DNA From a 14,000-Year-Old Cave Skeleton in Alaska Is Rewriting Human History

[00:00:00]digging around it, brushing away the mud, pulled it out, and there was this giant stone head gazing back at him.

[00:00:06]>> A 14,000-year-old skeleton pulled out of a windswept Alaskan cave and forgotten in a Danish museum drawer for 60 years, is doing something nobody thought was possible. It is rewriting human history, not adjusting it, not fine-tuning it, rewriting it. The DNA inside those bones does not match the story you were taught in school about how the Americas were people. It does not match the story most archaeologists were teaching 10 years ago. It matches a population nobody knew existed. Walking a continent nobody can visit on a timeline nobody is ready for.

[00:00:41]And what those bones are saying is only the beginning. Opening the cave. The recent discovery of the caves in the car lands on Prince of Wales Island suggests that one of the most important geological resources in the world might be right here in our backyard in southeast Alaska.

[00:00:55]>> Picture the place for a second because the place matters. The Seward Peninsula points west out of Alaska like a finger reaching for Russia. And on its windswept edge, a series of limestone caves cut into the hills above a dry tributary of the Immaruk Basin. The locals call them Trail Creek Caves. From the mouth of the largest cave you can see for miles across treeless tundra, the same sweep of land mammoths and bison crossed when the ice age was ending and the world was just beginning to warm. That is where Larson was working in the late 1940s. He and his crew dug through layer after layer of cave fill. And what they pulled out was extraordinary stone tools. Ivory implements carved from mammoth and walrus tusk. The bones of mammoth, bison, and caribou. And among the animal remains, fragments of human bone, pieces of skull, pieces of jaw, teeth that had once chewed food in a mouth that lived and breathed at the end of the last ice age. Scientists have been listening to the story being told by a man who was born thousand years ago.

[00:01:59]>> Here's the thing. Larson had no idea what he had. Nobody in 1949 did. The science needed to read what was inside those bones did not exist yet. So Larson did what archaeologists of his generation did. He packed the crates. He shipped them home and the Trail Creek bones disappeared into the back of a museum collection in Copenhagen and went to [music] sleep for 60 years. In 2010, under the Native American Graves Protection and Repatriation Act, those bones came home. They were returned to Alaska, and they arrived at exactly the moment when the science needed to read them had finally caught up with what they contained. Carbon dating placed the oldest fragments at roughly 14,000 years before present. The shape of the bones was studied. And then as ancient DNA recovery techniques [music] exploded in capability through the 2010s, the genetic material preserved inside those teeth and jaw fragments became readable for the first time in the 14,000 years they had existed. Researchers extracted DNA. They sequenced genomes. They compared the results to every other ancient American genome on file. And what came back was nothing like what the textbook said, the model that broke.

[00:03:12]Before we get to what those Trail Creek bones actually said, you need to understand what they were about to contradict. Stay with this because the official story dominated American archaeology for nearly a century. And the people who built it did not give it up easily. The story went like this.

[00:03:28]Around 13,000 years ago, a population of big game hunters from Northeast Asia walked across the exposed continental shelf of the Bearing Land Bridge into Alaska. From there they moved south through an ice-free corridor that opened between the Cordilleran and Laurentide ice sheets. As the glaciers retreated, they spread across two continents in a few thousand years. They became the ancestors of every indigenous American population from the Arctic to Tiara del Fuego. One migration, one source, one people, one elegant chapter in a textbook. It had a name, Clovis first.

[00:04:07][music] There has been a narrative that archaeology has sought to impose upon us. And this was this was called the Clovis first idea that there was a people who archaeologists called the Clovis people. We don't know what they called themselves uh in North America named after a New Mexico town where in 1932 archaeologists found beautifully fluted stone projectile points lodged among mammoth bones. The points were unique to the Americas. Nothing like them existed in the old world. When radiocarbon dating became available in the 1950s, the Clovis culture was placed at roughly 13,000 years ago. And for the next half century, those fluted points defined the floor of American prehistory. The people who made them were the first Americans. Anyone who said otherwise was wrong. And let that land for a second. The field did not just hold this view. It enforced it.

[00:05:00]Sites that appeared to predate Clovis were challenged on methodological grounds. Researchers who argued for earlier human presence were marginalized. Careers were damaged.

[00:05:10]Funding dried up. Entire dissertations were essentially blacklisted because their dates did not fit. The Clovis first model was not just a hypothesis.

[00:05:20]It was orthodoxy. And orthodoxy in any science [music] is a dangerous thing. It shapes what people are willing to look for, what they are willing to admit they have found, and what they are willing to risk their careers defending. Then the evidence began to leak through. Anyway, in southern Chile, a site called Monte Verde returned firm dates of 14,500 years ago, more than a thousand years before Clovis. The site preserved wooden structures, plant remains, and stone tools in a pete bog that had protected them from decay. The dates were challenged. They were defended.

[00:05:55]Additional excavations confirmed them.

[00:05:57]Monte Verde held. Then Meadowraftoft Rock Shelter in Pennsylvania returned dates potentially as far back as 16,000 years. The Buttermilk Creek complex at the Freriedken site in Texas yielded stone tools beneath the Clovis layer at 15,500 years. [music] And within a decade, we discover the Clovis site, which is out on the Texas New Mexico border. And the Clovis site is of these elephant hunters who are actually even older than the Folsome people. Each site survived intense scrutiny. The pre-Clovis evidence was real, and the wall the Clovis first defenders had built around their model started to crack. And the bones from a Danish museum drawer freshly returned to Alaska were about to push that wall over. Because what came next was not stone tools and not footprints and not a contested date. It was a genome, a complete one. And the people who built the old model were not ready for what it said. Before we get into the genome that finished the old model off, do me a favor, hit subscribe.

[00:07:03]The kind of story we are walking into ghost populations, footprints that should not exist, a continent now drowned under the bearing sea, is exactly what this channel exists to tell. Stay with it. Ancient DNA changes everything. The technology that made any of this possible was refined mostly in Europe. cold climates, favorable burial conditions, genetic material that could survive across tens of thousands of years inside frozen bone. Applying the same techniques in the Americas was harder. Warmer soils, acidic conditions, different burial practices. All of it ate DNA faster. Think about it like this. For years, geneticists working on early American remains were trying to read a book that had been left in the rain for 10,000 years. The pages were warped. The ink was bleeding. Whole chapters were illeible. But the methods improved. The amount of bone needed for a sample shrank. From a full bone to a tooth to a fragment of a tooth to a few milligs of inner earbone, the densest piece of the human skeleton where DNA survives best. By the mid2010s, geneticists were pulling whole genomes out of remains that 20 years earlier would have been considered unreadable.

[00:08:17]The first results lined up with what people expected. The ANZIC child, an infant burial from Montana found alongside Clovis stone tools, was sequenced in 2014. The lab went quiet when the results came back. Clovis people were ancestral to modern Native Americans, and they already represented a population that had branched out from an even earlier founding lineage.

[00:08:40]Kenowick man, a 9,000-year-old skeleton from Washington State.

[00:08:45]>> It is certainly America's most important skeleton. It is an exceptionally rare discovery. It does not happen but once in a lifetime.

[00:08:54]>> Who had been at the center of years of legal battles was sequenced in 2015. The result came back unambiguous. He was closely related to modern Native American populations of the Pacific Northwest. The spirit cave mummy from Nevada around 10,600 years old was sequenced in 2018 confirmed as Native American and returned to the Fallon Pyute Shosonyi tribe. So far the textbook held the deep past of the Americas connected to the people who lived there when Europeans arrived. That part of the story survived. But something else was emerging in the data.

[00:09:30]Something that did not fit. The ancient Americans were not one people spread evenly across two continents. They were multiple groups, genetically distinct branches that had split from each other thousands of years before the Clovis horizon. Different roots, different regions, different signatures left [music] in their descendants. The single migration story was breaking and the next genome would shatter it completely.

[00:09:54]The ancient Bingians. Hold the Trail Creek bones in your mind for a second because what was about to happen in another Alaskan cave was the moment the genetic record finally caught up with the bones in those Danish drawers. And the result was something nobody had predicted. In 2013, an archaeologist named Ben Potter from the University of Alaska Fairbanks was excavating a site called Upward Sun River in the Tanana River Valley of Interior Alaska. The site dated to about 11,500 years ago, post Clovis. But among the oldest human occupations anywhere in Alaska, inside a residential pit beneath a layer of burned material, Potter's team found something that stopped them cold. The remains of two infants undisturbed in pristine archaeological context. Potter sent the bones to a team led by SK Willers at the University of Copenhagen.

[00:10:50]Willer Sleb's lab is one of the best ancient DNA operations on Earth. They had pulled genomes from places nobody thought possible. And they got a complete genome out of one of the Upward Sun River infants.

[00:11:03]>> Really exciting. I mean, now it's a matter of breaking, you know, the genetic code and try to reconstruct this individual.

[00:11:09]>> We could see that because we start getting Y chromosome sequences coming up.

[00:11:14]>> The bones were given a designation USR1.

[00:11:17]Picture the lab the day the result came in. Picture the screen. The genome loaded against the reference database and the first comparison started running and nothing matched. Read that again.

[00:11:28]Nothing matched. The USR1 genome did not line up with any known Native American population. Not the Clovis ancestors.

[00:11:36]Not modern indigenous groups in North America. Not modern groups in South America. It was something else. A separate lineage that had split from the ancestors of every other Native American. roughly 20,000 years ago and then it had stayed in Bingia isolated for thousands of years afterward. When the paper hit nature in 2018, the field reacted. The researchers gave the population a name, the ancient Bingians.

[00:12:02]It's clear they had a common ancestor, but according to the team, the baby's genetic code was so different that it's part of a whole separate lineage, a third group that they're calling the ancient Bingians that supports what's known as the Bingian standstill model.

[00:12:15]These were not the ancestors of most modern Native Americans. They were a parallel branch, a population that had broken off from the broader Native American founding lineage. While both groups were still in Bingia before the migration south even began, and then they had stayed, they had lived in Alaska. They had hunted on the same tundra, walked the same valleys, watched the same northern lights wheeling above them in winter for thousands of years.

[00:12:41]And eventually, for reasons the genetic record cannot fully explain, they had vanished, leaving little if any trace in the populations that survived. An entire branch of the human family gone. So quietly that until 2018, nobody knew they had ever existed. The peopleing of the Americas was not a single wave. At minimum, it was two. The ancient Bingians who stayed in the north and the ancestors of everyone else who pushed south. And what was hiding in the Trail Creek bones was about to make the picture even stranger. What Trail Creek confirmed. Now we come back to those bones from the Danish museum drawer, the ones Larsson had pulled out of the limestone in 1949, the ones that had waited 60 years to speak. When researchers compared the Trail Creek genomes to USR1 and to the rest of the ancient American genetic record, the picture sharpened. The individuals buried in those caves on the Seward Peninsula belong to the same deep population structure that the upward sun river genome had revealed. They were part of the genetic mosaic of late ice age bingia. A mosaic, the single migration model had never predicted and could not explain. They were not outliers. They were the increasingly common pattern. Stay with this. Every additional bingian genome that has come out of an Alaskan cave or a frozen lake bed in the years since has slotted into the same picture. Not one founding population. A network related but genetically distinct groups sharing a landscape, occasionally mixing, occasionally splitting off, occasionally dying out without descendants. The Trail Creek data also helped confirm an idea that had been floating for years on theoretical grounds, but was now getting hard genetic backup. It is called the Bingian standstill. Here's the idea simply. The ancestors of Native Americans did not cross from Siberia to the Americas. In one fast push, they paused. For thousands of years, the ancestral population lived in Bingia itself, the exposed continental shelf that connected Asia and North America when sea levels were hundreds of feet lower than today.

[00:14:49]>> Americans were were a seafaring people and all that about what glaciers are melted and not melted and when this and that corridor and land bridges open was a moot point because these were people that um just came down the coast.

[00:15:02]John Hoffer at the University of Colorado was one of the researchers pushing this model hardest and the genetic evidence was now backing him up.

[00:15:11]Bingia was not just a land bridge. It was not a hallway you walked through. It was a refugeium, a habitable region of shrubby tundra, river valleys, and pockets of woody vegetation. Sheltered enough to support large mammals and the humans who hunted them. While massive ice sheets covered most of Canada and blocked [music] any southern movement, it was in effect a homeland, a lost continent of grass and bone we can no longer visit because [music] the sea has eaten it. The genetic evidence now pointed to an isolation that lasted as long as 8,000 years. from roughly 25,000 years ago when the founding population diverged from their Asian ancestors [music] until somewhere between 17 and 15,000 years ago when they finally began to move south as the ice retreated 8,000 years in one place going their own way.

[00:16:02]Long enough for entire branches of the family to split off. Long enough for the ancient Bingians to become a separate people. long enough for the bones in those Danish drawers to belong to a world that until very recently did not officially exist. And what came out of the Amazon a few years later should not have existed at all. The ghost in the Amazon. Hold the Trail Creek bones in your head one more time because what came next made even the Bingian story looks simple. In 2015, two of the most influential figures in ancient DNA research, Pontis Scogland and David Reich, were running comparisons across indigenous populations of the Americas.

[00:16:41]And in a small handful of South American groups, particularly in the Amazon basin, they kept hitting a signal that did not belong. Picture the moment the screen lights up. The match is small, statistically faint, but it keeps coming back. And the closest comparison the algorithm can find is not from anywhere in the Americas. It is not from anywhere in Asia. It is from Australasia, from the indigenous peoples of Australia, New Guinea, and the Andaman Islands.

[00:17:09]>> The superighways that came out of some of the models actually seem to match up a lot of the old stock routes and the Aboriginal trade lines that um that we know from say the 19th century.

[00:17:20]>> Read that again. a trace of Australasian ancestry in the Amazon. The signal was tiny. It only showed up in certain South American populations and was absent everywhere else in the New World, but it kept coming back. The researchers gave the source a name, population Y from the Tupi word for ancestor, and they had no good explanation for how it had gotten there. Australasians had never been proposed as contributors to the peopleing of the Americas. The distances were enormous. The migration routes were not obvious. The timing made no sense in any model anyone had built. But the data did not care. The most likely explanation is a ghost. A population that contributed ancestry to some early Americans and then went extinct without leaving any unmixed descendants. Maybe they came across separately. Maybe they were absorbed into a later wave. Maybe they were displaced or eliminated entirely by populations that arrived after them. Whatever happened, their genetic legacy survives only as a faint trace, detectable in a few Amazonian groups [music] and invisible everywhere else on Earth. They were here. They lived. They had children. Some of those children's children's children are still walking around the Amazon today, carrying a small piece of them. And we have no idea who they were, what language they spoke, what they looked like, how they got here. The peopleing of the Americas was not one migration.

[00:18:45]It was not even two. It was a process. A long layered multissource process that unfolded over thousands [music] of years and involved populations we are only now starting to recognize. And then in 2021, the timeline itself broke. White sands.

[00:19:03]Hold this one tight because this is the finding that pushed the whole story past the point of return. A team led by Matthew Bennett published evidence of human footprints preserved in the lakeshore sediments of White Sands, New Mexico.

[00:19:16]>> I think [music] uh is the age of the footprints. They're much older than anything that has been found uh in in North America previously. And so that really changes the way that we think of uh how the Americas were were people.

[00:19:28]Not stone tools [music] that could be argued over. Not animal bones that could be dismissed as natural deposits, footprints, the unmistakable impressions of bare human feet pressed into mud that hardened, was buried, and was eventually exposed by paleontologists.

[00:19:44]And the dates came back at 21 to 23,000 years ago. Let that land 21,000 years ago at the absolute height of the last ice age. The dating was challenged hard.

[00:19:56]The original method relied on radioarbon ages from aquatic plant seeds which can absorb older carbon from groundwater and produce ages that are too old. The Clovis first defenders pounced. Then in 2023, a follow-up study by Pigatti and colleagues confirmed the original dates using completely independent methods.

[00:20:17]Pollen, quartz grains, multiple lines, multiple labs. The footprints were real.

[00:20:23]The dates were real. Humans had been walking around the interior of New Mexico while the Great Ice Sheets still covered most of Canada. Think about what that does to the map. 21,000 years ago, the ice-free corridor was not yet open.

[00:20:37]The traditional inland route into the Americas did not exist. The coastal kelp highway might have been usable, but the timing pushes the very edge of what the coastal hypothesis can support. And the genetic record does not yet clearly account for a human population in the interior of North America that early.

[00:20:55]So, who were the White Sands people?

[00:20:57]Were they distant relatives of the ancient Bingians? Were they an even earlier and unknown population that left no genetic descendants? Were they the source of the mysterious population Y signal showing up in the Amazon thousands of years later? Right now, nobody can say for certain. The questions are multiplying faster than the answers. And the bones in the Danish museum drawer were always part of this story. [music] They just had to wait 60 years for the world to catch up. The stories we were told. Jennifer Raph, an anthropologist at the University of Kansas, has spent years pulling all of this together. Her book, Origin: A Genetic History of the Americas, is the closest thing the field has to a current map of where the evidence actually stands. And her assessment is sobering.

[00:21:43]>> DNA can tell you some things. It can tell you estimate when events may have happened, like gene flow between two populations, uh, migrations, genetic bottlenecks, and you can kind of tie those to events in the archaeological record.

[00:21:57]>> The simple stories are gone. The textbook diagrams with single neat arrows pointing from Siberia to Alaska to the rest of the hemisphere are obsolete. What we have now is a research frontier. A field where new genomes arrive faster than consensus can form.

[00:22:14]Where each ancient skeleton answers some questions and opens new ones. Where the official narrative is being rewritten in real time. The Trail Creek bones are part of that rewriting. They are not the oldest findings of the past decade. They are not the most famous, but they are representative. They show what Bingia genuinely looked like at the end of the ice age, a region structured into multiple lineages, some of which would become modern Native Americans, some of which would stay in Alaska, some of which would simply vanish. They show in genetic black and white that the founding of the Americas was not a moment. It was a process, one that unfolded over thousands of years, involved multiple distinct populations, and left signatures in living people we are only now learning how to read. And here's the part that should keep you up at night. Trail Creek waited 60 years in a Danish drawer to be read. How many other museum drawers in how many other countries are sitting on bones that have never been sequenced? How many skeletons excavated in the 19th and 20th centuries before ancient DNA was even a concept are sitting in storage right now containing genomes that could resolve outstanding questions about who we are and where we came from. Trail Creek was one set of bones. There are thousands more. The honest answer to how many other collections hold similar data is we do not know. We are only beginning to ask.

[00:23:40]The Bearing Landbridge National Preserve, which protects the Trail Creek Caves region encompasses some of the most archaeologically significant terrain in North America.

[00:23:49]>> Ever since this place was first discovered over 12,000 years ago by people crossing the Landbridge in search for food and shelter, the Taurus had always been a place where inspiration and community can be found.

[00:24:00]>> But here's the cruel part. Most of the original landbridge is gone. It is submerged under the Bearing Straight.

[00:24:07]Whatever campsites, kill sites and burials existed across thousands of years of Bingian human life on the exposed continental shelf are now hundreds of feet underwater beyond [music] the reach of conventional archaeology. The coastal migration route, if it was used, left most of its evidence on shorelines that are now drowned. generations of human life, tools, fire pits, children's bones, stories, all of it sealed under cold, dark water at the bottom of a sea that did not exist when those people were alive. Stay with this. The deep history of the peopleing of the Americas is [music] to a real extent lost to the sea. What remains are the margins sites like Trail Creek Caves perched above the ancient coastline where the evidence of Bingian occupation survived because the water never reached that high.

[00:24:58]Researchers like Dennis Oor, Connie Mulligan, and Jennifer Raph keep extracting DNA from Bingian and early American remains. Each new genome adds another data point. Each new comparison sharpens or complicates [music] the picture. The ancient Bingians were real.

[00:25:15]The pre-Clovis populations were real.

[00:25:17]The ghost population whose trace shows up in the Amazon was real. The footprints at White Sands [music] were real.

[00:25:24]>> Those are the things that are really um exciting here because most footprint sites around the world are relatively small. They're constrained by a small piece of exposure, but here the footprints expand over well tens of kilometers and that is unique. And the bones from the Trail Creek caves, the ones Larson pulled out of a windswept Alaskan limestone cave in 1949, were not a footnote. They were a witness. They were always part of this story. They were just waiting for someone to know how to listen. The Clovis first model is dead. The single migration story is dead. The ice-free corridor as the one true route is dead. What is taking their place is not a clean replacement theory.

[00:26:03][music] It is the data itself. Somewhere right now, in a museum drawer in Copenhagen, in a laboratory freezer in Cambridge, in an archaeological archive nobody has thought about in 50 years, the next set of Trail Creek bones is sitting, already excavated, already cataloged, already waiting. The genome is intact. The story [music] is intact.

[00:26:25]All it needs is for someone to pull the drawer open and run the sample. The Trail Creek bones waited 60 years to speak. Some of those other bones have been waiting longer. So, here is the question I want to leave you with. If a single drawer in a single Danish museum was hiding a finding this large, what is sitting in the drawer next to it? Drop your thoughts in the comments. Do you think the next big rewrite comes from another cave? From underwater archaeology along the drowned Bingian coast, or from a museum collection nobody has touched in half a century? I read every comment. I want to know where you think this goes next. Subscribe so you don't miss the next one because the next drawer is already open in someone's lab. And when the next genome lands, we'll be the ones telling you what it actually says. The bones are home. The DNA has been read. And the story they are telling is one we were never meant to