200,000 years more recently than the species was thought to have survived anywhere on the Chinese mainland. And for much of that time, the chamber that held them was a porcupine's den.

Homo erectus cranium from Hexian, China
A cast of the Homo erectus cranium from Hexian, Anhui — one of the Chinese specimens whose protein signature the Guangxi teeth were compared against. Photo: Jonathan Chen / Wikimedia Commons, CC BY 4.0

Baxian and Naxian are sister caves in the karst country of Chongzuo, in the Guangxi Zhuang Autonomous Region — 15 kilometres apart on the same limestone corridor above the meandering Zuojiang River, 13 and 20 kilometres north-west and north of the city, their entrances five to seven metres above the valley floor.

Both end in small rear chambers packed with the same cemented, cataclastic breccia. Together the two deposits have yielded more than 18,000 mammalian fossils; the excavation team was originally hunting something else entirely, the three-metre ape Gigantopithecus. What the breccia also held was nine hominin teeth.

Dating them took five independent clocks — infrared-stimulated luminescence on feldspar and optically stimulated luminescence on quartz from the sediments, uranium-series and coupled uranium-series/electron-spin-resonance dating of teeth, and uranium-series dating of the flowstone that caps the fossil layer. Forty-five radiometric ages came back; forty-three were run through two Bayesian models. The result: Baxian between 324,000 and 184,000 years ago, Naxian between 319,000 and 241,000 — statistically indistinguishable from each other, and a 95 per cent probability that Homo erectus was on this landscape somewhere inside the Baxian window. Until now, no securely dated Homo erectus fossil in mainland China was younger than about 400,000 years.

Identifying the teeth was a separate argument, and it was won on shape before it was won on molecules. Scanned by micro-CT, the buried surface beneath the enamel cap — the enamel–dentine junction — carries a wrinkled, branching pattern that is a signature of Chinese Middle Pleistocene Homo erectus, and the Guangxi teeth show it, in simplified form, alongside short square upper molars, asymmetric lower premolars with a protruding distolingual corner, and a shelf-like protostylid.

Geometric morphometrics put seven of the nine unambiguously inside the Homo erectus range and statistically apart from Neanderthals, Denisovans and modern humans. An eighth, the lower premolar NX-A2-56, fell outside every comparison group as a statistical outlier, but kept the plesiomorphic talonid and the vertical wrinkling of the enamel–dentine junction that mark it as erectus-like rather than something new. Eight, then, of nine.

The ninth argument is the one that will be quoted for years, because it did not exist as a method a decade ago. Enamel is the hardest tissue a human body makes, and it is stuffed with proteins — amelogenin, ameloblastin, enamelin — that survive in cold and even in warm caves long after DNA has dissolved. Five of the teeth gave up original enamel proteomes, with 556 to 808 amino-acid positions read per tooth.

At position 253 of ameloblastin, four of them carry glycine where almost every other hominin carries alanine, the residue of a single codon change on chromosome 4. That substitution is absent from the Neanderthal and Denisovan genomes, from the great apes, from the 1.77-million-year-old Homo erectus of Dmanisi, and from all 1,461,678 living people in the gnomAD database. It has turned up in exactly one other place on Earth: six Chinese Homo erectus teeth around 400,000 years old from Zhoukoudian, Sunjiadong and Hexian, published earlier in 2026.

But at a second position, 273, the Guangxi teeth carry the ancestral methionine while those six carry the derived valine — enough, the authors argue, to place them on their own branch: a population of Homo erectus that no one had ever sampled molecularly. The same extracts also sexed the owners. Four teeth carried amelogenin-Y peptides and belonged to males; the fifth, a premolar, did not, and belonged to a female.

Karst landscape in Guangxi, southern China
Tower karst in Guangxi, the landscape of steep limestone hills and hidden caves that runs from Chongzuo to the Vietnamese border. Photo: McKay Savage / Wikimedia Commons, CC BY 2.0

Then there is the tooth that does not fit. PA1796, an upper second molar from Baxian, has a short, rhomboidal enamel–dentine junction sitting between Homo erectus and Neanderthals, taller dentine horns, simpler trigon and talon basins, a more developed and lingually expanded talon — and none of the Homo erectus protein marker. It is not a modern human, not a Neanderthal, and not comfortably an erectus.

The obvious candidate is a Denisovan: Homo longi, the lineage known from the Altai, the Tibetan Plateau, Harbin, the Penghu channel and a cave in Laos, and already present in East Asia in this period. The authors say so — carefully. The evidence is circumstantial, and they know it.

"Every label we put on that ninth tooth is a little bit of speculation," the molecular palaeoanthropologist Frido Welker of the University of Copenhagen told Nature. If it is a Denisovan, the valley at Chongzuo held at least two human lineages at the end of the Middle Pleistocene, not long before the first modern humans appear in the region.

The rainforest claim rests, in the end, less on the hominins than on everything around them. The breccia is dominated by forest animals: orangutans, Asian elephants, the rhinoceros Dicerorhinus, tapirs, muntjac deer, leopards, and the extinct Stegodon and Megalovis — the classic Stegodon–Ailuropoda fauna of southern China, the assemblage in which Homo erectus had never before been found.

Carbon isotopes turn that list into a measurement. Plants under a dense canopy are poor in carbon-13, and the animals that eat them carry the signal into their enamel; converted to plant-equivalent values, the Baxian fauna averages −27.6 per mille and the Naxian fauna −29.4, against a dense-canopy threshold of −27.5. Neither cave shows any trace of C4 grasses.

The frequency of canopy-level values is higher than in equatorial African rainforest and comparable to western Amazonia. Clumped-isotope thermometry on the cave sediments points to humid tropical conditions, wetter and cooler than the region today.

The hominins' own teeth are thinner evidence: only three Homo erectus teeth, all from Naxian, were measured for carbon, and they fall at −15.0 to −14.0 per mille — right on the boundary between closed canopy and open C3 vegetation, exactly where the much larger faunal sample sits, but only three data points.

Orangutan
Orangutans lived in the Chongzuo caves' catchment, and their teeth lie in the same breccia as the hominins'. Photo: Ridwan0810 / Wikimedia Commons, CC BY-SA 4.0

What the teeth say about movement is the quiet surprise of the study. Strontium isotopes in enamel are fixed by the geology of whatever a child ate and drank, and in the Baxian premolar PA1799 the ratio ranges from 0.7110 to 0.7220 — a spread that cannot come from one hillside. Matched against a strontium isoscape built for the region with machine learning, the highest values point outward from the cave corridor, potentially as far as the lower Red River valley of northern Vietnam.

The two caves, 15 kilometres apart with overlapping dates, look like one population roaming a forested landscape rather than a band pinned to a single valley. All of that is childhood only, frozen in the two to six years when those teeth formed; where these people went as adults, the enamel cannot say.

And the study is careful in a way its headlines were not: one paragraph calls the evidence for rainforest use unequivocal, the next describes Homo erectus as "at minimum, capable of persisting among rainforest fauna." Both are in the paper. The second is the one the data supports.

How they found food in a closed canopy, what plants and animals they took, what tools they carried into the forest — none of that is in nine teeth. What is newly on the table is larger: if Homo erectus could live in the rainforest, then the spread of rainforest was probably not what killed it in Java, and the species' long, slow ending needs a different explanation. Nine teeth from a porcupine's den have reopened a question palaeoanthropology thought it had closed.

📊 KEY FACT-CHECK — what these papers do and do not establish

Established: eight of nine teeth are Homo erectus by dental morphology, geometric morphometrics and enamel proteomics; the deposits date to 324,000–184,000 years ago (Baxian) and 319,000–241,000 (Naxian); this is the youngest securely dated Homo erectus in mainland China; the associated fauna is closed-canopy rainforest; the four proteomically identified individuals carry a previously unsampled Homo erectus enamel variant.

Not established: the species of the ninth tooth, PA1796 (not erectus, not Neanderthal, not modern human; a Denisovan attribution is suggested on circumstantial grounds only); whether Homo erectus lived in the rainforest as opposed to using it — the hominin isotope sample is three teeth; and how far these people ranged as adults, since strontium records childhood alone.

[학술 탐사] 중국 열대우림 호저의 굴에서 나온 이빨 아홉 개

우리가 아는 호모 에렉투스는 늘 '열린 땅'의 인간이었다. 190만 년 전 아프리카를 나와 캅카스와 중국 북부를 거쳐 자바까지 내려갔고, 기다란 다리로 사바나와 성긴 숲을 걸으며 주먹도끼를 들었고, 불을 쓴 최초의 인류로도 일컬어진다. 그리고 숲이 무성해지면 끝이었다—빽빽한 열대우림은 오직 호모 사피엔스만이 뚫어낸 서식지라고, 반세기 가까이 그렇게 가르쳤다. 2026년 11월 18일 《네이처》에 실린 논문 두 편은 그 문장을 정면으로 뒤집는다.

Korean Version: [학술 탐사] 중국 열대우림 호저의 굴에서 나온 이빨 아홉 개

Socko/Ghost

A King With No Body, a Grave With No Bones: The 'Sand Body' Found Beside Sutton Hoo's Royal Mounds — and the Second Person Beneath It [서턴 후 왕릉 옆 묻힌 두 사람]

In the final week of a four-week excavation in the summer of 2025, the Time Team diggers working in Garden Field at Sutton Hoo uncovered a g...

가장 최신 글