Ancient Mysteries - The Most Unsolved Mysteries We Still Can’t Explain

Episode Date: October 7, 2026

Some mysteries survive for centuries because the evidence is incomplete. Others become even stranger the more we discover.In this documentary, we explore some of the world’s most fascinating unsolved mysteries — ancient underground cities, enormous structures, mysterious archaeological sites, unexplained engineering, lost knowledge, strange discoveries, and places that continue to challenge researchers.From vast chambers hidden beneath the ground to prehistoric monuments built thousands of years earlier than expected, these stories raise questions about what ancient people knew, how they accomplished extraordinary feats, and how much of the past may still be waiting to be discovered.Some mysteries have several possible explanations. Others remain the subject of serious debate. And in a few cases, we may simply be missing the evidence needed to understand what really happened.We have discovered the sites. We have studied the evidence. But we still don’t have all the answers.💬 Which mystery do you think is the hardest to explain?🔔 Subscribe for more documentaries about unsolved mysteries, archaeology, ancient civilizations, mysterious discoveries, lost history, and unexplained places.

Visit BetMGM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and GameSense remind you to play responsibly. 19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact Connects Ontario at 1-866-531-2,600 to speak to an advisor,

free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. Hey there, History Hunters. In 1963, a man in the Turkish village of Derinkuyu knocked down a wall in his own house. Behind it was a tunnel. That tunnel dropped 85 metres straight down into an underground city with 18 levels, built to hide 20,000 people. It had been sitting under his living room the whole time.

Worst renovation surprise in history, that is the pattern for today. The biggest secrets on this planet get found by farmers, divers, and guys who just wanted a bigger kitchen. Ten places, three levels of impossible, things that could not be built with the tools of their time, things that could not have arrived where we found them, and writing that nobody can read. And one question the whole way through, how many more are under our feet right now, waiting for someone to hit the right wall, hit that like button, and drop a comment telling me what city you're watching from. Let's go. So let's go back to that wall in Capadocia, because what happened next is even better than the moment of discovery. The man did

what any of us would do. He got a flashlight, or more likely an oil lamp, and he went in. And the tunnel did not stop. It went down past a room, and then another room, and then a staircase carved out of solid rock, and then a corridor so low he had to bend double, and then it opened into a space the size of a church, which, as it turns out, is because it was a church. What that homeowner had accidentally opened was the top floor of the largest excavated underground city on the planet. Not a bunker, not a hideout. A city, with all the boring municipal infrastructure that word implies. Stables, storerooms, kitchens with soot still baked into the ceilings, wine and oil presses because apparently even Doomsday requires a beverage program.

Wells, chapels, a school, communal rooms that could hold hundreds of people at once. And the whole thing punched straight down into the earth like someone had dropped a skyscraper into the ground upside down and then walked away without telling anyone. To understand how this was even physically possible, you have to understand that Cappadocia is not made of normal ground. Millions of years ago, the volcanoes in this part of central Anatolia, including the enormous Erciers and Hassan Dag, spent an extremely long time redecorating the region with ash. That ash settled, compacted, and turned into a rock called Tuff, and Tuff is the single most user-friendly building material nature has ever produced. It is soft enough to carve with hand tools made of iron, bronze, or in a

pinch a very determined stick, and here is the magic part. When you expose a fresh surface of it to air, it hardens. So you dig your room out of butter, and then the walls turn to stone on their own schedule. It is basically the geological equivalent of a self-assembling piece of furniture that actually works, which puts it several thousand years ahead of anything currently sold in a flat cardboard box. That is why the region is full of holes. Locals had been carving into cliffs and hillsides here for as long as anyone can remember, making storage cellars, animal pens and cool rooms, for keeping food from spoiling in a climate that swings between roasting summers and brutal winters.

Except at some point in Derinkuyu, somebody looked at their perfectly nice storage cellar and thought, What if we did not stop? And then a whole lot of people spent a whole lot of generations not stopping. The complex goes down roughly 85. meters. That is a 28-story building of nothing, carved out and hauled away by hand. Estimates put the number of levels around 18, though the honest answer is that levels blur together because this was not built to a master plan by an architect with a drafting table. It grew. Different centuries added different floors for different reasons, like a house that gets extended

every time a new generation moves in, except the extensions go downward and the reason is usually that an army is coming. The population figure people people love to throw around is 20,000, and that number deserves a little unpacking, because the mental image most people form is wrong. This was not 20,000 people living underground permanently, raising kids in the dark, developing a pale, subterranean civilization with its own folk music. Nobody was moving in. This was a shelter, designed for weeks or months at a stretch, and the design tells you exactly that. The upper levels held the stables, which sounds backwards until you spend three seconds thinking about what a few thousand animals produce and where the smell would go if you

put them at the bottom. Livestock stayed near the surface. Easy to bring in fast, easy to ventilate, and mercifully far from where anyone was sleeping. Below that came the living quarters and kitchens. Deeper still came storage, wine presses, wells, and the chapels. The deeper you go, the more permanent and the more precious the function, which brings us to the part that separates Derinkuyu from a really ambitious basement, the engineering, and the ventilation system is where you stop laughing and start being impressed. There are more than 50 ventilation shafts drilled vertically through the rock, some of them running 55 metres straight down, and the network connected to them is estimated in the hundreds of channels. These shafts did double duty. They pulled fresh air down into the lowest

levels, and some of them also functioned as wells. Here is the clever bit, and it is genuinely clever. Certain wells were deliberately not connected to the surface. They tapped into the water table from below, with no opening at ground level at all. Why would you do that? Because the first thing any competent besieging army does when it finds out your hiding underground is poison your water. If the enemy cannot find the well, the enemy cannot poison the well. The people who designed this had clearly thought about their enemies in detail the way you think about someone who has already ruined your life once. And that theme repeats everywhere down there. The round. The round The ground stone doors are the famous example, and they are famous for good reason.

Picture a giant stone wheel, up to about a metre and a half across, half a meter thick, weighing somewhere in the range of a couple hundred kilograms, and in some cases pushing toward half a ton. It sits in a carved recess beside a narrow passage. From the inside, a couple of people can roll it across the corridor and seal the passage completely. From the outside, you have got nothing, no handle, no leverage, no grip. Nothing to push against but a flat disk of rock wedged into a tunnel that is deliberately too narrow to let more than one attacker approach at a time. Naturally, several of these doors have a hole through the centre, and archaeologists have offered various theories about what that hole was for,

including citing the corridor beyond and, less charmingly, sticking things through it at whoever was on the other side. Either way, the customer service philosophy is clear enough. The door only works one way, and you are on the wrong side of it. Now multiply that. There is not one door, there are doors at multiple points, on multiple levels, so that even if attackers somehow got through the first, they would find themselves in a sealed section of tunnel with another disk of rock in front of them, no light, no air movement, and an unknown number of extremely unfriendly locals somewhere in the dark below. Whoever built this did not just plan for a siege.

They planned for a siege that goes badly, and then a fallback, and then a fallback for the fallback. This is defence in depth in the most literal sense, the phrase has ever had. The passages themselves are part of the weapon. They are low, narrow and awkward on purpose. You cannot walk upright. You cannot swing a sword. You cannot bring in a formation. You have to shuffle forward bent at the waist, one man at a time into darkness, in a maze you do not know, while the people who carved it know every corner. An invading soldier in full kit going down there was not conducting an assault. He was volunteering for the worst afternoon of his life, which raises the obvious question. Who was so terrified, for so long, of what? And here is where the tidy documentary

answer falls apart because the dating is a mess. One school of thought credits the Frigians, who were running things in Anatolia in the 8th and 7th centuries before the common era, and who were genuinely skilled at working rock. The idea is that they cut the first levels, and everyone who came afterward inherited a very good starting point. Supporters of an early date like to point at xenophon, the Greek soldier and writer, who marched an army through Anatolia around 400 before the common era, and described people living in underground dwellings with entrances like the mouth of a well, wide enough at the bottom for livestock, with the humans coming and going by ladder. That is an eyewitness account of the underground

lifestyle in this general region, written down before the Roman Empire existed. It is not proof that he was standing on top of Derengueu, but it certainly means nobody can call the concept a medieval invention. The other school argues that the vast majority of what you can walk through today is Byzantine work, expanded and militarized between roughly the 7th and 10th centuries, when Cappadocia had the extraordinary bad luck of sitting directly on the frontier between the Byzantine Empire and the advancing Arab armies. For a couple of centuries, this region was a corridor that armies used the way commuters use a highway. Raids came through in cycles, sometimes seasonally, and the Christian population of Cappadocia responded by refining the low.

local hobby of digging into an art form. The chapels with their cross-shaped floor plans are clearly Christian. The barrel-vaulted hall on the second level, usually identified as a religious school with side rooms opening off it, is clearly not the work of Bronze Age herders. The truthful answer is almost certainly both, and that is what makes the dating so difficult. When you carve a new room out of rock, you erase the evidence of the old room. There is no foundation layer to excavate, no neat stratigraphy of dropped pottery, because every century of occupants used the same space, swept it out, recut the walls, and left. Radiocarbon needs organic material, and the useful organic material has generally been carried

off, eaten or burned. So you end up with a monument that is simultaneously very old and constantly renovated, which is the archaeological equivalent of asking how old a knife is when the handle has been replaced twice and the blade three times. What is not in dispute is that people people kept using it, right up until modern history. The Cappadocian Greeks who lived in the area were still going underground, during periods of danger well into the 20th century, and they left the region in 1923 during the population exchange between Greece and Turkey, which means the community that actually knew how the place worked, which passages went where, which shafts were what the local names for the rooms were, packed up and left within living memory. The knowledge

did not get destroyed by some ancient catastrophe. It just walked out the door, and 40 years later a man with a hammer had to rediscover it by accident, because nobody who stayed behind knew what was under the neighbourhood. That single fact should genuinely unsettle you a little. This was not lost in the midst of antiquity. This was lost in the 1920s, and Derinkuyu is not alone, which is the detail that turns a curiosity into a pattern. Capadocia has something in the range of 200 underground settlements of various sizes, and the region keeps producing more. In 2013, during a routine urban development project in the town of Nevesahir, contractors demolishing old houses around the Castle Hill hit voids, not small ones. What emerged over the following years appears to be an

underground complex potentially larger than Derring QU, with tunnels, living spaces and rock-cut infrastructure spread across a huge area. It was found the way these things are always found, by people trying to do something completely unrelated, in this case building housing, which had to be immediately and expensively stopped. Somewhere in that story is a municipal planner who opened an email on a Tuesday morning and had his entire year cancelled by the 8th century. But the loose thread that nobody has pulled all the way, the one that keeps people up at night, is the tunnels between cities.

Dair and Kuyu is connected, or at least appears to be connected, to Kay Mackley, another substantial underground city roughly 8 to 9, kilometers away. A tunnel of that length is not a panic exit. You do not dig nine kilometers through rock in case you need to run away quickly. You dig nine kilometers because you want two separate populations to be able to move people, food, messages, and possibly animals between two fortified positions without ever appearing on the surface. That is not a shelter. That is a network, and a network requires coordination, shared planning, shared labor, and shared authority across communities that in most historical

reconstructions are supposed to have been fairly local and fairly independent. Somebody was organising this at a regional level, and we do not know who or under what authority. And then there is the question nobody has a satisfying answer to. Where did the rock go? Do the arithmetic even roughly. 18 levels, dozens of chambers, kilometers of passage, ventilation shafts running the full depth. The volume of material removed from Derinkuyu alone runs to hundreds of thousands of cubic meters. Every basket of it was carried up by hand through those deliberately narrow corridors by people who then had to put it somewhere. There is no mountain of spoil sitting next to the town. Some of it was certainly used as building material and construction fill on the surface. Some of it spread

across fields. Some of it lost to 3,000 years of weather and farming and later construction. But the disposal operation was clearly systematic and clearly sustained, and for a project meant to hide people from invaders, hiding the evidence of the digging is not a footnote. A giant fresh pile of pale rock chips outside your village is a sign that says, look here. Today, visitors can walk through roughly 10% of Derinkyu. The rest is closed, partly for safety, partly because it is unexcavated, and partly because the honest answer is that nobody has mapped all of it. You can go down eight levels and stand in a room where a family waited out a raid 1,200 years ago, and above your head a time, a

12 more levels of rock, and below your feet is a well that was dug so that nobody outside could find it. It is one of the very few places on earth where the phrase we have only scratched the surface is not a figure of speech, but a description of the excavation status. Now, if the mystery of Derring-Q-U is who was so afraid and why, the next one flips the problem completely. Because on the other side of the continent there is a place where the digging is even more precise, the volume of removed rock is even more absurd, and there is no fear, no war, no invaders, and no explanation at all. In fact, there is not even a mention. The village of Shianbeokun sits in Jijang province, in eastern China, near the city of Kuzhou, and for as long as anyone in the village could

remember it had ponds. Deep, still, cold, weirdly dark ponds that locals had always assumed were bottomless. That is one of those phrases that sounds superstitious until you realise that you realize it just means nobody in living memory had ever managed to find the bottom, which for a farming community is functionally the same thing. Visit BetMDM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and Game Sense remind you to play responsibly, 19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact Connects Ontario at 1-866-531-2,600 to speak to an advisor.

of charge. BetMGM operates pursuant to an operating agreement with I Gaming, Ontario. With Export Development Canada, Canadian businesses are competing, leading, and winning in more international markets, like how agri-food businesses are strengthening global food supply, and critical minerals are powering industries of the future. Their ambition can't be contained. And with EDC behind them, neither can their growth. Export Development Canada is the financial partner for Canadian businesses going global. Visit EDC.ca for more. In the summer of 1992, a group of villagers decided to test the theory, motivated not by archaeology but by fish. The plan was to drain one of the ponds, see what was down there, and ideally harvest whatever

had been quietly growing in the dark for a few generations. They pulled their money, rented pumps, and started emptying the pond into the surrounding fields. They pumped for 70,000. days. Take a moment with that. Seventeen days of continuous pumping, water levels dropping, walls appearing, and at some point the strong suspicion that this was no longer about fish. What emerged was not a pond bed. It was a floor. The pond turned out to be the flooded top of a chamber more than 30 metres high, with walls cut straight and clean, a ceiling angled at a consistent pitch, and enormous stone columns left standing in the middle of the room to hold the roof up. The fish harvest, by the way, was catastrophic.

philosophically disappointing. The Great Hall consisted of essentially nothing living, because whatever was in there had not been a fish paradise. They spent 17 days and their savings to catch approximately no fish, and one of the great archaeological puzzles of the modern world, which is either the worst or the best return on investment in the history of Chinese agriculture. Then somebody noticed the other ponds. They drained those two, and the same thing happened, and again. What is now known as the Longue Caves consists of more than 20 separate excavated chambers, with some counts pushing higher as more flooded sections are examined, spread through a hill of siltstone across an area of tens of thousands of square metres. Seven of them are open to visitors. Others remain

full of water, which is its own kind of unfinished business. The scale is hard to convey with numbers alone, so try it this way. Walk into one of the big chambers and the ceiling is higher than a 10-story building. The floor area of a single grotto can exceed a thousand square meters. The support columns are not decorative, they are structural, tapered, positioned with an understanding of load that no accidental process produces. The ceiling sloped downward at a consistent angle, and the floors below them slope at the same angle, meaning the person who cut this maintained a constant ceiling height while descending, over an area the size of a warehouse, in total darkness, with no laser levels and no way to see the far wall. And then there are the marks.

Every surface, and this is the part that gets under people's skin, is covered in the same pattern. Parallel chisel lines, cut in bands, each band roughly 60 centimetres wide, with the lines running at an angle of about 60 degrees. Ceiling, walls, columns, floors. The pattern is consistent within a chamber, and it is consistent between chambers in caves that were separately excavated. The lines are even in depth, and even in spacing, which means a stoneworker either had extraordinary discipline

or was following a specification that everyone else was also following. Some researchers think the marks are purely a byproduct of a particular quarrying technique. Others think they were deliberately finished that way. Either way, somebody enforced a standard across a project of enormous size, which implies supervision, training and a management structure. The stone masons of Longue had a style guide, and they stuck to it better than most modern workplaces stick to anything. Now the arithmetic, which is where the whole thing goes off a cliff, add up the volume of the known chambers, and you get an estimate in the neighbourhood of 1 million cubic

metres of rock removed. One million. For scale, that is comparable in raw excavation to serious imperial infrastructure, the kind of project that requires a state, an army of labourers, a supply chain of food and tools and years of sustained effort. And here is the problem, delivered in four parts. Part one, there is no spoil. A million cubic metres of siltstone has to go somewhere. There is no artificial hill nearby, no fill terrace, no identified quarry dump, no ridge of debris that geologists can point at and say there it is.

The rock left the site and vanished into the landscape without leaving an obvious signature. Part two, there is no camp. A project of this size means thousands of workers over a long period, and thousands of workers means housing, cooking, waste, broken tools, worn out chisels, discarded pottery, burials. The archaeological debris field around a major ancient construction site is usually the easiest thing to find because people are messy. Around Longyu, that expected mess is conspicuously thin. Part three, and this is the genuinely strange one, there is no record. China has one of the deepest, most obsessive documentary traditions on the planet.

Chinese officials wrote things down. They wrote down grain shipments, tax disputes, court gossip, unusual weather, comets, floods, the price of salt, and the personal failings of their colleagues in exhaustive detail. A construction project of this magnitude would have involved conscripted labour, budgets, and a responsible official who very much wanted credit for it. And yet across the entire surviving corpus of Chinese historical writing, nobody mentions the enormous cave complex in Jijang, not a line, not a tax record, not a complaint. Somehow the paperwork culture that documented everything failed to document this, which is roughly

as likely as a government today building a subway system and forgetting to invoice anyone. Part four, the caves were never used for anything obvious. There are no inscriptions covering the walls, no shrines, no accumulated layers of habitation, No grave goods, no storage remains. Aside from one carving found in the first cave showing a horse, a bird and a fish, plus a few marks, the interiors are essentially blank. The most enormous unexplained excavation in the region contains almost nothing that tells you what it was for, so the theories multiply, and they all have holes. Maybe it was a quarry, and the chambers are simply the negative space left by extracting building stone. That would explain the missing rock,

since the rock became somebody's city, and it would explain the systematic toolmarks. It does not explain the load-bearing columns, the deliberate ceiling geometry, or why a quarry would carve staircases and pools, or why the chambers were kept separate with walls in some places only about half a metre thick, when merging them would have been easier and yielded more stone. That last detail is the one that really bothers people. Whoever was digging knew where the neighbouring chamber was, came within touching distance of it and stopped. That is not the behaviour of someone, chasing raw material. That is the behaviour of someone following a plan. Maybe it was a mausoleum complex or a set of tombs for the elite of the ancient UA state, whose territory covered this region.

But nobody has found the burials. Maybe it was underground barracks, a secret assembly point where an army could be trained and hidden before a campaign, which would neatly explain the total absence of documentation, since the entire point would be that nobody writes it down. It is a satisfying story. It also requires you to believe that thousands of soldiers left no trace of having been housed there, and that the secret held so completely that not one veteran official or gossiping court historian ever spilled it in the following 2000 years. Maybe it was storage on a strategic scale. Grain and supplies kept cool and dry underground. Maybe it was a mine for a specific material. Maybe it was a construction project by a state that fell before it could use the result,

which is the historical equivalent of a company going under three weeks before the new headquarters opens. Dating is nearly as unsatisfying as the purpose. Because you cannot radio-carbon data whole, researchers have to work from indirect evidence and estimates commonly land somewhere before the Han period with a frequently cited figure of around 200 before the common era, or earlier, placing the work in or before the era when China was being violently unified. That would mean this was cut by people using iron hand tools

oil lamps and human muscle, working in a hill they could not see through, and hitting tolerances that would make a modern contractor sweat. And there is one final irony that I love, because it explains why we have this puzzle in such pristine condition. The water saved it. For centuries the chambers sat completely flooded, sealed off from air, temperature swings, plant roots, and human interference, which is basically an ideal preservation environment for cut stone. The moment the villagers pumped that first pond dry, the clock started. Exposed to air, humidity cycles and visitors, the silt stone began to weather, and conservation has been a live problem ever since. The caves survived 2,000 years because nobody could get in, and the instant we got in, they started deteriorating.

Somewhere in that there is a lesson about curiosity that I am choosing not to think about too hard, so we have two enormous voids at opposite ends of Asia, one built by people who were clearly terrified, over centuries, with every design decision screaming that someone was coming to kill them. The other built by people who left no fear, no fortifications, no bodies, no inscriptions and no paperwork, at a scale that should have required a small nation's economy to pull off. In the first case, we know why and cannot pin down when. In the second, we can roughly guess when and have absolutely no idea why. And in both cases, the discovery came down to somebody doing routine local business, a wall in the way of a renovation, a pond with rumoured fish.

That is the recurring theme, and it is about to get considerably more uncomfortable, because the next site on the list was not found by accident at all. It was found by people who went looking, and what they pulled out of the ground rewrote the start date of organised religion. To get why that matters, you need one piece of background, and I promise to keep it short, because most of you have already heard of the site that broke the timeline, In southeastern Turkey near the city of San Ljurfa, there is a hill called Gobeckli Tepe. Under it sit rings of massive tea-shaped limestone pillars, some over five metres tall and weighing

ten tonnes or more, carved with foxes, boars, cranes, scorpions and snakes, arranged into enclosures by people who had no pottery, no metal tools, no wheel, no writing, no domesticated crops, and no cities. It is roughly 11,000 years old, which puts it several thousand years before Stonehenge, before the pyramids, back when the standard textbook description of humanity in this region was small mobile bands of hunter-gatherers chasing gazelle around, and occasionally inventing a new type of arrowhead. Small mobile bands do not, according to every model anyone taught for a century, quarry multi-ton monoliths, dress them, decorate them with relief sculpture, drag them across a hilltop,

stand them upright in coordinated rings, and then come back and do it again. Farming was supposed to come first. Surplus food, then settlement, then social hierarchy, then somebody with enough spare time and authority to say we are building a temple. Gobeckli Tepe arrived and flipped the order, suggesting that the gathering place might have come before the farming, and that the enormous effort of feeding a construction crew might have been one of the reasons people started messing with wild grain in the first place. The wild ancestor of in corn wheat grows practically in view of the site, which is the kind of coincidence that makes archaeologists get very quiet and start writing papers. Also worth mentioning, because it is the funniest failure in modern archaeology,

is that Gobiklitepe was not exactly hidden. A survey team from American and Turkish universities walked over it in 1963, noticed broken limestone slabs sticking out of the dirt, concluded they were probably gravestones from an abandoned medieval cemetery, wrote it up, and moved along. The hill sat there for another 30 years while a farmer plowed it and occasionally hit something annoying with his equipment. Then, in 1994, the German archaeologist Klaus Schmidt, read the old survey notes, thought the description sounded off, went to look, and realised within minutes that those were not gravestones. So the record now shows that the discovery of the oldest known monumental architecture on earth

was delayed by three decades because somebody filed it under Boring. Which brings us to the part of the story that almost nobody outside the field was paying attention to until recently. Gebekli-Tepa was never alone. About 35 kilometres east, in the rocky Tektec Highlands, sits a hill called Karahan-Tipe. It was spotted during a regional survey in 1997, and its surface was covered in the tops of pillars, more than 250 of them poking out of the soil like the world's least subtle hint. And then, for over 20 years, essentially nothing happened, because excavating a site like this

cost a fortune, and everyone's attention and funding were pointed at the famous hill down the road. Serious digging only began in 2019, as part of a Turkish research programme called Tastipela, which translates roughly as the Stone Hills, and which is currently working through a dozen sites of the same era spread across the region. That is the reveal that reframes everything. This is not one weird hill, this is a landscape, a cluster of contemporaneous sites, all neolithic, all in the same cultural world, all sharing symbols, techniques and architecture, spread across an area you could drive across in a single afternoon, assuming you had a car, which they very much did not.

The right mental image is no longer a lone mysterious monument out in the wilderness. It is a network, with regional style, shared iconography, and presumably people travelling between locations for events, which means the story is not one inexplicable temple, built by an inexplicable genius. It is an entire society that we had catalogued as too primitive to have a society, and Karahan Tepe is where that society gets genuinely strange. The centrepiece so far is a chamber that has been cut down into the bedrock itself,

not built, cut, and the distinction is everything. At the more famous site the pillars were quarried elsewhere and hauled into position, which is impressive muscle work. Here, the builder started with a lump of natural rock, went in with hand tools and removed everything that was not the room. Eleven pillars stand inside it, shaped like a little. fallacies, up to about a meter and a half tall, and they are not standing in the floor. They are the floor. They are the same continuous piece of stone as the bedrock they rise from, left behind while

everything around them was chiseled away. Nobody set them up because they were never separate. This is not construction, it is sculpture at architectural scale, which means there was no fixing a mistake. Cut too deep and you do not order a replacement pillar. You have permanently ruined the room, and possibly your reputation in front of everyone you know. Then there is the head. Carved into the wall of that chamber at roughly the height where a person's face would be is a human head with a long neck, emerging out of the rock so that it appears to be leaning into the room and looking at the pillars.

It has a face, it is watching. Depending on the light and your general emotional stability, it either reads as a solemn guardian figure or as the single most unsettling thing ever produced by the Stone Age. And I want to be clear that this is a sub-rength. objective judgment, based entirely on how I would feel about being in that room alone at night with a flame that keeps flickering. Some interpretations link the head to a snake, because of the neck, some link it to ancestor representation. All of them agree that whoever cut it wanted anyone

entering to feel observed. There is also a smaller adjacent chamber connected to the main one by a low opening, and the whole complex has features that suggest liquid was involved, with the possibility that part of the space was deliberately filled with water. Ritual, in other words, and a specific and elaborate one with a defined sequence of spaces you move through, in the dark, under the eye of a carved face, surrounded by rows of stone. This is not a group of people leaving offerings under a nice tree. This is choreographed religious architecture,

made by hunter-gatherers, at a date when the accepted narrative says nobody had yet worked out how to keep a goat. Karahan-Tipe has also produced sculpture of a quality that keeps surprising people, and neighbouring sites in the same programme are producing even more. At Sayberg, excavators uncovered a relief panel showing a narrative scene, with a human figure flanked by two leopards and a separate scene with a bull. Narrative is a big word in this context. It means the carving is not just an image of an animal, it is an event, a story,

something with a before and an after, which implies shared mythology, that everyone looking at the wall was expected to recognise. That is a culture with content, not just decoration. And here is the practical problem that turns all this into a genuine puzzle rather than a nice art history discovery. You cannot cut a room out of bedrock or carve two dozen relief panels in your spare time between hunting trips. Projects like this take crews, and crews take food, and food for a crew of hunter-gatherers has to be brought in from a wide area, because you cannot pop down to the granary you do not have. Excavations at these sites have turned up enormous limestone basins, some with capacity well over 100 litres,

and residue analysis has raised the very fun possibility that some of them held fermented grain, meaning the ancient equivalent of beer may have been part of how you paid your workforce. Add in the animal bone deposits from mass consumption of wild game, and the picture becomes feasting on an industrial scale. Somebody organised the hunt, somebody organised the transport, somebody organised the brewing, and somebody had the authority to say that all of this was happening on a particular date at a particular hill. That is administration. Unsurprisingly, there was no paperwork since writing was still about 6,000 years away, so it was all held in people's heads, which means all of it depended on tradition, memory,

and someone with enough standing to be obeyed. Every one of those things is a form of institution, and institutions are supposed to be the reward you get for inventing agriculture, not the thing you do beforehand for fun. Visit BetMGM casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMGM and GameSense remind you to play responsibly. 19 plus to wager.

Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact Connects Ontario at 1-866-531-2,600 to speak to an advisor. Free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. There is one more wrinkle that has quietly. changed how researchers talk about all this. The early framing was Temple versus Village,

a sacred site out in the landscape with nobody living at it. Newer work, including at Karahan Tepe, has found evidence of ordinary residential structures alongside the special buildings, so the emerging model is less mysterious cathedral in the wilderness, and more town with an extremely intense ceremonial district, which is somehow both more normal and more impressive, because a permanent settlement that predates farming is its own headache for the timeline. And then, because these people were not done being confusing, there is what happened at the end. At Quebecli Tepe, the enclosures were deliberately backfilled, filled in, on purpose, with rubble and sediment and debris by human hands.

Nobody knows why. Retirement of a sacred space, ritual burial of the building, a scheduled closing ceremony after a set number of years, or simply the local approach to renovation. take your pick. But it is the reason the site survived in such extraordinary condition, and it is the exact opposite of the disposal problems we have been looking at so far. Elsewhere, the mystery is where all the excavated rock went. Here the mystery is why anyone would go out and find that much material and then shovel it into their own monument.

The scale of what remains unexcavated is the part that should keep you interested. Ground penetrating surveys indicate that only a small fraction of Gobeckli-Tepipe has been dug, with common estimates somewhere around 5%, and Karahantepe has barely been opened. There are more hills in the programme, and there are almost certainly hills that are not in the program yet because nobody has looked at them properly. Since the last time a survey team looked at one of these properly, they decided it was a graveyard and went home. Every season now produces something that would have been the find of the decade 40 years ago, and the pace is such that any documentary about this region is out of date by the time it finishes editing.

What we still do not have, and we'll probably never have, is the why. No text, no myth, no name for the carved face, no explanation of the 11 pillars, no clue what the leopard scene meant to the person who cut it. We have the stage, the props, the theatre, and the seating, and we have lost the entire script. The gap is not technical. We can date the layers, analyse the residue and model the labour. We simply cannot ask, and the people who could answer stopped being able to be able to. We can't ask, and the people who could answer stopped being able to. to about 11,000 years ago. Now, that mystery is about the mind. The next one is about muscle, and it is the single most argued about pile of rock on the planet. Above the Peruvian city of Cusco, at roughly 3,700 meters, where the air already makes visitors feel like they have been personally insulted by the atmosphere, three enormous terraced walls run in a zigzag along a hillside for about 400 meters. This is Saxe-Huhrmann, and if you have ever seen a photograph of impossible ancient there is a very good chance you are looking at it. The walls are built from blocks of limestone and other hard stone, dry laid with no mortar of any kind, and the individual pieces are ridiculous.

The largest are commonly estimated at well over 100 tonnes, with figures ranging into the 200-ton neighbourhood depending on which stone you are measuring, and how generous your estimate of density is. The biggest visible blocks stand taller than a person by several times, and they are not cubes, not one of them is a cube. Every single stone is an irregular polygon, with eight, ten, twelve or more separate faces, each one shaped to match the equally irregular stones around it. This is where the famous line about paper comes from, and for once the cliché is basically accurate. The joints between these blocks are so tight that there is no visible gap to slide anything into. You cannot get a blade in, you cannot get a card in, water does not run

through the seams, and every joint is unique, because every stone is unique, which means there was no standardisation, no repeatable module, no shortcut. It is a jigsaw puzzle where each piece weighs as much as a loaded truck. The pieces were all custom cut, and there is no picture on the box because the box does not exist, and neither does the concept of a written plan. Which is the actual scandal here, and it usually gets lost under the more sensational claims. The Inca had no writing system. They had Kipu, the knotted record records, which reliably encoded numbers and accounting and possibly more that we cannot read, they did not leave technical manuals. So for a building program of this magnitude, executed by an empire that stretched thousands of kilometres along the Andes, and ran on organised labour at enormous

scale. There is no surviving description of the method written by the people who used it. Everything we have comes from Spanish accounts written after the conquest by men who are describing something they had not watched being made, or from later chroniclers working from oral testimony or from modern experimental archaeology, and the Spanish accounts, to their credit, are hilariously overwhelmed. Pedro Chiesa de Leon, who travelled through Peru in the 1540s and 1550s, and had the good sense to ask questions, was told the work involved something on the order of 20,000 labourers, organised into groups handling different jobs, quarrying, hauling, cutting, and placing. Garcilaso de la Vega, born in Cusco to a Spanish father and an Inca noble woman, grew up literally

next to this thing and wrote about it as an adult in Spain. His account is the source of the estimate that construction ran for decades, and of one of the great anecdotes in Andean history, the story of the tired stone. According to what he was told, a colossal block was being hauled toward the site, when it got away from the cruise on a slope, killed a horrifying number of the men moving it, and was abandoned where it lay, after which people said the stone had grown tired and refused to go further. Whether the details are exact or inflated by memory and time, the story tells you something real, which is that moving these things was lethal, and everybody knew it. So how did they do it? And I want to be genuinely fair here, because this is a subject

where the internet immediately jumps to the conclusion that it must have been done with tools nobody has found, and the honest answer is much less exciting and much. much more impressive. Start with the material. The Inca did not have iron tools, but they did have very good stone hammers, chunks of hematite, quartzite, and similar hard rock in various sizes, and these are found in quantity at Andean quarry sites. The technique was not carving in the European sense with chisels cutting lines, it was pounding. You strike the surface repeatedly with a hard hammer stone, and the target rock crumbles away in small amounts, and you keep going, switching to smaller hammers as you get closer to the final surface until you have shaped the face you want.

It is slow and it is boring and it works. The Swiss architect and researcher Jean-Pierre Protson spent years on this problem in the 1980s, went to Inca quarries, picked up the hammer stones that were lying there and demonstrated experimentally that you can shape and dress Andean stone this way, at rates that are entirely plausible for a large organised workforce. He also worked out how the fitting was likely done. done, and the answer is trial and fitting. You lower a block into place, see where it makes contact, lift it, pound down the high spots, lower it again, check, repeat. The reason the joints are so

perfect is that somebody kept putting a 100-ton stone down, taking it back off and adjusting it, possibly for a very long time, in a process that would drive any modern project manager to an early grave. Evidence for this workflow is visible on the stones themselves. Many blocks at Andean sites have protruding nubs and bosses left on their outer faces, which are exactly what you would expect if ropes and levers were being attached for manoeuvring, and which in many cases were simply never removed after installation. There are unfinished blocks abandoned along transport routes, showing the process interrupted mid-job. There are quarries with partially cut stone still attached to the bedrock. The chain of production is documented in the landscape even though it is not

documented on paper. What is genuinely not solved is the heavy transport. Material for the complex came from multiple sources, some limestone from relatively close by, other stone hauled from quarries tens of kilometres away, and the routes involve river valleys and serious elevation change. The standard proposal is ramps, log rollers or sledges, packed earth roadways, enormous numbers of people on ropes, and an organisational system called metre, in which communities across the empire owed labour to the state on rotation. That system is well documented. It built roads, terraces, storehouses and bridges across the Andes,

and it is the closest thing to the answer we have. But nobody has taken a 150-ton irregular block up a mountainside, with ropes and logs in a modern experiment and published the result, so the final step of the argument remains a reasonable inference rather than a demonstrated fact. Naturally, that gap is where every wild theory on the internet has set up permanent residence. There is a second argument that is just as unresolved and gets less attention, which is what the place was for, it gets called a fortress, and the layout supports that reading. The zigzag walls remove the possibility of attacking a flat face,

forcing anyone assaulting them to expose a flank at every angle, which is a design idea that European military engineers would independently arrive at centuries later and feel very clever about. There were towers on the summit, including a round one called Muyukmarker, with internal water channels and rectangular ones alongside it, And the site absolutely was used as a fortress, most famously in 1536 when Manco Inca launched a massive rebellion against the Spanish and Inca forces held these heights above the occupied city. The fighting there was brutal, it went on for days, and it killed Juan Pizarro, brother of the conquistador Francisco Pizarro, who took a stone to the head during an assault. But being used as a fortress in an emergency does not prove it was designed as one, and plenty of researchers argue the complex was primarily.

ceremonial, a religious and administrative centre, with vast open plazas where enormous state occasions were held. The two readings are not mutually exclusive, because Inca State Religion and Inca State Power were the same institution wearing different hats. There is also the long-standing tradition that the plan of Cusco was laid out in the shape of a puma, with this complex forming the head and the zigzag walls forming the teeth, which is either a genuine piece of Inca-urban symbolism or a later interpretation applied to an existing city. And specialists have been politely disagreeing about it for a long time, and now the part that I find the most bleakly entertaining, because it explains why what you see today is a fraction of what stood. After the conquest,

Spanish Cusco needed building material. There was an enormous, freshly conquered, structurally perfect supply of pre-cut stone sitting on the hill above town, and colonial authorities did the obvious thing. They quarried it. For decades, the the site was systematically dismantled, and its stones went into churches, government buildings, and the mansions of the new elite down in the city. This was not a secret or a rumor, it was policy, and later Spanish officials even issued orders trying to slow the destruction. Once it became clear the place was being erased. Here is the punchline. They only took the stones they could move. The upper courses, the smaller blocks, the towers, the finished structures on the summit,

all of that went into the colonial city. The bottom terraces, The ones built from the truly monstrous blocks stayed exactly where they were, because a crew with 16th century equipment took one look at a 150-ton polygonal boulder and concluded that it could stay. So the reason we can stand there today and marvel at the giant foundation stones is not that they were the most important part, it is that they were the leftovers. The world's most famous demonstration of impossible ancient engineering survive specifically because it was too inconvenient to steal, which may be the most honest summary of how heritage preservation has worked throughout most of human history. That also means the version everyone photographs is a stump. Contemporary descriptions talk about a structure of

considerable height and complexity, with buildings, chambers and towers rising above the terraces. What remains is the base course of an enormous complex whose upper stories are now distributed across the historic centre of a colonial city, which is a genuinely strange thing to think about while walking around Cusco looking at church walls. And there is a final detail that quietly settles part of the argument about whether the Inca knew what they were doing. The Cusco region sits in serious earthquake country. In 1650 and again in 1950, major quakes hit the city

and did enormous damage to the colonial buildings. Repeatedly, the Spanish structures cracked, collapsed, or lost their upper floors, and the Inca walls underneath them did not. Dry-laid polygonal masonry with no mortar, has a property that rigid mortared walls lack. When the ground moves, the blocks can shift slightly against each other, absorb the energy and settle back into position, because they were never glued in place to begin with. Those absurd multi-angled joints are not just showing off, they interlock, resisting the sliding and separation that brings a straight-coarsed wall down. So the

answer to why every stone had 12 faces might be less about aesthetics, and more about the fact that the builders lived somewhere the ground kicks, and they had figured out, without written engineering how to make a wall that survives it. Which is the theme worth holding on to here, Karahantepe shows a society doing something it supposedly had no capacity to organise. Saxe-Huhrman shows a society doing something it supposedly had no tools to accomplish.

In both cases, the sensible explanation is not that somebody else did it. It is that we have persistently underestimated how much can be achieved with enormous amounts of coordinated human effort, patience measured in decades, and a complete disregard for the modern assumption that a job should be finishable within one person's career. But there is a limit to how far coordination alone gets you, and the next site is where the effort argument runs straight into open water, because it is one thing to drag stone across a hillside.

It is another thing entirely to move three quarters of a million tonnes of basalt across a coral reef in the middle of the Pacific Ocean and build a city out of it on top of the sea. Pompey is a small volcanic island in the Western Pacific, part of the Federated States of Micronesia, and it is about as remote as land gets. It is green, mountainous, drenched in rain almost daily, ringed by mangroves and a barrier reef, and it is roughly 4,000 kilometres from Australia, and about the same from Japan. The modern population of the whole island is in the neighbourhood of 36,000 people. Nothing about it, says Imperial Capital, and yet off its eastern shore, built out over the reef flat in shallow water,

sits one of the most improbable construction projects ever attempted by human beings. Nan Madol is a city of about 92 artificial islets, spread across roughly 75 hectares of tidal water, separated by a grid of canals and enclosed by stone walls that in places rise to 7.6 metres and run up to 5 metres thick. Not a city beside the water, a city in the water, on platforms that did not exist until somebody built them, in a lagoon, on a reef, in the middle of the Pacific Ocean.

The name itself refers to the spaces in between, meaning the canals, which is a refreshingly practical thing to call a place compared to most ancient capitals, which tend to be named after a god or a guy. The building material is the part that breaks people's brains, so let me explain what these stones actually are, because the common description makes them sound man-made and they are not. When certain lava flows cool, slowly and evenly, the rock contracts and cracks into long prismatic columns, usually with five, six or seven sides.

You have seen the effect if you have ever looked at photographs of the giant's causeway in Northern Ireland. Nature produces enormous stacks of these things, and they look so manufactured that the standard human response on first sight is to assume someone made them. Pompuy has them in quantity, and the builders of Nan Madol figured out that a natural supply of long, straight, heavy, roughly hexagonal stone beams is basically, a construction kit sitting in a hillside. So they built with them the way you would build a log cabin. Lay a row of columns in one direction, lay the next row across them at right angles. Repeat, filling the interior with coral rubble as you go. The crisscross stacking locks the courses

together without mortar, which is essential when you're building in a place where saltwater moves under you twice a day and the occasional typhoon comes through to see how your work is holding up. The result is a set of massive platforms and walls that have stood for some something like 800 years in one of the more punishing climates on Earth. Now the numbers. Individual columns commonly weigh several tons, plenty of them run 5 tons and up, and the heaviest pieces in the complex are estimated in the range of tens of tons, with figures around 50 tonnes quoted for the largest cornerstones.

Total mass of material moved and placed across the whole site is estimated at roughly 750,000 tons. Take a second with that. quarters of a million tonnes of stone positioned by hand on an island where the total population probably never exceeded the low tens of thousands by people with no metal tools, no wheels, no draft animals, no pulleys, and no cranes. And that is before we get to the transport problem, which is where this stops being merely impressive and starts being an unsolved question. The columnar basalt quarries are not next door. The main sources are on the other side of the

island, and even the closer candidate sites involve serious distance. So the So the columns had to be extracted from the hillside, moved down through rainforest and volcanic terrain, then somehow carried across water, over a reef, into the lagoon, and finally lifted into position, in some cases to the top of a wall taller than a two-story house. And here is the thing that keeps this in the unsolved column rather than the solved one. Nobody has managed to demonstrate the water stage convincingly. The standard proposal is rafts, either bamboo or logs, floated at high tide with the buoyancy of the water taking most of the weight. That is a completely reasonable idea, and it is almost

certainly roughly what happened, but attempts to reproduce it with real basalt columns have a persistent tendency to end with the raft underwater, because Ba'alt is dense, unwieldy, and entirely uninterested in your project timeline. Some accounts of these experiments involve modest columns and still describe difficulties, which is a polite way of saying the ocean one. There is also a floating theory that the stones were rafted at high tide and settled into place as the water dropped, letting the tide do the heavy lowering, which is genuinely clever and fits the tidal environment. What is missing is not plausibility, it is proof, and there is a shortage of volunteers willing to spend a year of their life sinking rafts in a lagoon to find out.

Then there is the location itself, which deserves a serious question. Why here? Because as a place to live, Nanmodole is objectively terrible. There is no fresh water on the site. None. Wells dug into the islets fill and empty with the tide, so they are brackish, which means every drop of drinking water had to be brought in from the main island by canoe. There is no farmland, there is no soil worth speaking of. Nothing edible grows there. Every yam, every breadfruit, every fish that was not caught right off the wall, every coconut, every bit of firewood, all of it had to be imported daily indefinitely across water by people who lived somewhere else. Naturally, that is not a minor logistical footnote.

That is a permanent supply operation supporting a resident elite, and it tells you exactly what kind of society we're talking about. Somebody had the authority to make thousands of people row food to an island of rock every single day, so that a small number of important individuals could live on top of a reef and be impressive, which they very much were. The complex is divided into functional districts, and the layout is not random. There is an administrative and residential sector, including Pankadira, associated with the ruling authority. There is a mortuary sector dominated by Nandauas, the royal tomb enclosure, whose outer walls are the tallest at the site, and whose corner stones are the heaviest, because when you are building the place where your dynasty is buried, you obviously do not use the small ones.

Inside the outer wall is a second wall, and inside that a central tomb of stacked basalt. There were islets for priests, for canoe construction, for coconut oil preparation, and for specialised ritual functions. And then there is Ided, which is my favourite piece of the whole story. It was the site of a ceremony centred on a sacred eel, which was understood to embody a deity, and which was ritually fed the cooked liver of a sea turtle. Turtles were caught, brought in, prepared with ceremony, and the honoured guest at the meal was an eel

living in a stone enclosure in the middle of an artificial city. There are a great many ways to signal that your civilization has surplus and organisation and maintaining a full-time ceremonial eel with a specialist diet is one of the more convincing ones. The archaeological deposits from that islet include enormous accumulations of turtle bone and charcoal, which is how we know this was not an occasional thing but a sustained institution. The people who ran all this were the Soderler dynasty. The general picture from archaeology and oral tradition together is that there were people active on the reef area from remarkably early, with evidence of human presence going back to around the first couple of centuries of the common era.

But the megalithic city as we see it belongs to a much later push, with the great construction phase generally placed from around the 12th century onward. The Sodler line unified the island under a single centralized authority, which was itself a break with how Pacific Island societies of this size usually organised themselves, and they ruled from this offshore stone complex for several centuries. Centralised is doing a lot of work in that sentence. Tradition describes the later Soderler rulers as increasingly oppressive, demanding tribute and controlling religious practice, and one of the recurring themes in the oral history is that they required chiefs from around the island to live at Nanmadol where they could be watched.

That is a policy so recognizably authoritarian that it shows up independently in the history of empires all over the world, and it means the city was not only a palace and a temple complex, it was also functionally a very expensive way to keep an eye on people. It ended around 1628, and the story of Howe is one of the great narratives of the Pacific. A warrior named Isichel arrived from Kossrai, an island several hundred kilometers to the southeast, with a force of fighters. Tradition holds that he was a semi-divine figure with a personal grievance against the ruling order, that he landed under the cover of ordinary visitation, and that the invasion turned into a full-scale battle across the canals of the city. The last Saudelaire was defeated, and according to the story he returned to the

treated into a stream and transformed into a fish, which is either mythological embellishment or the single most decisive resignation in the history of government. Isukalekal did not restore the old system. He established the Nanmwaki system, dividing the island into chiefdoms with their own paramount chiefs, a structure that persists in Pompeian society to this day. He also moved into Nan Madol himself, at Pecap, because when you conquer a stone city built on the ocean, you do not immediately move into a hut. But the city, but the city is a city. function was already broken. Without the centralised tribute machine that fed it, an island with no water and no food is not a capital. It is a maintenance problem. Over the following generations it was

gradually abandoned and by the early 1800s essentially nobody was living there. And then something happened that has quietly cost us most of what we would like to know. The place became taboo. Pompeians came to regard Nanmodole as a dwelling place of spirits, dangerous to disturb, especially after dark. Visit BetMGM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and GameSense remind you to play responsibly, 19 plus to wager. Ontario only.

Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact connects Ontario at 1-866-531-2,600 to speak to an advisor, free of charge. BetMGEM operates pursuant to an operating agreement with Eye Gaming Ontario. Where some see heroes and others see egos. Bloomberg sees the era of billionaire athletes. A fad to some, the future of money to others.

We see crypto's trillion dollar swings. The end of jobs or the end of human struggle. We see the endless funds fueling the AI hype. While others follow the noise, we follow the money. Learn more at Bloomberg.com. Visiting during the day for specific purposes was one thing. Staying overnight was another, and the prohibition around the tombs in particular was serious. The effect of that taboo on the historical record is enormous, and it is the exact opposite

of what you might expect. Normally, an abandoned monumental site loses its story because the population disappears. Here the population never went anywhere. The descendants of the builder are still on the island, still speaking the language, still maintaining an oral tradition that is genuinely rich about the political history, the dynasty, and the conquest. What faded was the practical knowledge, because you do not pass down detailed instructions about a place you're not supposed to go near. The gap in the tradition gets filled instead by the origin legend, and I want to describe it fairly, because it is not a joke to the people who tell it. Two brothers, Olusipa and Olosopa, arrive from somewhere else, sorcerers looking for a site to build an altar so they could properly honour

the deity of agriculture. They tried several locations, failed, and eventually settled on the reef. And the way they moved the stones, according to the tradition, was with the help of a flying dragon that carried the basalt through the air. That is the answer that has survived. When outsiders started asking in the 19th and 20th centuries how a city of 750,000 tonnes of stone got built on a reef, the local explanation was magic, not because Pompeians were unaware of how construction works, but because the technical memory had been severed by centuries of prohibition, and the mythological memory had not. The outsiders who came asking, incidentally, did not distinguish themselves. The German ethnographic expedition, led by Paul

Hamburg in the years around 1910, produced the first serious documentation of the site, mapping the islets and recording traditions, and that work remains valuable. Less useful was the behaviour of a certain German colonial governor a few years earlier, who decided to excavate at the royal tomb enclosure, despite being warned repeatedly by locals, that this was an extremely bad idea. He opened it, removed material, and then died shortly afterward, reportedly the very next day. Medically speaking, this was probably a sunstroke or an infection in a tropical climate with no antibiotics. Narratively speaking, the story of the governor who ignored the warnings and did not live to file the paperwork,

has enjoyed a long and healthy afterlife on the island and in every book written about the place since, including, unfortunately for my credibility, this one. What modern archaeology has been able to establish is real, but limited. Excavation and radiocarbon work by researchers, including teams led by American and Pompeian archaeologists, have anchored the construction sequence, confirmed the long occupation history, and documented the ritual deposits. survey work has mapped the islets and their functions. What none of it has produced is a technical answer to the transport question or a written record of any kind, because there was no writing system here

and the accounting was done in memory and obligation rather than in symbols. There is one more piece that shifts the perspective, and it is the same lesson we keep running into. Nanmadol is not unique. About 550 kilometres away, on the island of Kossray, there is another megalithic complex called Le Lé-Lu, built with the same columnar basalt in the same criss-cross technique, with its own walled compounds and canals, and its own dynastic history. Two islands, separated by open ocean, building comparable stone cities out of comparable material in overlapping periods,

with a tradition that explicitly connects them through the story of the conqueror, who sailed from one to the other. So this was not one isolated, eccentric project by one unusually ambitious ruler. It was a regional style of monumental. monumental power in the middle of the Pacific, executed by seafaring societies whose navigational and organisational abilities, Europeans consistently underestimated for the entire period they were writing things down about them. Today the site is a World Heritage location, added to the list in 2016. It was placed on the danger list at the same moment, which is one of the more depressing

double announcements in Heritage Administration. The threats are not tourists or looters. are mangroves growing through the walls, roots levering the columns apart and silt filling the canals that once let canoes move through the city. The ocean is not tearing it down so much as the jungle is quietly digesting it, and the resources available to a small island nation for large-scale conservation of an offshore megalithic complex are exactly what you would guess they are. At least that one is above the surface, where we can measure it and argue about it in daylight. The next one is not. The next one sits 25 metres underwater off the coast of Japan, and the argument about it is not

between believers and scientists. It is between scientists, and they have both been down there, and neither will move. Yonaguni is the last piece of Japan before you run out of Japan. It sits at the western end of the Ayama chain in Okinawa Prefecture, about 110 kilometres from the coast of Taiwan, with a population somewhere around 1,700 people, a couple of villages, some cattle, a legendary local liquor, and a stretch of ocean with currents strong enough to make experienced divers reconsider their hobbies. In winter, schools of hammerhead sharks gather in the water off the island, and that fact is the entire reason this story exists. In 1986, a local diver named Kihachiro Aratakei was in the water off the southern coast, scouting locations that could

be developed for shark-watching tourism. Instead of sharks, he found a wall. Then he found that the had a corner. Then he found that the corner had steps. What he had swum into is a massive rock roughly 50 metres long and 20 meters wide, rising about 27 meters from the seabed, with its upper surface only a few meters below the waves. It has broad flat platforms. It has terraces stacked above one another. It has faces that met at angles close to 90 degrees. It has features that divers immediately nicknamed the stairs, the loop road, the twin pillars, and the turtle. because the human brain cannot look at a rock for more than four seconds without assigning it a job title. Photographs of it look staged.

Video of it looks like a set from a film about a sunken civilization, and I mean that literally, because footage of the site has been used in roughly every documentary about lost civilizations produced in the last 30 years, usually with dramatic music laid over it to help you reach the conclusion the editor has already reached. And here is why this chapter is different from everything before it. There is a real argument about this rock, and it is not the usual one. It is not a case of credentialed researchers on one side, and enthusiastic hobbyists with a website on the other. Both of the main figures in this dispute are professional geologists.

Both have been to the site personally. Both have looked at the same surfaces with their own eyes, in the same water, and they have reached opposite conclusions, and neither has been able to close the case. On one side is Masaki Kimura, a marine geologist at the University of the Rukius, who has dive the formation repeatedly over decades, and who argues that it is at least partly the work of human hands. His position has shifted over the years, which is worth noting because critics often describe it as one fixed claim. In broad terms, he holds that natural rock was shaped, cut, and modified into a structure, that the site includes worked elements such as an upright steely, a channel

functioning like a drain or water feature, quarry-style marks, and holes that look like sockets for wedges, and that all of this was done when the area was dry land. He has also identified what he interprets as carvings, including a face and animal figures, and he argues for a whole complex of related features spread across the surrounding seabed. The dating logic follows from sea level, and that part is not controversial at all. During the last glacial maximum, with enormous volumes of water locked into ice sheets, the ocean stood more than 100 metres lower than it does now, and the shallow shelf around Yonaguni was above the waterline. As the ice melted, the sea came up and the shelf went under.

Depending on which depth you take and which curve you follow, the top of this formation would have been dry until somewhere in the region of 10,000 years ago. So if it was built, it was built by people living at the tail end of the last ice age. That is the claim that makes everybody sit up because 10,000 years ago in the Japanese archipelago means the Joman period, and the Jomon are already famous for doing something remarkable early, namely producing some of the oldest pottery on the planet. What they are not known for is monumental stone architecture. There is no established megalithic tradition in Japan at that date, no comparable structures on land, no documented toolkit for heavy quarrying in this region at that time.

A cut and terraced platform on this scale would not be an addition to the record. It would be a demolition of it. On the other side is Robert Schoch, a geologist at Boston University, and I want to underline who this is, because it matters in normal. for how the disagreement should be read. Shock is not a defender of conventional chronology by temperament. He is the man who argued that the weathering on the enclosure around the Great Sphinx in Egypt shows the signature of heavy rainfall, implying a monument considerably older than Egyptology accepts, a position that made him deeply unpopular with a lot of archaeologists

and extremely popular with a lot of television producers. If anybody on Earth had a professional incentive to look at a submerged rock formation formation and announce a rewritten timeline, it was this guy. He went to Yonaguni, he dived it himself. He came back and said it is natural. His argument is plain geology. The rock is a fine-grained sandstone and mudstone belonging to a formation that runs through this part of the rukus. Sedimentary rock like this is laid down in layers, and it splits along those bedding planes, which are flat, parallel and close to horizontal. It also fractures along joints, which in this tectonically busy region tend to run vertically. Put horizontal bedding together with vertical

jointing and what you get automatically with no help from anyone is flat surfaces meeting other flat surfaces at right angles. Add wave energy, ferocious currents and the occasional major earthquake and block separate slide and drop producing steps, ledges, terraces and channels. Nature does not need a site supervisor to make a 90 degree angle in this rock. It makes them as a default setting, and then comes the single most effective argument in the entire debate, which is almost embarrassing in its simplicity. Go and look at the coast of Yonaguni above the waterline. The cliffs and shore platforms on the island show the same style of formation, with stepped terraces, flat tables and sharp angular edges in the same rock produced by the same processes,

in full daylight, where nobody is claiming a lost civilization was involved. If features that look manufactured turn up a short distance away on dry land with no controversy attached, it becomes very hard to insist that the underwater version has to be artificial. There is also the matter of what is not there. Every genuine ancient construction site leaves debris, broken tools, flakes and chips from working stone, pottery if the culture had it, and the Joman absolutely did, charcoal from fires, animal bone from meals, burials somewhere nearby, foundations, post holes, dumps. This is the ordinary garbage of human existence, and it is usually more informative than the architecture. At Yonaguni, that assemblage has not turned up. No artifact from the site has been recovered and

shown to be a manufactured object of the right age. Underwater preservation is brutal and the currents are savage, so absence of evidence counts for less here than it would on land, but it still counts, and after nearly 40 years of attention, the silence is loud. The counter-arguments from the artificial side are not nothing, and it would be lazy to pretend otherwise. Supporters point to blocks sitting in positions that do not obviously match a simple collapse pattern, edges that stay straight over in probably long runs, and step proportions that seem to repeat. They point to channels cutting across the formation in directions that do not follow the obvious joint orientations, and they make the strongest version of the case, which is that even if the

underlying mass is natural, and Camura himself accepts a large natural component. People could have taken an existing rock terrace and improved it, squaring edges, deepening steps, and using it as a platform. That version, the modified natural feature is far more defensible than the sunken city version, and it is much harder to disprove, because a light human touch on a natural formation is exactly the sort of thing that vanishes after 10,000 years underwater. Unfortunately, the discussion keeps getting dragged sideways by the company it attracts. Early on, some of the more expansive readings tied the site to Mew, a supposed sunken Pacific continent invented in the 19th century by writers with a taste for romantic geography and no evidence at all. The moment that word enters

a conversation, a large slice of the academic world stops listening and a large slice of the alternative history world starts. The formation also became a fixture of popular books and television, most prominently through Graham Hancock, whose interest brought it a global audience, and, at the same time, made mainstream researchers even less eager to be seen anywhere near it. The result is a feedback loop that actively prevents the question from being answered. The site gets famous, professionals get wary, fewer professionals study it, less hard data exists, and the vacuum fills with speculation, which makes professionals warier still.

That is where things stand today, and it is genuinely underwent. satisfying. The formation has no official status as a cultural property. The Japanese cultural authorities do not treat it as an archaeological site, so there is no state protection program, and no funded long-term research plan of the kind that could settle this within a decade. The main human activity at the location consists of recreational divers being hauled sideways by one of the strongest currents in the region while trying to hold a camera steady. So what would actually settle it? This is worth spelling out, because it is the difference between a mystery and a mystery that is merely inconvenient. Toolmarks would settle it,

not marks that look like toolmarks to somebody who wants them to be toolmarks, but marks with the specific microscopic signature of percussion or cutting, properly documented in places where natural fracturing cannot produce them. Datable worked material would settle it, meaning a human-made object tied to the formation and given an age. A quarry waste field would settle it, since removing stone produces debris and debris piles up in predictable places. Evidence of assembly would settle it, meaning blocks resting on the bedrock rather than continuous with it, material that had to be transported, anything that required a decision rather than a fracture. And there is an argument sitting quietly underneath all of this that I think is underrated. In every case we have looked at so

far the giveaway was never the finished shape. It was the process trail, the quarries, the abandoned and half-finished pieces, the standardized tool patterns, the disposal problem, the tools left lying around. Human construction is messy in a recognisable way, and the mess is usually easier to find than the monument. At Yonaguni, we have a monument-shaped object with essentially no mess. That is either evidence that nobody built it, or evidence that 10,000 years of one of the most energetic marine environments on the planet scrubbed the mess away. Both statements are reasonable. them can be tested, and it is not the second one. What I find genuinely valuable here is that this is an honest deadlock, and honest deadlocks deserve more respect than they usually get. It is very

easy to sort the world into serious researchers and cranks, and then stop thinking. That shortcut does not work at this site. Two geologists with real credentials went into the same water, looked at the same terraces, and disagreed about whether the right angles are a human fingerprint or the standard behaviour of sandstone. Neither is lying, neither is a fool. They are weighing ambiguous evidence, and their conclusions are shaped partly by which claim each of them considers the more extraordinary one, which is not a flaw in science, it is science. If you want my read, the geological case is stronger, and the above-water comparison is very hard to get around. The modified platform idea stays open, and it would be fascinating if it were ever confirmed. But right now,

the case rests on how the formation looks rather than on independent traces of human activity. How something looks is not nothing, but how something looks is also the reason a hill on Mars once made international headlines for resembling a face, which sets up the strangest decision on this entire list, because the next site is the precise opposite situation. There is no ambiguity about whether it is artificial. There is no argument about who is inside or when it was sealed. We have the ancient text describing what was put in there with him, and modern instruments have already found supporting evidence in the soil, and it has been sitting unopened for 50 years

because the people with the shovels have deliberately decided not to use them. In the spring of 1974, Shanxi Province was in a drought, and a group of farmers in Lintong district east of Cheyenne were doing the least glamorous job in agriculture, digging a well. A couple of metres down, their tools started coming up with pieces of hard-fired clay, then a torso, then, memorably, a head. staring up out of the dirt at a group of men who had come out that morning hoping for groundwater. Local reaction was not immediate excitement.

Broken pottery is common in that landscape, and the initial assumption was an old kiln site. Some of the fragments were carted off. There was talk among villagers about bad luck and unhappy spirits, which under the circumstances is a completely rational risk assessment. It took a local heritage official named Zhao Kangmin, recognizing what the fragments were, collecting them, and beginning to reassemble figures before anyone understood that these farmers had punched a hole into the outer edge of the largest funerary complex ever constructed by a human being.

The well, incidentally, never produced usable water, which makes this the single worst plumbing project and the single best excavation in modern Chinese history. Visit BetMDMGM Casino and check out the newest exclusive, the Price is Right Fortune Pick. BetMDM and Game Sense remind you to play responsibly, 19 plus to wager. Ontario only. Please play responsibly.

If you have questions or concerns about your gambling or someone close to you, please contact connects Ontario at 1-866-531-2,600 to speak to an advisor, free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. Where some see heroes and others see egos. Bloomberg sees the era of billionaire athletes. While others follow the noise, We follow the money. Learn more at Bloomberg.com.

What came out of the ground over the following years is familiar to almost everyone. Pit 1 alone is over 200 metres long and holds an estimated 6,000 figures in battle formation, infantry and ranks with chariots between them. Pit 2 contains cavalry, chariots and archers. Pit 3 appears to be a command post, small, with a chariot and officers, which is exactly the kind of arrangement you would expect if somebody would. was modelling an actual army rather than decorating a hallway. Estimates for the complete force run to around 8,000 warriors, over 500 chariot horses, and more than 100 cavalry mounts.

The production system behind them is a story in itself, and it is more interesting than the myth people usually repeat. The famous line is that every face is unique, and broadly that is true, but they were not individually sculpted from scratch by artists having a moment. They were mass-produced with a modular workflow. Bodies were built from standardized components, legs, torsosos, arms, heads, made with moulds and repeated forms, then assembled and finished with individualising details, ears, hair, mustaches, expressions, applied by hand. It is an assembly line with a customisation step at the end, which means the chin state was doing scalable manufacturing with quality control more than 2,000 years before anyone wrote a business book about it.

about the quality control because the workshops signed their work. Inscriptions stamped or incised into figures record the names of foreman supervising production, dozens of them, along with workshop identifiers. That is not artistic signature, that is accountability. If a figure failed, the state knew exactly whose team to have a conversation with, and given the general management style of the Chin Dynasty, that conversation was probably not a performance review. The same standardization shows up in the weapons. Bronze crossbow triggers recovered from the pits are made to tolerances tight enough that components are effectively interchangeable between units, which is a genuinely advanced manufacturing concept and a considerable military advantage when parts break in the field.

And they were all painted. Every figure was originally finished in bright colour, reds, greens, blues, purples, whites, over a lack of ground, with faces and hands in flesh tones and armor picked out in detail. The pigment list includes a synthetic barium copper silicate, now known as Chinese purple, a manufactured compound rather than a ground-up mineral, which tells you the chin had a chemical industry as well as a ceramics industry. Here is the part that turns this from an art history note into the central problem of the entire chapter. The lacquer layer under the paint has spent 2,000 years in damp soil, and it is saturated. When it comes out of the ground and hits open air, it dries, and as it dries it curls and flakes off the ceramic, taking the

colour with it. Not over a season, over minutes. Early excavated figures lost their surface colour within a very short window of exposure, sometimes described in single-digit minutes, and in the worst cases the paint was gone before anyone had properly photographed it. Which is why the army you see in every documentary is uniformly grey, that is not the original appearance, that is damage. We opened the box and the contents faded, while we watched. Conservation eventually caught up. Chinese specialists working with German colleagues developed treatments using consolidating agents applied to the surface within moments of exposure, keeping the lacquer stable long enough to fix the pigment in place, and newer excavations

have recovered figures with substantial colour preserved. That is a real triumph. It is also a solution that took decades to develop, and it is specific to painted lacquer on terracotta, which is one material out of the many that are likely waiting in the ground. Because everything described so far as outside the tomb, all of it, the warriors were buried in pits roughly a kilometre and a half east of the actual burial mound, positioned like a garrison guarding an approach. They are not the tomb, they are the parking lot. The tomb itself is a mound at the foot of Mount Lee,

currently standing about 51 metres high, after two millennia of erosion and farming, and originally considerably taller. Its base covers a vast rectangle, and it sits inside a double-walled enclosure within a necropolis, whose full extent runs to something on the order of 98 square kilometres of associated features. Surveys and limited excavation around the perimeter have found pit after pit of supporting content. A pit of acrobats and strong men, modelled with realistic bodies rather than military uniforms, because the emperor was apparently not planning an eternity of pure administration.

A pit with bronze waterfowl, cranes and swans, arranged along a simulated watercourse with figures of musicians. A pit containing dozens of suits of armour and helmets made from limestone plates laced together with wire, which are useless in combat and clearly made for the afterlife. Stables with real horse skeletons, mass graves of labourers who died on the job, and in 1980 two half-scale bronze chariots with drivers and horses assembled from thousand, of individual components including gold and silver fittings, umbrella mechanisms and working parts, which remain among the most technically astonishing bronze objects ever recovered anywhere. That is the outskirts. Nobody has opened the centre. What we think is in there comes primarily

from Sima Tien, the great historian of the Han dynasty, who compiled his account roughly a century after the emperor's death. His description is short, specific and completely mad. He reports that 700,000 conscripted workers were employed on the project. He says the builders dug down through subterranean layers and sealed the chamber with molten metal. He describes the interior as containing palaces and towers, filled with treasures and rare objects. He says the ceiling was made to represent the heavens with celestial bodies, and the floor was laid out as the terrain of the empire, with the rivers of China, including the Yanksy and the Yellow River, reproduced in mercury and mechanically made to flow.

He mentions lamps fuelled with whale oil, intended to burn indefinitely, which is a lovely idea and a chemically doomed one. He records that crossbows were rigged to fire automatically at anyone breaking in, that the craftsmen who knew the layout of the inner passages were sealed inside, and that childless women of the court were killed to accompany the emperor. For a very long time, the sensible academic reaction to all that was to file it under literary embellishment. A historian writing a century later, describing a tyrant everyone had reason, to portray as monstrously excessive is not a neutral source. Rivers of liquid mercury sounded like

exactly the kind of detail that grows in the telling. Then somebody tested the soil, beginning in the 1980s and repeated with better methods since. Geochemical surveys of the ground above and around the mound have measured mercury concentrations, and the results are not subtle. The soil above the central area of the mound carries mercury levels far above the natural background of the surrounding region, and the anomaly is not spread evenly. It concentrates over the centre, exactly where a burial chamber would be, and drops off toward the edges. Some analyses have gone further and argued that the distribution pattern of the anomaly shows a rough correspondence with the geography of China's waterways, with heavier readings in the areas that would correspond to the eastern northeast. That interpretation is debated and

depends on assumptions about how mercury vapor migrates through 2,000 years of packed earth, and I would not build a religion on it. But the basic first thing is. But the basic form of the fact of the finding is solid and repeatedly confirmed. There is a large quantity of mercury under that hill, it is not natural, and the ancient text that everybody had written off as poetic exaggeration described exactly that, which means the rest of Sima Chian's account has to be taken more seriously than it used to be, not literally, not all of it, but seriously. The other thing modern instruments have found is the shape of what is down there. A major geophysical survey program run in the early 2000s used resistivity, magnetic and other remote sensing methods across

the mound, and produced a picture of a large rectangular chamber structure at depth, with dimensions in the range of 80 by 50 metres, roughly 30 metres below the surface, surrounded by rammed earth walls of considerable thickness. Crucially, the surveys found no signs of the collapse or infill pattern you would expect from a chamber that had caved in or been broken into and emptied. That last point matters enormously, because the historical record contains a serious threat to the whole story. Four years after the emperor died, the dynasty he built to last 10,000 generations fell apart in a wave of rebellion, and the warlord Xiang Yu is recorded as having sacked the site. The terracotta pits show the evidence.

Timber roofing was burned, figures were smashed, and many of the weapons appear to have been carried off, which is what soldiers do when they find 8,000 racks of usable bronze. So the outer works were definitely violated, but the survey data suggests the central chamber was not, which sets up the situation we have now, which is that the largest unopened intact tomb of a major world civilization is sitting under a hill, its approximate contents described by a surviving text, its most outrageous detail already partly confirmed, and everyone has agreed to leave it alone. That decision was not made out of laziness or lack of money. China has both the archaeological expertise and the resources to excavate it.

The decision is deliberate policy, and the reason it exists has a name and a date. In 1956, a Chinese team opened the dingling tomb of the Wanli Emperor of the Ming Dynasty, and the result was a catastrophe that is still taught as the definitive cautionary tale. The excavation went ahead without the conservation technology needed to handle what came out. Silk textiles, some of the finest ever produced, were exposed to air and light and degraded rapidly, with much of the material effectively destroyed. Objects were mishandled and stored badly. Then, during the political chaos of the following decade,

the remains of the emperor and empresses were dragged out and burned by a crowd. An intact imperial tomb with contents of incalculable value was reduced within a couple of decades to a lesson in what not to do. The response was a state policy against the excavation of imperial tombs, focusing instead on rescue archaeology, meaning you dig when a site is threatened by construction, looting or erosion, and otherwise you leave it in the ground. Apply that lesson to the mound at Mount Lee and the argument writes itself. The terracotta paint problem, as mentioned, demonstrates how fast an unprepared excavation destroys information,

and that was ceramic and lacquer, which is one of the more durable combinations you can find. What Simichien's description implies is inside, and what any tomb of this period would plausibly contain, is a nightmare list for conservators. Silk, bamboo and wooden documents, lacquerware, pigments, organic materials that have been sitting in a sealed, stable, oxygen-poor environment for 2,000 years and would begin degrading on contact with the atmosphere. Bronze and jade would survive. The written records, the textiles and the pigment would be the things at risk, and they are precisely the things that would tell us the most. Then there is the small matter of the mercury. A chamber of the chamber. A chamber of the chamber of

that appears to contain a substantial quantity of liquid mercury is not a normal excavation environment. Mercury vapour is a serious neurotoxin and opening a sealed space full of it in a region with a huge resident population and a major tourism economy is a public health problem with a very short list of people volunteering to sign off on it. And there is a third argument that gets less airtime and is the most convincing to me. If you open it now, you open it with 2020's methods and 2020's methods are the worst methods that will ever be available for the job again. Every year that the tombstays closed, the toolkit improves. Imaging gets better.

Conservation chemistry gets better. Sampling gets less destructive. Analytical techniques that can pull dietary and geographic information out of a fragment of bone or a smear of residue did not exist a generation ago. Waiting is not indecision. Waiting is a strategy, and it is the same strategy a good doctor uses when the correct treatment. has not been invented yet. So the current approach is to look without touching.

Remote sensing continues, with progressively more sensitive geophysical methods mapping the structure. And there is real interest in mion tomography, which is the closest thing we have to a magic trick in this field. Cosmic rays constantly bombard the atmosphere and produce mons, particles that rain down through everything, including rock and earth.

Denser material absorbs more of them. So if you place detectors underneath or beside a large, mass and count the mions arriving from different directions over a long period, you can build a density map of what is inside, essentially an X-ray of a mountain using free particles from space. The technique has already been used to find a previously unknown void inside the Great Pyramid and applying it to a rammed earth mound in Shanxi is exactly the kind of thing it is good at. It will not read you a bamboo scroll, it could tell you the shape and internal layout of the chamber without moving a single spoonful of dirt. Which brings

this chapter to a question that is worth sitting with, because it is the only one on this list where the mystery is voluntary. Every other unknown here is imposed on us by time, erosion, water, taboo, or plain bad luck. This one is a choice. What do we lose by not looking? Realistically, quite a lot. There are almost certainly administrative documents down there, and Chin Dynasty documents would be transformative, because this is the state that standardised Chinese writing, weights, measures, currency, and axle-wits, and whose legal system we know mainly from fragments recovered elsewhere, there may be maps. There is presumably an accurate representation of how the Chin understood their own territory and their own sky. There is the answer to what actually

killed the emperor, a man who spent his final years hunting for an elixir of immortality, and who was quite possibly poisoned by the mercury-based preparations his own alchemists fed him, which would make the rivers under that mound the single most on the nose piece of symbolism in the ancient world. And there is the closing of a two thousand-year-old question about whether the most famous description of a tomb ever written was repartage or fantasy. What we gain by waiting is the chance to get all of that once, properly, instead of half of it now, and a permanent apology later. There is no second attempt. A tomb can only be opened for the first time on one day in history, and whoever is standing there on that day gets to be either the person who recovered it or the person who ruined it.

For now, the mound stays a hill with trees on it. Tourists photograph the soldiers a kilometer and a half away, and the emperor keeps his appointment with an empire made of clay, mercury and patience. That is what happens when you know exactly what you have and decide not to touch it. The next object on this list is the opposite situation in every possible way. It was hauled out of the sea by accident, thrown into a museum storeroom, ignored as a lump of corroded scrap, and it turned out to contain a machine that had no business existing for another thousand years. If you win a $3,000 a month payday for life, what would you feel free to do? Maybe take a long weekend, every weekend, or try a bunch of new hobbies.

Would you feel free to upgrade and listen ad-free? Don't worry, we get it. Every $20 ticket could win you $3,000 a month for life, and, supports life-saving cancer research at the Princess Margaret. Feel free to buy your payday-for-life ticket today. Raffle number RAF-157-3916. Please play responsibly. In the autumn of 1900, a boat full of sponge divers from the island of Simi was working its way home across the Aegean when the weather turned.

Their captain, Demetrius Contos, put in at Antichythra, a tiny scrap of land between Crete and the Peloponnese with cliffs, goats, and very little else. While waiting out the wind, they did what sponge divers do, which is go-looking for sponges. The diver named Elias Stadiartis went down about 45 metres in a canvas suit and a copper helmet and came back up in a hurry, describing a seabed covered in dead bodies, rotting men, horses. The captain, showing admirable scepticism about a report of an underwater cavalry massacre, went down himself and came back holding a bronze arm. What the crew had stumbled onto was the wreck of a ship that had gone down around seven

to 60 years before the Common Era, loaded with cargo that was almost certainly headed for Rome, and almost certainly not purchased with anyone's consent. Roman appetite for Greek art in this period was enormous, and the acquisition methods were flexible. The site produced monumental bronze statues, marble figures half eaten by marine organisms, glass vessels, jewelry, furniture fittings, and coins that helped date the loss. The Greek government organised a proper salvage with the Navy. and over the following months the divers worked at depth that were flatly dangerous with the equipment of the time. The cost was real. One diver died, and others were left paralysed by decompression sickness,

back when nobody fully understood why coming up fast turns a person's blood into a shaken bottle of soda. Everything went to the National Archaeological Museum in Athens, where the spectacular pieces got the attention, and the rest sat in storage, including a shapeless lump of corroded bronze and rotted wood that looked exactly like what you would expect from 2,000 years of seawater. It looked like garbage. It got treated like garbage. Then in the spring of 1902, an archaeologist named Valerius Stice was looking at the material and noticed that the lump had cracked as it dried out, and inside the crack was a gear wheel, with teeth, cut into bronze, that should not exist. Not in a Greek shipwreck from the first

century before the common era. The immediate reaction in the field was a mix of fascination and deep suspicion, with a lot of people assuming a later object had somehow ended up mixed into the deposit, because a geared machine simply did not fit into anybody's understanding of ancient technology. Broadly speaking, the ancient Greeks were credited with mathematics, philosophy, and astronomy of a very high order, and with mechanical toys and siege engines, but not with precision instrument making. For the next half century, the object was a curiosity that nobody could resolve, because the only way to see inside a corroded bronze lump with destroying it, is to look through it, and the tools for that had to be invented first.

The breakthrough phase started in the 1950s with Derek de Sola Price, a British physicist turned historian of science working at Yale, who spent 20 years on the problem. Working with the Greek nuclear physicist Karolampos Caracalos, he used gamma-ray radiography to image the interior, counted gears, counted teeth, and in 1974 published a study that laid out the basic architecture. He identified a gear with 127 teeth and recognized what that number meant, which is that this device was tracking the motion of the moon, because 127 is half of 254, and 254 sedereal lunar months fit neatly into 19 solar years. That is the metonic cycle, the relationship that lets you keep a lunar calendar synchronized

with the solar year, and building it into brass means the maker was encoding astronomical theory as a physical ratio. the big picture right and a lot of the details wrong, which is the standard fate of anyone who goes first. The next major push came from Michael Wright, a curator at the Science Museum in London, who used linear tomography to image the fragments in slices, corrected a large part of the gearing scheme, and did the thing that separates plausible reconstruction from theory, which is build working physical models and see whether they actually turn. Then in 2005, a joint research project brought in equipment that changed everything. Two technologies mattered. The first was a surface

imaging technique that photographs an object under light from many different angles and then combines the images so you can relight the surface interactively, which makes faint inscriptions leap out of corroded metal. The second was a custom-built micro-focus computed tomography machine, an eight-ton monster hauled to Athens specifically for this, capable of resolving the internal structure of the fragments in three dimensions at very high detail. The results were spectacular. The team could read text that had been invisible for two millennia, roughly tripling the known inscriptions, and could count teeth on gears buried inside solid corrosion. So here is what the machine actually is. There are 82 surviving fragments, most of them small. Between them, around 30 gears survive,

and reconstructions require a total original gear train of at least 37, with some proposals running considerably higher. The gears are bronze, hand cut, with triangular teeth, and the smallest teeth are about a millimeter across. The whole thing was housed in a wooden case roughly the size of a large book, with bronze plates front and back covered in explanatory text, functioning as an engraved instruction manual for the user. You operated it with a hand crank on the side, turning the crank advanced time, and every dial on the machine moved to show what the sky was doing on that date. on the front there was a large dial with two concentric scales, one marked with the 12 signs of the zodiac in 360 degrees, and one marked with the 365 days of the Egyptian calendar. That outer calendar ring was designed to be removable and repositionable, which is how the maker handled the leap year problem, since the Egyptian calendar drifts by a day every four years.

Pointers on the front showed the position of the sun and the moon against the zodiac, and there was a small rotating ball, silver, on one side and dark on the other that displayed the current phase of the moon. There is strong evidence, and detailed reconstructions have been built on it, that the front also displayed the positions of the five planets known to antiquity. On the back were two large spiral dials. The upper spiral runs five turns and is divided into 235 cells, one for each lunar month in the 19-year cycle, so the pointer follows the spiral around like a record groove tracking the calendar. Attached to it a For two smaller dials, one handles the Callipic cycle, a 76-year refinement of the 19-year cycle that improves accuracy.

The other, and this is my favourite part of the entire object, is a Games dial, a four-year cycle listing the great athletic and religious festivals of the Greek world, including the Olympics. Somebody built a precision astronomical computer and then thought, yes, but people are also going to want to know when the games are. The lower spiral is the eclipse predictor. It runs four turns divided into 223 months, the Saros cycle, which is the period after which the geometry of sun, moon and earth, repeats closely enough that eclipses recur in a predictable pattern. Individual cells carry small inscriptions, glyphs, that specify whether an eclipse is

solar or lunar, at what hour it is expected, and additional characteristics. And attached to that is a dial for the Exeligmos cycle, which is three Saros periods, included because each Saros as repetition is offset by about eight hours, so after three of them the shift comes back to a whole number of days, and the correction can be applied. This device did not just say an eclipse is coming, it told you roughly when in the day to look, and it carried its own correction factor, and then there is the detail that elevates the whole thing from very clever to genuinely startling. The moon does not move at a constant speed across the sky. Its orbit is elliptical, so it speeds up and slows down over the course of the month,

an effect Greek astronomers had studied and modelled mathematically. The builder of this device translated that theory into mechanics using a pin and slot arrangement, where one gear carries a pin that sits in a slot on another gear mounted slightly off centre, so as the pair rotates, the driven gear alternately races ahead and falls behind. The output is a lunar pointer that varies its speed exactly the way the real moon does.

That is not a calculation. That is an analog simulation of a variable rate motion, built out of bronze by hand, using a mechanism that a modern engineer would recognize instantly as a legitimate solution to the problem. So who made it? This is the part where the honest answer is that we do not know, but we have good circumstantial material. The month names inscribed on the calendar dial are Corinthian in type, which points toward the Greek Northwest, the Corinthian colonies, or possibly Syracuse in Sicily. Syracuse is an interesting address, because it is where Archimedes lived and died.

Roman sources, particularly Cicero, mention a device made by Archimedes that modelled the motions of the sun, moon, and planets, which was carried off to Rome after the fall of the city, and was still being shown off generations later as a marvel. Cicero also refers to a similar instrument built by Posidonius of Rhodes. So the written record does contain a tradition of geared astronomical models. It just does not contain a single surviving example other than this one. Dating is contested. One approach reads the eclipse predictions on the back and works backward to find the starting date that best fits, producing an argument for around 205 before the Common Era, which would place it uncomfortably close to Archimedes himself.

Other analyses, based on the shipwreck date and on the astronomical assumptions in the device, favour somewhere in the range of 150 to 100 before the Common Era. Either way, this is a Hellenistic object, and either way it is older than most people's intuition allows, and now the actual mystery of this chapter, which is not the machine, it is the silence around it. Comparable mechanical sophistication does not reappear in Europe, until the astronomical clocks of the 14th century, machines like the Astrarium built by Giovanni de Dondi over 16 years, which modelled planetary motion with dozens of gears. That is a gap of roughly 1,400 years between this object.

and the next European device that can do what it does. A gap that large is not a mystery about genius. It is a mystery about transmission. Skills like this do not appear once, a maker capable of cutting triangular teeth a millimeter across, laying out compound gear trains, and encoding astronomical cycles into ratios is not somebody's first attempt.

Behind this object, there has to be a workshop tradition, apprentices, earlier and cruder versions, later and improved versions, and a customer base wealthy enough to pay for them. None of that survives. Not one other example, not one clear technical description in the surviving literature explaining how these were made.

There are two explanations and the truth is probably a mixture. The first is simple and depressing, and it is bronze. Bronze is valuable and infinitely recyclable. An old machine that no longer works, whose maker is dead and whose repair knowledge is lost, is not a museum piece to the people living with it. It is metal. Across antiquity in the entire medieval period, bronze objects were systematically melted down

and turned into new bronze objects, which is why the number of surviving large Greek bronze statues is tiny compared to the number we know existed, and why a disproportionate share of the ones we do have came out of shipwrecks. The sea, in other words, is the only archive that could not be raided. This device survived specifically because it sank before anyone could sensibly reuse it, which means our entire knowledge of ancient precision mechanics rests on a single storm and one bad night of navigation. The second explanation is that the tradition genuinely was tiny.

These were not consumer goods. Each one would have taken a highly skilled maker an enormous amount of time, cost a fortune, and served a market consisting of extremely rich people who wanted a demonstration of cosmic order sitting on a table. A technology with a customer base of a few dozen individuals across the Mediterranean is one bad century away from extinction, and the Mediterranean of this period was rich in bad centuries. It is also worth killing off the version of this story that gets told loudest, which is that the knowledge vanished completely into a black hole. It did not entirely.

There is a Byzantine-geared sundial calendar surviving in fragments from around a thousand years ago, much simpler but unmistakably part of the same idea. The scholar al-Baruni describes a geared calendrical device in the Islamic world in the same era. Islamic astronomers built sophisticated astrolabes, some of them geared, and that tradition fed directly into medieval Europe. So the thread was never quite cut. It got very thin, and it lost the top end of its capability for well over a millennium, but the concept of gearing the heavens survived in the Eastern Mediterranean and the Islamic world, and eventually resurfaced in the cathedral clocks of Europe. What did not survive is whatever the peak of the

Hellenistic version looked like. And there is one more twist that I love, because it shows how much is still coming out of these fragments. Recent work applying statistical methods borrowed from gravitational wave astronomy to the spacing of surviving holes in the calendar ring has argued that the ring originally held 354 holes rather than 365, which would mean it tracked a lunar year rather than a solar one. That is a 120-year-old object, examined by generations of specialists, and researchers are still changing our understanding of what its individual components did, using techniques developed for detecting collisions between black holes. The wreck itself is not finished either. Modern expeditions have returned to the site repeatedly with rebreathers and robotic

assistance, and have recovered more statuary, ship timbers and small finds. The ship was large, and it broke up across a slope, and nobody is willing to say the seabed has given up everything. At least we can read that one. We know what its dials meant, what its numbers referred to, and what its inscriptions say, because it was written in a language we have, and about a subject we understand. The last two things on this list do not offer that courtesy. They are covered in writing, and nobody alive can read a single line of it. On the southern side of Crete, above the Masara plain, sit the ruins of Feistos.

one of the great palace complexes of Minoan civilization, second only to Nosos in fame, and considerably nicer in terms of the view. In the summer of 1908, an Italian archaeologist named Luigi Pernier was working through a set of rooms in a basement level north of the main palace, digging through dark earth mixed with ash, and on the 3rd of July he lifted out a disk of fired clay about 16 centimetres across, roughly the size of a dessert plate, almost intact. Both faces were covered in symbols, arranged in a spiral running from the rim toward the centre, divided into groups by short vertical lines. There were 241 sign impressions in total, sorted into 61 groups, drawn from an inventory of 45 distinct symbols,

and the symbols are gorgeous, a human head wearing what looks like a spiked crest or mohawk. A walking man, a running figure, a shield with dots, a ship, a fish, an eagle. a flower, a branch of something, a beehive, a hide, an ox-hoof. This is not a dry accounting script full of strokes and wedges. It is a parade of tiny pictures, and it looks like it should be readable by anyone with a bit of patience and imagination. It is not, it has never been. And a century and change later, it is arguably the most famous, unreadable object on the planet. Before getting into why, there is a technical detail that deserves top billing, because

most summaries bury it, and it is the single most remarkable fact about the thing. The signs were not scratched or carved into the clay. They were stamped. Each symbol was impressed into the soft surface with a prepared punch, one strike per sign, and the same punches were used repeatedly across both faces. You can see it in the impressions, which are identical for repeated signs in a way that hand carving cannot produce. Sit with what that implies. Somebody manufactured a set of at least 45 individual stamps, presumably, from a hard material, each carrying a specific symbol, and use them to compose a text by pressing them into clay and sequence. That is movable type. It is not a printing press. There is no ink

and no paper and no mechanism, but the core concept, a reusable set of individual character punches assembled into arbitrary text, is exactly the idea that Johannes Gutenberg would build an industry on somewhere around 3,000 years later. This object, dating to the Bronze Age, is the earliest known example of the principle, and it exists. exists in a total print run of one. Because nobody makes 45 custom punches to write a single document. That is an absurd amount of preparation for one clay disc. Manufacturing a stamp set is an investment in repeat production. The existence of this object logically requires that other stamp texts were made, probably many, and that a workshop somewhere on Crete had the equipment and the

habit. Not one of those other texts has ever been found. Same structural problem as the geared machine from the last chapter, in a different medium and one. 1,500 years earlier, and no more comforting for the repetition. So why can nobody read it? The answer is a numbers problem, and it is worth walking through because it explains what decipherment actually requires, as opposed to what people imagine it requires. Start with the sign count. Writing systems fall into rough categories based on how many distinct characters they use. An alphabet, where signs represent individual consonant and vowel sounds, usually run somewhere between 20 and 40 characters. A syllabary, where each sign represents a whole syllable

like k, k, k, k, k, k, k, k, typically needs 50 to 100. A logographic system where signs stand for whole words or concepts needs hundreds or thousands. This disc has 45 distinct signs, which lands right in the awkward zone, most probably a syllabary, possibly with some logograms mixed in, which is what several other Aegean scripts did. Now the corpus size. This is the killer. Decipherment is fundamentally a statistical exercise. You look for repeated sequences, positional patterns, sign frequency,

and the way words change at their edges, and you use those regularities to reverse engineer the structure of the language underneath. To do that, you need volume, you need thousands of examples, ideally in similar formats, ideally with some content you can guess at, like inventories where the same commodity appears

with different numbers after it. The entire known text of this script is 241 signs. That is not a corpus. That is a paragraph. It is enough to see that some sequences repeat and that certain signs cluster at the beginnings or ends of groups and it is nowhere near enough to build a system on. Compare that with the great success story in this field, Linear B.

When Michael Ventris cracked it in 1952, showing that the script recorded an early form of Greek, he had two colossal advantages. First, thousands of clay tablets from multiple sites with repeated formats, lists and totals. Second, an underlying language that turned out to be one we already knew. He still needed the accumulated work of others, in particular the analytical groundwork of Alice Cobra,

who spent years identifying sign patterns without ever getting to see the payoff since she died before the breakthrough. The disc has neither advantage. There is no archive. and there is no known language behind it. Nobody can even say with confidence whether the language it records is Minoan

or something else entirely brought in from Anatolia or the wider Aegean, because we do not have the Minoan language either. Linear A, the script the Minoans used for their own tongue, remains undeciphered too, with a substantially larger corpus and a century of professional attention. The relationship between the disc is script

and Linear A has never been demonstrated. They share the island and roughly the era, and that is about it. And there is one more layer of difficulty that people underestimate, which is that we do not even know for certain which way to read it. The current consensus is that you start at the outer rim and spiral inward, based mainly on evidence of stamping order, where later impressions overlap earlier ones,

and on the way sign groups crowd near the centre, as the writer ran out of room, which is a mistake every human being who has ever handwritten a birthday card will recognise instantly. But this remains an inference, not a certainty, and a text you might be reading backwards is a rough starting position, none of which has stopped people from trying. The number of published decipherments runs into the many dozens, and the range is spectacular. It has been declared a prayer, a hymn to a mother goddess,

a calendar tracking lunar cycles, a court list, a record of soldiers, a narrative of a journey, a musical notation, a board game, a geometric theorem, a legal document, and a message in essentially every language spoken anywhere near the Mediterranean, including several that had not been invented yet at the time it was made. One widely publicised modern attempt, associated with the researcher Gareth Owens working with a phonetics specialist, assign sound values borrowed from Linear B to visually similar signs, and produces a reading interpreted as a religious text mentioning a goddess and a pregnant mother. It has had a lot of press, it has not been accepted, and the objection is basic, assigning phonetic values by visual resemblance to a different script from a different period

is a hypothesis, not a method. And once you have made those assumptions, you can produce sentences from almost anything. The uncomfortable truth underneath the whole industry is that with 241 signs and no anchor, an interpretation can be internally consistent and still be entirely fictional. There is no way to check it. That is why professional Aegean scholars mostly do not spend their careers on this, and why the field is dominated by outsiders, retirees, and the extremely determined. It is the crossword puzzle from hell, with no answer key and no guarantee there is an answer. Then there is the other storyline, the one that turns this from an academic puzzle, into a genuinely spicy dispute. Some people think it is a fake. The accusation went mainstream in 2008.

On the 100th anniversary of the discovery, when the American archaeologist and antiquities dealer Jerome Eisenberg published an argument that the disc was manufactured by Pernier himself as a career move. The context he pointed to is real. Cretan archaeology at the start of the 20th century was fiercely competitive. Arthur Evans was at Nossus producing spectacular results and enormous publicity. Italian teams were working the south of the island and a young archaeologist who produced a unique, beautiful headline-generating object would improve his position substantially. The technical arguments made were roughly these. The object is entirely unique, with no parallels in the archaeological record, which is exactly what a forgery looks like. The edges are unusually clean and regular for a

bronze-age clay object of this type. The fine circumstances, as recorded, are not as precise as we would want, and the excavation records leave room for doubt about the exact context. And a stamped document with no other stamped documents anywhere is at minimum convenient. The counter-arguments are stronger, and the field has largely stayed unconvinced by the forgery case. The most important piece of evidence is a bronze double axe found in a cave at Arcalochori in Crete in the 1930s, decades after the disc was found, carrying inscribed signs that resemble several of the disc as symbols. If the disc were an invention, an object discovered later by different people, at a different site, should not share its symbol set.

There are also a handful of other possible sign parallels on Cretan material. Beyond that, forging a script is a different order of difficulty from forging an artifact. A convincing, invented writing system has to display the statistical behavior of real language, and analyses of sign frequency and group structure on the disc keep coming back looking like an actual script rather than a decorative arrangement. But here is the part that keeps the argument alive, and it is a procedural point rather than a scientific one. The dispute could be substantially narrowed by physical testing. Aluminoussance dating measures how long it has been since fired ceramic was last heated,

and while it is not perfectly precise, it would distinguish between an object fired in the Bronze Age and one fired in the early 1900s, which is a gap of roughly 3,500 years and not exactly a subtle call. Request to perform such testing on the disc have historically been declined by the Museum in Heraclian, where it is the Star Exhibit. The stated reasons are the reasons any museum gives, and they are not unreasonable in themselves. The technique requires taking a sample, which means removing material from a unique object. The disc is fragile, iconic and irreplaceable. The institution is not obliged to damage a national treasure to satisfy a claim most specialists consider weak. Cretan authorities also have a long institutional memory of foreign researchers

making demands about Cretan heritage, and are not especially motivated to accommodate an accusation that a foreigner faked their most famous artifact. The effect, however, is that the one test that could close the question has not been run in a way that satisfies everyone, which guarantees the accusation stays in circulation forever. It is a small case study in how heritage protection and open inquiry can pull in opposite directions with nobody behaving badly. Whatever it is, the disc has had an afterlife that is somewhere between charming and absurd. It is the unofficial emblem of Crete. jewelry, ceramics, souvenir plates, wine labels, hotel signage, and an enormous quantity of tourist merchandise. There is a decent chance that more copies of this text exist today, printed on

refrigerator magnets, than were ever produced by the civilization that made it. The world's oldest example of typography finally achieved mass production, roughly 3,500 years late, in gift shop form, and still nobody knows what it says? So what would actually crack it? The requirements are clear and unromantic. More text is the main one. A cache of stamped clay documents from a Minoan storeroom would transform this overnight, because the pattern analysis that fails on 241 signs starts working at a few thousand. A bilingual would do it instantly, some object carrying the same content in this script and in one we can read, which is how Egyptian hieroglyphs fell to the Rosetta Stone. Even a securely dated second disc from a clean excavation would settle the forgery question and give the statistics

something to bite on, and the honest possibility that nobody likes is that none of that exists to be found. Crete has been dug over intensively for more than a century. If stamped documents were common, we should probably have more of them by now. The plausible explanation is that this technique was used on materials that do not survive, or that it belonged to a very small tradition that ended quickly, or that the surviving examples are still sitting in unexcavated deposits, or in a museum drawer somewhere labeled as decorated pottery fragments by someone in 1911 who was in a hurry. That last one is not a joke, by the way. Given that the most important gearwork of antiquity spent decades in a storeroom being ignored, the odds that something significant is currently

mislabeled in a museum basement are not low, which leaves us with a disc of clay, 45 symbols, a technology that should not exist for another three millennia, an argument about whether it is even real, and a message from a Bronze Age cretan that has survived. perfectly intact for 3,500 years, while being completely, permanently silent. At least in that case, the silence has an excuse. The civilization that wrote it collapsed, its language died, and everything around it was lost. The final item on this list has no such alibi. It was made in Europe, in the age of paper and universities, by someone who was almost certainly not trying to be mysterious, and it has defeated every professional codebreaker who has ever picked it up.

In 1912, a book dealer named Wilfred Voynich was doing business in Italy. He was an interesting person in his own right, a Polish-born former revolutionary who had done time in Siberia, escaped, ended up in London, and reinvented himself as one of the most successful antiquarian booksellers in Europe, which is a career arc you do not see much anymore. That year he was at the Villa Mondragon near Frascati, a Jesuit property where the order was quietly selling off part of a library to raise money,

and he bought a chest of manuscripts. One of them was unlike anything he had ever handled, and he had handled a lot. It is a codex of roughly 240 parchment pages, modest in size, bound in later limp vellum with no title, no author, no date, and no colophon. Every page is written in a flowing script that does not match any known writing system on earth,

using an alphabet of somewhere between 20 and 30 distinct characters that appears nowhere else in the historical record, and it is illustrated, extensively, in colour, with pictures of plants that do not exist. That is the Voynich manuscript, currently held at the Beecher Rare Book and Manuscript Library at Yale University, under the catalogue number MS408, fully digitised, freely available online to anyone with an internet connection, and completely unread. Start with the physical object, because that part is settled, and it eliminates a lot of theories that used to be popular. In 2009, samples of the parchment were radiocarbon dated at the University of Arizona,

and the result came back with a range of 1404 to 1438 at high confidence. That is the first half of the 15th century, before Gutenberg, in the world of manuscript books and university faculties and monastic libraries. Separately, materials analysis of the inks and pigments found iron, gall ink and colorants consistent with the period, applied in a way consistent with the writing being done not long after the parchment was prepared. No anachronistic pigments, nothing that screamed later addition. That combination kills the most economically attractive theory, which is that the whole thing is a modern forgery cooked up to be sold. It also constrains a very fun older theory,

which I will come back to, and it means that whatever this is, somebody in early 15th century Europe sat down with expensive materials and produced 240 pages of it. Now the contents, which are organised clearly enough that researchers have divided the manuscript into sections, based purely on what the illustrations show without understanding a single word, the largest is the herbal section, over a hundred pages, each dominated by a drawing of a plant with text wrapped around it, exactly like the herbals that were a standard genre of medieval and early modern book production. The problem is the plants. A handful have been tentatively matched to real species. The great majority cannot be identified, and many of them look like composites, a root from one plant, leaves from another,

a flower that belongs to nothing in particular, assembled into a species that has never grown anywhere. If you handed this to a botanist without context, they would tell you the illustrator had either never seen the plants they were drawing or was working from a description, or from another book, or from imagination. Then there is an astronomical section, with circular diagrams, star-like elements, and a set of zodiac pages showing the familiar signs surrounded by rings of small nude female figures, many of them standing in what look like barrels or tubs, each holding a star. Several of those zodiac pages carry month names written in a different hand, in a dialect that reads as something close to oxytan, or a related southern French form,

which is one of the very few pieces of readable writing in the entire book, and is almost certainly a later owner's annotation rather than part of the original. Then comes the section usually called biological or balneological, and this is the one that makes people stop and stare, page after page of small naked women in green liquid, in pools and basins, connected by an elaborate network of pipes, tubes, and vessels that look organic and plumbing related at the same time.

It is bizarre, it is strangely charming, and the leading sober interpretation is that it relates to bathing, balneology, and possibly women's medicine and reproduction, which were genuine subjects of medieval medical writing. Whether it is that, or an alchemical allegory, or something else entirely nobody can confirm, because the captions are in the same unreadable script as everything else. There is a cosmological section including a spectacular fold-out of nine connected circular diagrams that has been interpreted as a map, a cosmology, an alchemical scheme, or a set of islands depending on who is looking. There is a pharmaceutical section with drawings of jars and containers

alongside plant parts, which looks exactly like an apothecary's reference. And there is a final section of dense text organised into short paragraphs, each marked with a star in the margin, which everybody calls the recipe section, for the simple reason that it is laid out like one. So the structure of the book is completely legible. It reads as a technical reference work on plants, stars, bodies and preparations. It has a table of contents feel to it without a table of contents. The person who made it was following the conventions of a recognisable genre. The words are what will not yield. Here is where things get properly strange, because the text does not behave like nonsense, and it does not behave like a normal European language either. Analyze statistically it shows

the hallmarks of real writing. Word frequencies follow the kind of distribution you see in natural languages, where a few words are extremely common and most are rare. Words have internal structure, with recognisable prefixes, middles, and endings, and the components combine according to consistent rules, meaning there is a grammar of some kind operating at the level of word construction. Certain characters appear almost exclusively at the beginnings of words, others at the ends, others only at the start of a line. There is order everywhere. At the same time, the text does things no natural language does. It is more repetitive than any known language, with the same word sometimes appearing two or three times in a row, which happens in real writing occasionally and not like this.

The entropy of the text, which is a measure of unpredictability, is unusually low, meaning that if you know one character you can guess the next one better than you should be able to in ordinary prose. Words that are similar to each other cluster together more than they should, and the line seems to function as a unit in a way that is uncommon in normal writing. With position on the line influencing which characters appear, there is another wrinkle. A cryptanalyst named Prescott Currier noticed in the 1970s that the manuscript is not statistically uniform. Different sections behave differently, enough to suggest two distinct systems or dialects, which are now conventionally called Currier A and B, apparently.

associated with different hands. So the book may be the work of more than one person, or one person using more than one system, which is not what you would expect from a simple invented gibberish. The list of people who have failed on this thing is where it stops being a curiosity and starts being a legend. The most notable is William Friedman, arguably the most important cryptologist in American history, the man who led the team that broke the Japanese diplomatic cipher machine the Americans called Purple, an achievement that shaped the Pacific War. Friedman was obsessed with this manuscript for decades. He organised study groups, including one during the Second World War made up of colleagues who cracked enemy codes for a living, and later

another with computing assistance. They applied the full arsenal of professional cryptanalysis to it. They got nowhere. Friedman's eventual conclusion, which he famously encoded as an anagram so he could claim priority without committing publicly, was that the text is not an encryption of an existing language at all, but an artificial language constructed on some system, which was a very live intellectual project in later centuries, and not an absurd thing to attempt. Others crashed harder. In 1921, a Pennsylvania philosophy professor named William Newbold announced a decipherment based on the claim that microscopic markings within the penstrokes concealed a shorthand, and that the manuscript proved a 13th century friar had used a telescope and a microscope centuries early. It was a sensation.

It was also wrong, comprehensively demolished a few years later when it was demonstrated that the microscopic marks were simply the cracking of dried ink on parchment. Newbold had been reading the physical decay of the ink as a secret message, which is the single most tragic thing that can happen to a researcher. The parade has continued ever since. It has been declared a cathar text, a Hebrew text scrambled by anagram and identified with machine learning, a shorthand Latin medical compendium, a Turkish dialect, and most recently and most loudly a proto-romance language, an announcement in 2019 that generated headlines worldwide and was rejected almost immediately by medievalists and linguists, to the point where the publisher

issued a statement distancing itself from the claim. The pattern is consistent. Someone produces a method, applies it, generates a translation, and the translation turns out to be flexible enough that the method could produce almost any text you wanted from the same source, which leaves the hoax hypothesis, and it deserves a fair hearing because it is the most serious rival to the it is a real language position. In 2004, Gordon Rugg, a researcher at Keel University with a background in both psychology and computing, demonstrated that you can generate text with Voynich, like properties using a technique available in the 16th century. Take a grid of syllables, prefixes, and suffixes. Take a card with holes cut in it, a device known as a card-un-grill, place it over the grid, read off

whatever shows through the holes, slide the card, repeat. The output is meaningless, but it has structure, repetition, and internal consistency, and it looks disturbingly like the manuscript. Rugg's point was not that this is definitely how it was made, but that a hoax capable of producing this text was technically feasible without modern tools, which removes the argument that no medieval person could have faked it. The counterattack came from information theory. Analysis of the text have found that word distributions are not uniform across the book, that particular words concentrate in particular sections, and that the pattern of concentration resembles the way content words behave in real documents about real subjects. In other words,

the vocabulary seems to track the topic, with plant-related sections using their own recurring terms. Random generation should not do that, at least not without a lot of extra effort by a forger who cared deeply about statistical realism 500 years before anyone invented the field. So the honest summary is that both camps have real evidence, and the manuscript keeps refusing to fully satisfy either. Then there is the provenance, which is a genuinely great story, and also explains where the hoax idea came from originally. The earliest documented owner takes us to Prague around 1600. A signature, partly erased and recovered under ultraviolet light, appears on the first page belonging to Jacobus Horsicki de Tepernick, an imperial pharmacist and herbalist in the service of the

Holy Roman Emperor Rudolf II. The book then passed to an alchemist named Georg Beresh, who was so frustrated by it that he sent copied pages to the Jesuit polymath Athanasius Kircher in Rome, asking for help and got nothing useful. After Beresh died, the manuscript went to his friend Johannes Marcus Marcy, a physician and scholar who sent it to Kircher in the 1660s, along with a letter that survives to this day. That letter is the single most important document about the book and it is the source of two claims. First, that Rudolf II acquired the manuscript for 600 ducats, an eye-watering sum. Second, that the emperor believed it was the work of Roger Bacon, the 13th century English friar and natural philosopher, a belief the letter reports second-hand

and does not endorse. Now, Rudolf II is a critical character here, because he was the most enthusiastic patron of the strange and the esoteric in Europe. His court in Prague was a magnet for astronomers, alchemists, artists, and a substantial population of charming men with unverifiable claims. Among those who spent time in that orbit were the English scholar John D. and his associate Edward Kelly, a figure with a documented history of persuading wealthy people to fund things that did not exist. So a theory took root that someone in that circle manufactured a beautiful, impressive, utterly meaningless book and sold it to an emperor known to pay generously for wonders. It is a great theory. It has one heavy problem, which is that the parchment was made around

150 years before Kelly arrived in Prague. Not impossible since blank medieval parchment could survive and be reused, but you would need a large stock of consistent leaves from the same period, and the ink analysis pushes against a long gap between preparation and writing. After Kircher, the trail goes quiet inside the Jesuit collections for two and a half centuries until the sale at the villa, and from there to a New York dealer, and finally, in 1969, as a gift to Yale. Which brings me to the detail I keep coming back to, because it is what makes this the right place to finish. Look at the handwriting. Whoever wrote this wrote fluently. The script flows. There are essentially no corrections, no hesitations, no crossed out passages, no visible struggle. Nobody laboriously

enciphering text letter by letter writes like that, and nobody inventing gibberish as they go, writes like that, either without an enormous amount of practice. The person holding that pen either knew the system cold or was copying from something that already existed, and they kept it up for 240 pages, illustrated it in colour, and used expensive parchment to do it. That is not the behaviour of someone playing a prank on the world. Pranks are short. That is the behaviour of someone producing a working reference book, for use, in a system that made sense to them, and presumably to at least one other person, because the effort of writing a book nobody can read is a very strange way to spend a year of your life. And here is the thing that quietly connects everything we have looked

at. In every case on this list, the object survived and the context did not. The rock-cut refuges outlived the memory of who was hunting the people inside them. The stone city on the reef outlived the knowledge of how the columns crossed the water. The bronze calculator outlived every other machine its workshop ever built. The stamped clay disc outlived its language. The stamped clay disc outlived its language. And this manuscript outlived whatever key made it readable, which may have been nothing more than a single sheet of correspondences, or a teacher, or a shared understanding between two people who saw no reason to write it down, because they were not planning on dying. That is the actual mechanism behind almost every historical mystery worth the name. Not a conspiracy,

not a lost superior civilization, not visitors from anywhere. Just the ordinary, relentless fact that context is fragile and objects are durable, and the two do not travel through time at the same speed. Knowledge lives in people, and people are terrible at anticipating which of the things they know will turn out to matter, which is why the pattern held all the way through this list, a farmer looking for water, a crew-hunting sponges, villagers hoping for fish, a diver scouting sharks, a man swinging a hammer at an inconvenient wall in his own house, wanting nothing more than a slightly bigger room, and instead opening a hole into a city that had been holding its breath under his floor for a thousand years. None of them were looking

for history. It was simply there, waiting, the way it is currently waiting under an unknown number of fields, floors, ponds, and quiet hillsides, in places where nobody has yet had a reason to dig. So the only real question left is which one goes next, and who is going to be holding the hammer? If you made it this far, you're exactly the kind of person this channel is built for. Subscribe so you're here when the next wall comes down, and I will see you in the next one.