The video rests on a quiet fact that deserves a whole essay of its own: we can date the Maya collapse almost to the year because the Maya wrote it down. Their monuments carry dates in a calendar accurate to the day, and their scribes painted books in a true writing system — one of only a handful ever invented from scratch anywhere on Earth. When I said in the video that “the dates stop,” I could only say it because, a thousand years later, we can read those dates.

But for most of the last five centuries, we couldn’t. And the story of how we regained the ability is one of the strangest in the history of knowledge, because it turns on a single man who both destroyed the Maya’s written world and, without meaning to, saved the means of recovering it.

His name was Diego de Landa, a Franciscan friar in the Yucatán. On the 12th of July, 1562, at the town of Maní, he staged an auto-da-fé — a public burning. Convinced the Maya books were, in his words, full of “superstition and the devil’s falsehoods,” he threw the accumulated written memory of a civilization onto a fire. By his own account he burned twenty-seven hieroglyphic codices and some five thousand sacred images, while torturing the Maya who had kept them. It is one of the most complete acts of cultural erasure ever recorded. Of the countless books the Maya once produced, only three or four survive today. Everything else is ash.

Here is the paradox. Years later, back in Spain and defending his conduct, Landa wrote a long account of Maya life — the Relación de las cosas de Yucatán. In it, trying to prove he understood the people he had brutalized, he asked a Maya scribe to show him the “alphabet,” and he wrote it down: a set of glyphs matched to Spanish letter-sounds. Landa thought he was recording an alphabet. He was wrong — but the mistake was priceless.

For four hundred years, scholars stared at Landa’s garbled glyph list and got nowhere, because they trusted his framing: they too assumed the Maya wrote in an alphabet, and it never worked. Then, in 1952, a young Soviet linguist named Yuri Knorozov, working in Leningrad from reproductions, at the height of the Cold War and largely cut off from Western Mayanists, looked at the same list and saw what everyone had missed. Landa’s signs were not letters. They were syllables. The Maya script was logosyllabic — part sound, part word-picture — and Landa’s flawed “alphabet” was, correctly read, the Rosetta key.

Knorozov was ridiculed at first, especially by the reigning Western authority, who dismissed the breakthrough partly because it came from the Soviet Union. He was right and they were wrong. Today, on the foundation he laid, scholars can read more than ninety percent of Maya writing — which is precisely why we can trace the collapse city by city, date by date.

So the next time someone calls the Maya a “lost” civilization, remember the shape of it: a friar burned their library to erase them, accidentally preserved the one page that would let us hear them again, and a linguist an ocean and four centuries away used it to give them back their own voice. The Maya wrote their history down. We are only now finishing the work of learning to read it.

Sources

• “The Auto-da-Fé of Maní: Diego de Landa and the Destruction of Maya Literature” — Mayan.org — https://mayan.org/mysteries/auto-da-fe-mani/

• “Diego de Landa: The Friar Who Burned — and Saved — Maya Knowledge” — Mayan.org — https://mayan.org/civilization/landa/

• “Burning the Maya Books: The 1562 Tragedy at Maní” — Popular Archaeology — https://popular-archaeology.com/article/burning-the-maya-books-the-1562-tragedy-at-mani/

• Maya hieroglyphic writing & decipherment (Knorozov) — Penn Museum, Expedition — https://www.penn.museum/sites/expedition/maya-hieroglyphs/

Fact-check note: Landa’s own account gives 27 codices and ~5,000 images burned at Maní (12 July 1562); only 3–4 Maya codices survive. Knorozov’s 1952 phonetic/syllabic breakthrough using Landa’s “alphabet” is well established; “>90% deciphered” is the standard current figure. The often-repeated tale that Knorozov rescued a Maya codex reproduction from a burning Berlin library in 1945 is disputed/apocryphal and is deliberately NOT included here.

The video leans on one witness: Procopius, the emperor’s historian, who watched Constantinople empty from inside its walls. He is the source almost everyone quotes. But he was not the only person who wrote down what he saw in 542 — and the other account is, in some ways, the more harrowing of the two, because that man was not safe behind city walls. He was out on the roads, walking straight through it.

His name was John of Ephesus. He was a churchman traveling overland from Syria and Palestine toward Constantinople in the very months the plague was moving along the same routes. Where Procopius gives us the capital, John gives us everything between the cities — and what he describes is a landscape coming apart. Villages standing silent with no one left in them. Crops ripe in the fields with no one to harvest them, simply rotting where they grew. Cattle and sheep wandering loose, untended, “roaming” the hills because their owners were dead. He wrote of walking past places where the living had become too few to bury the dead, so the dead were left where they fell.

And then there is the image that stays with you: ships drifting on the sea with no one aboard alive — crews struck down mid-voyage, the vessels carried by wind and current until they ran aground on empty shores, floating tombs washing up on the coast.

John’s account is, in fact, the longest description of the plague that survives from anyone who lived through it — longer than Procopius’s. He gives numbers too, and they climb even higher: he claims the dead were leaving Constantinople at five, ten, sixteen thousand a day, until the men posted at the city gates to tally the corpses gave up counting altogether. That last detail is quietly devastating. The counting stopped not because the dying stopped, but because the number lost its meaning.

Here is the part the video could not fit, and it matters for the whole argument about what we can and cannot know. We do not have John of Ephesus’s account in his own book. The section of his history that described the plague — Part Two of his Ecclesiastical History — is lost. It survives only because later writers copied it: a Syriac chronicler known as Pseudo-Dionysius of Tell-Mahre quoted him at length, and where even that copy dropped the opening, a twelfth-century patriarch named Michael the Syrian happened to preserve the missing beginning. Our fullest eyewitness to the first pandemic in history reaches us secondhand, stitched together from copies of copies, with a gap that one medieval scribe accidentally filled and another accidentally left out.

Sit with that, because it is the deeper truth beneath the maximalist–minimalist debate in the video. The reason serious historians can argue about whether this plague killed thirty million or a fraction of that is not carelessness. It is that our evidence is genuinely this thin and this fragile — a bacterium’s DNA in a few Bavarian teeth, an emperor’s historian with a talent for drama, and a churchman’s vanished book that survives only because two later monks, centuries apart, chose to copy it.

The plague was real. The witnesses were real. And the record they left us is a handful of fragments we are still, fifteen hundred years later, trying to read in the right order.

Sources

• John of Ephesus — Wikipedia — https://en.wikipedia.org/wiki/John_of_Ephesus

• “John of Ephesus describes the Justinianic plague” — Roger Pearse — https://www.roger-pearse.com/weblog/2017/05/10/john-of-ephesus-describes-the-justinianic-plague/

• “Viewpoint: New Approaches to the ‘Plague of Justinian’” — Past & Present (2022) — https://academic.oup.com/past/article/254/1/315/6427314

• “Rejecting Catastrophe: The Case of the Justinianic Plague” — Mordechai & Eisenberg, Past & Present (2019) — https://academic.oup.com/past/article-abstract/244/1/3/5532056

Fact-check note: John of Ephesus’s plague account survives not in his own (partly lost) Ecclesiastical History but through later Syriac quotation — chiefly Pseudo-Dionysius of Tell-Mahre (the Chronicle of Zuqnin), with an opening portion preserved via Michael the Syrian. His daily-death figures (5,000–16,000) are his own claims, as contested as Procopius’s “10,000 a day,” and are presented here as testimony, not established fact.

In the video, there is a single sentence that I could not stop thinking about long after the script was finished. When Tambora erupted, an entire language died with it — the Tambora tongue, unrelated to anything spoken around it, gone before anyone thought to write it down.

That line is true. But it is not the whole story, because in 2004 a small part of that lost world was found again.

The man who found it was Haraldur Sigurdsson, a volcanologist from the University of Rhode Island — the same scientist who, decades earlier, had helped establish Tambora as the largest eruption in recorded history by measuring the volume of rock it expelled. In the summer of 2004, guided by local villagers and by the ghostly outlines that Tambora’s own pyroclastic flows had preserved, Sigurdsson and his team dug into a gully on the volcano’s western flank. Six weeks later, roughly three meters down, they reached a house.

Inside it, they found two people.

Everything had been carbonized. The heat of the surge had not simply buried the house; it had turned it to charcoal, the way Vesuvius had done to the bakers and merchants of Pompeii seventeen centuries earlier. The beams, the foundation stones they had collapsed around, bronze bowls, ceramic pots, iron tools — all of it charred and frozen in place, exactly as it had been on the evening of April 10th, 1815. Sigurdsson called it, cautiously, “the Pompeii of the East.” An entire way of life, sealed under ash in a single night, waiting two hundred years to be read.

And the reading produced a genuine surprise.

Almost nothing had been known about the Tambora people. The one fragment that survived was a short word-list, collected by a European visitor shortly before the eruption, which showed that the Tambora language belonged to no family spoken anywhere near it — not to the Austronesian tongues of the surrounding islands, not to Malay, not to Javanese. It was an isolate, an orphan. But when Sigurdsson’s team examined the design and decoration of the carbonized artifacts, they found the answer to where these people had come from. The style pointed north and west — toward Vietnam and Cambodia. The Tambora language, linguists now believe, was related to the Mon-Khmer family of mainland Southeast Asia.

Think about what that means. On one small volcanic island in Indonesia lived the last speakers of a language that had traveled, at some unknown point in the deep past, across a thousand miles of ocean from the Asian mainland — a final, isolated branch of a tree whose other branches grow in the jungles of Cambodia. They farmed their rice, traded their bronze, and spoke a tongue that connected them to a world they had probably long forgotten. And then, in a single evening, that entire branch was cut off. Not just the people, but the last living record of where they came from.

This is the part the runtime could not hold: that Tambora did not only starve Europe and freeze New England and give the world Frankenstein. It also erased a small human story so completely that it took two centuries and a volcanologist with a trowel to recover even its outline.

The mountain kept its secrets in charcoal. We are only now learning how to read them.

Sources

· “URI volcanologist discovers lost kingdom of Tambora” — University of Rhode Island — https://today.uri.edu/news/uri-volcanologist-discovers-lost-kingdom-of-tambora/

· “Scientists unearth ‘Pompeii of the East’” — The Christian Science Monitor — https://www.csmonitor.com/2006/0302/p14s03-sten.html

· “The lost kingdom of Tambora is found” — Phys.org — https://phys.org/news/2006-02-lost-kingdom-tambora.html

· Tambora culture — Wikipedia — https://en.wikipedia.org/wiki/Tambora_culture

1815 eruption of Mount Tambora — Wikipedia —

There’s a reason Dharmasvamin’s 1235 visit to Nalanda didn’t happen in isolation, and the video didn’t have room for it: he was one of hundreds of monks, over roughly five centuries, who made the same brutal overland journey from Tibet into India for the same reason — to study at Nalanda and its sister monasteries, and to carry what they learned back home. That steady, unglamorous traffic turns out to be the actual reason so much of Nalanda’s intellectual world still exists to be studied today, and it has almost nothing to do with the raid everyone focuses on.

Here’s the detail worth knowing: a huge amount of what was taught and written at Nalanda no longer survives in its original language. Sanskrit Buddhist philosophy — the actual treatises of towering figures like Nagarjuna, Vasubandhu, Asanga, and Dharmakirti, scholars whose work anchored centuries of Nalanda’s curriculum — exists today, in large part, only because it was translated into Tibetan. The Tengyur, the vast Tibetan collection of translated commentaries and treatises, contains thousands of these works. Many of their Sanskrit originals are simply gone. Lost to time, to fire, to damp, to neglect, to exactly the kind of violence and slow institutional decay the video describes. The only reason we can read them at all is that a translation, made by someone like Dharmasvamin or one of his predecessors, happened to survive in a different script, in a different country, copied and recopied by monks who had no way of knowing they were preserving the only surviving version of anything.

That’s worth sitting with next to the story of Rahula Shribhadra teaching his last students in the ruins. Dharmasvamin wasn’t a tourist visiting a famous, fading site out of curiosity. He was there to study grammar under one of the last people alive who’d been trained in an unbroken line going back centuries, and part of what a serious student did, in that tradition, was carry the material home and put it into another language before there was no one left who could teach it directly. Nalanda’s real afterlife isn’t the ruin you can visit in Bihar today. It’s sitting in translated form in monasteries across the Tibetan plateau, in a canon most people who’ve heard of Nalanda have never heard of.

This connects to something that keeps coming up this sprint, in completely different centuries and completely different contexts: the actual record of an event and the physical site of that event are not the same thing, and the second one failing doesn’t necessarily mean the first one is gone. A letter can outlast the city that never sent it. Ice can remember a volcano no chronicle names correctly. A script can go unread for three thousand years and still be sitting there, waiting, once someone finally learns to read it again. And a university can burn, badly, repeatedly, over centuries — and still, in a sense, survive, because enough of what mattered about it had already been quietly copied out and carried somewhere else by people who were never trying to be heroes about it. They were just doing their coursework.

Sources: Tengyur — Wikipedia and 84000.co (Tibetan Buddhist canon, preservation of Nalanda-tradition scholars’ works); Dharmasvamin — Wikipedia; Nalanda mahavihara — Wikipedia.

There’s a detail the video didn’t have room for, and it changes how you should read everything else in it: for most of the last hundred years, scientists had the date of the 536 event slightly wrong. Ice core records placed the eruption around 529 — close, but off by seven years — until a 2015 study led by the climate scientist Michael Sigl went back and re-synchronized the chronology. The correction matters less for the specific number than for what it reveals about the method itself, because the method is the actual story here.

Here’s how you date a volcanic eruption from 1,500 years ago when no surviving human record names the mountain. You drill ice cores at both poles — Greenland in the north, Antarctica in the south — places that have been accumulating snow, layer by undisturbed layer, for tens of thousands of years, each layer a little time capsule of whatever was in the atmosphere that year. A large eruption throws sulfate particles high enough into the stratosphere that global winds carry them everywhere, including both poles, where they fall out with the snow and get sealed into the ice. If you find a sulfate spike in a Greenland core and a matching spike in an Antarctic core, precisely lined up in the layer-counted chronology, that’s not a coincidence. That’s confirmation the eruption was genuinely global — big enough, and high enough into the atmosphere, to reach both ends of the Earth in the same year. Scientists call this bipolar synchronization, and it’s essentially cross-examining two independent witnesses who’ve never met, the same evidentiary logic Elias Loomis used with handwritten letters after the Carrington storm, just conducted in ice instead of ink, over a much longer distance in time.

That’s what let Sigl’s team fix the date to 536 with real confidence, and it’s also what confirmed something the tree rings alone couldn’t fully prove: the eruption wasn’t a modest regional event that happened to get written about a lot. It was large enough, and violent enough, to leave an identical fingerprint on ice sheets eight thousand miles apart. The ice didn’t know about Procopius. The ice didn’t know about the Chinese famine chronicles. It had just been quietly falling, layer after layer, recording sulfate levels with no interest in being read by anyone, for the exact fourteen centuries it took until someone finally asked it the right question.

There’s something worth sitting with in the fact that the most reliable witness to 536 wasn’t a person at all. Procopius wrote precisely and carefully, and he still couldn’t tell you what caused what he was looking at. The ice never tried to explain anything — it just kept accumulating, indifferent to the question, until a research team with a corer and a mass spectrometer showed up fourteen hundred years later and it turned out to have been the most exact record in the room the entire time.

Sources: Sigl et al. (2015), bipolar ice-core synchronization and revised dating of the 536 CE volcanic event, as summarized via peer-reviewed climate science literature; Volcanic winter of 536 — Wikipedia; Büntgen et al. tree-ring isotope studies, cross-referenced against the ice-core chronology.

Buried in the video is a detail I didn’t have room to unpack: at Pylos, one of the Mycenaean palaces destroyed in the Bronze Age Collapse, archaeologists found clay tablets baked hard by the same fire that ended the palace, recording what looks like a kingdom organizing coastal watchmen and rowers in its final days — bracing for an attack the record never tells us the outcome of. That’s a remarkable thing to know. It’s an even more remarkable thing that we can read it at all, because for most of the twentieth century, nobody could.

The script on those tablets is called Linear B, and until 1952 it was a genuine mystery — a set of symbols found at Mycenaean sites across Greece and Crete that nobody could connect to any known language. Professional philologists worked on it for decades. The best-funded, most credentialed minds in the field tried and failed. The person who actually cracked it was Michael Ventris, a British architect — not a linguist, not an archaeologist, someone who’d been fascinated by the puzzle since he heard a lecture on it as a fourteen-year-old schoolboy and never let it go, working on it as a private obsession alongside his actual career designing buildings.

In July of 1952, Ventris announced on BBC radio that he’d broken it. The script wasn’t some unknown, undocumented language, as most experts had assumed. It was Greek — a startlingly early, archaic form of it, five hundred years older than Homer. That single decipherment didn’t just solve a puzzle. It rewrote the timeline of an entire civilization, proving Greek-speaking peoples were writing administrative records centuries before anyone had evidence for it.

He didn’t do it entirely alone at the end. Within weeks of the announcement, a young Cambridge philologist named John Chadwick got in touch, became the first scholar to publicly back the decipherment, and the two spent the next four years turning a breakthrough into a fully documented body of work — a joint book, Documents in Mycenaean Greek, that’s still foundational reading in the field. Ventris died in a car accident in 1956, weeks after that book was published. He was thirty-four.

There’s a pattern worth noticing across almost everything in this sprint so far. Richard Carrington was a brewer’s son with a private telescope, not a professional astronomer, when he recorded the first solar flare anyone had ever documented. The scientists who eventually pinned down Tunguska’s likely cause were working from tree-fall patterns and physics, not eyewitness testimony, because there wasn’t any that held up. And now: an architect, moonlighting on a linguistic puzzle that had beaten the professionals, who ended up being the reason we can read a Mycenaean kingdom’s last defensive preparations instead of just staring at symbols nobody understood.

None of these people were credentialed to make the discovery they made. All of them made it anyway, by paying closer, longer, more stubborn attention than anyone whose job it technically was. The Bronze Age Collapse destroyed the world that wrote those tablets. It took another thirty-two centuries, and one architect’s private obsession, for anyone to actually hear what they’d been trying to say.

Sources: Late Bronze Age collapse — Wikipedia; Michael Ventris — Wikipedia and Britannica; “The Decipherment of Linear B” — University of Cambridge Faculty of Classics; Linear B — Wikipedia (Pylos tablet content and context).

Six years after three legions were annihilated in the Teutoburg Forest in 9 AD, a Roman commander named Germanicus marched his army back to the site to see it for himself. What he found, according to the historian Tacitus, was a battlefield that had never really ended — bleached bones scattered across the ground, weapons half-buried in mud, altars where the Germanic tribes had apparently made ritual sacrifices of the Roman officers they’d captured. Germanicus’s men spent half a day gathering what remained of their countrymen. Tacitus records that Germanicus himself laid the first piece of turf on the burial mound they raised over the bones — a small, symbolic, deeply Roman act of putting a wound to rest.

Tiberius, the emperor back in Rome, was furious about it. Not because the burial was disrespectful, but because it was too respectful — a public, formal act of mourning that reminded every soldier under Germanicus’s command exactly how badly Rome had been beaten there, at exactly the moment Tiberius needed those same soldiers focused on finishing the war, not grieving the last one. There was also a religious complication: Germanicus held the office of augur, a priesthood that was supposed to avoid direct contact with the dead, and his own hands in that burial mound were, technically, a problem.

That’s what it actually looks like when a Roman disaster gets processed. It’s loud. It’s political. It’s expensive in soldiers’ morale and awkward for emperors and complicated by religious law, and six years later people are still visiting the site, still handling the bones, still arguing about what the mourning is doing to the army’s spirit. Modern archaeology at Kalkriese, widely believed to be part of the battlefield, backs up Tacitus here — the bones show signs of having lain exposed for years before anyone buried them, exactly as he described.

Now hold that up against the Ninth Legion. No burial expedition. No emperor arguing with a general about the optics of grief. No altar, no turf ceremony, no scattered weapons anyone bothered to record finding. Nothing, because as far as the actual surviving evidence shows, nobody in Rome experienced losing the Ninth as an event at all. There was no single afternoon to be angry about. The unit is in the record at York in 108, doing what legions do, and then it just isn’t anymore — not mourned, not memorialized, not fought over by emperors deciding how visible the grief should be.

That contrast is the real argument for taking the quieter version of this story seriously. Catastrophes leave fingerprints — bones somebody has to bury, emperors who have opinions about how the burying gets done, historians who can’t stop themselves from writing it down. The Ninth left none of that. Which means either the record of a real disaster simply didn’t survive eighteen centuries, which happens, or there wasn’t a single disaster to record in the first place — just a legion that got reassigned, reshuffled, and eventually spent, the way units are all the time, with nobody thinking to mark the exact year it stopped being the Ninth Legion of Spain and started being nothing in particular.

Both are real possibilities. Only one of them makes a satisfying story. History doesn’t owe us the second kind.

Sources: Legio IX Hispana — Wikipedia; Duncan B. Campbell, “The Fate of the Ninth: The curious disappearance of one of Rome’s legions”; Tacitus, Annals, Book I (Germanicus’s 15 AD visit to the Teutoburg battlefield), as summarized via Wikipedia’s “Battle of the Teutoburg Forest” and “Germanicus’s expedition into Germania” entries; Kalkriese site archaeological findings corroborating extended bone exposure, per the same sources.

For most of the twentieth century, the leading explanation for Tunguska was the dull, sensible one: a stony asteroid or comet fragment, tens of meters across, tore itself apart in the atmosphere and released its energy all at once, with nothing solid left over to leave a crater. Dull and sensible didn’t stop people from looking for something stranger, though — and in 1973, two American physicists, A.A. Jackson IV and Michael P. Ryan Jr., published a paper in Nature, one of the most rigorously reviewed journals on Earth, proposing that Tunguska hadn’t been an asteroid at all. It had been a black hole.

Watch the video here

Their idea wasn’t dressed up as speculation. It was a real physical model with a real testable prediction. A tiny black hole, they argued, moving fast enough and dense enough, could pass straight through the planet without releasing an underground shockwave — the rock’s rigidity would offer no resistance worth mentioning to something that small and that heavy. It would enter over Siberia and, following a near-straight line through the Earth, exit somewhere in the North Atlantic, between 40 and 50 degrees north, 30 to 40 degrees west. And when it exited, it should have produced its own shockwave — an air blast, an underwater disturbance, something a ship’s barograph or a shipping company’s log might have quietly recorded without anyone realizing what they’d caught.

That’s the part worth sitting with. Jackson and Ryan didn’t just propose something wild and walk away. They handed the rest of the scientific world a coordinate, a time window, and a specific signature to go check. That’s what separates a real hypothesis from a story — not how strange it sounds, but whether it tells you exactly where to look if it’s wrong.

People looked. Researchers combed shipping records and oceanographic data for the Atlantic exit event the theory demanded. Nothing matched. By 1974, a follow-up paper in the same journal — “Tungus Event Was Not Caused by a Black Hole” — closed the door, and by the late seventies the idea was, in the physics community’s own unglamorous phrase, buried.

What strikes me isn’t that someone proposed a black hole. Mysteries invite that; Tunguska has also attracted natural gas explosions, secret Tesla experiments, and a dozen other explanations serious and not. What strikes me is how fast and how completely a real scientific claim can be tested and closed, provided it’s specific enough to fail. The asteroid-airburst explanation isn’t the leading theory today because it’s the most exciting one. It’s the leading theory because, unlike the black hole, nothing has yet shown up to contradict it — and everyone still knows exactly what would.

That’s the quieter cousin of the question the video asks. Not just “how would we know if it happened again,” but: when the evidence runs out, what actually earns the right to fill the gap? Not the most dramatic answer. The one that told you, in advance, how to prove it wrong.

Sources: Tunguska event — Wikipedia; Britannica; NASA “115 Years Ago” retrospective; Jackson & Ryan, “Was the Tungus Event Due to a Black Hole?”, Nature (1973); Beasley & Tinsley, “Tungus Event Was Not Caused by a Black Hole,” Nature (1974).

By the time the sky caught fire on the first of September, 1859, the one thing that might have reported it instantly across an ocean was already dead. The first transatlantic telegraph cable — the pride of 1858, the cable Queen Victoria used to congratulate a president — had failed the previous October, after barely three weeks of real service, burned out by an engineer applying too much voltage in his hurry to make it faster. So when the storm hit, humanity’s fastest working link between continents was silence. News still moved at the speed of steamships and the mail they carried.

And yet within two years, a mathematician at Yale named Elias Loomis had done something that should have been impossible without a single working wire: he’d proven, to the satisfaction of the scientific world, that the aurora over the Rocky Mountains, the aurora over Colombia, the readings jumping off the scale at Kew Observatory in London, and the telegraph fires in Boston were not four strange coincidences. They were one storm, wrapped around the entire Earth at once.

He proved it with letters.

Loomis wrote to observatories, to newspapers, to ship captains who’d logged the sky in their voyage records, on continents that had no way of comparing notes with each other in real time. He collected every independent account he could find and plotted them by hand onto a single map. None of those witnesses — the gold miners making breakfast at 1 a.m., the readers in Queensland by aurora-light, the operator in Norway knocked out by a shock from his own key — knew that anyone else, anywhere, had seen the same impossible thing that same night. The proof wasn’t in any one report. It was in how many independent, unconnected people described the same event without ever comparing stories.

That’s worth sitting with, because it’s a discipline that has nothing to do with telegraph cables or satellites and everything to do with how you tell the difference between an anecdote and a fact about the world. One vivid account, however sincere, is a story. A dozen accounts from people who couldn’t have coordinated is evidence. Loomis didn’t have instant communication. He had patience, a lot of postage, and a refusal to declare the world’s business settled until enough of the world had independently said the same thing.

Which is the quiet, uncomfortable footnote sitting underneath this video’s bigger warning. We’ll very likely see the next Carrington-class storm coming — a coronal mass ejection gives some hours to days of notice. What a storm strong enough to burn out transformers and blind satellites might also do, for a while, is take away the very thing that let 1866’s rebuilt cable, and everything wired after it, replace Loomis’s method with something faster and lazier: the assumption that if it’s on the wire, it’s already been checked. In September 1859, nothing was on the wire. Somebody still had to write it down, send it, and wait to see if the world agreed.

Watch it here:

Sources: Carrington Event — Wikipedia; “The Carrington Event: History’s Greatest Solar Storm” — Space.com; “A Perfect Solar Superstorm” — HISTORY.com; “One Swallow Does Not Make a Summer: 160 Years of Richard Carrington’s Legacy” (Kew Observatory, Balfour Stewart, Elias Loomis) — arXiv:1910.02795; “The Grand Aurorae Borealis Seen in Colombia in 1859” — arXiv:1508.06365; “Transatlantic Telegraph Cable” — Wikipedia / History.com (1858 cable failure, October 20, 1858).

Companion essay to the third Arcanum World video.

In 1900, a Greek sponge diver came up from forty-five metres of cold black water terrified by what he had seen — bronze bodies, he thought, the dead at the bottom of the sea. He had not found bodies. He had found a Roman cargo ship two thousand years on the floor of the Aegean, and the corroded green lump of bronze that arrived in the Athens museum that summer would, a hundred years later, force the scholarly world to admit that some Greek workshop around 150 BCE had built the first computer — fourteen hundred years before anything like it would exist again.

The video this essay accompanies tells that story. This piece is about a different question. What other things have we already forgotten how to make?

The Antikythera Mechanism is the most spectacular case in the historical record. It is not the only one. Lost technologies are everywhere if you look. The pattern of their loss tells us something about how human knowledge actually works that the video could only gesture at.

The Romans built concrete that still stands. Their marine harbours have been underwater since the time of Augustus and are structurally sound. We have studied the material for two centuries. We know what was in it — a particular type of volcanic ash called pozzolana, lime, seawater, and a deliberate amount of tobermorite mineral that grows inside the concrete over the centuries, making the harbour walls stronger the longer they sit in salt water. Modern Portland cement does the opposite. It begins degrading the day it sets. Marie Jackson at Berkeley has now approximated the Roman recipe in the laboratory; we cannot yet make it at industrial scale at competitive cost. Every harbour built since the fall of the Empire has been built out of a worse material than the one the Romans used.

Damascus steel was forged in the Middle East between the third and seventeenth centuries. Crusader knights brought home stories of blades that could cut through European armour and silk scarves in the same swing. The technique died around 1750, partly because the particular ore the smiths used — wootz steel from southern India — became unavailable when the mines were exhausted. Modern metallurgists have spent decades trying to reverse-engineer the result. They have come close. They have not equalled it. The Damascus blades currently in museum collections cannot be made today, with anyone’s tools, by anyone alive.

Greek fire was the fleet-burning naval weapon that kept the Byzantine Empire alive for seven centuries longer than it would otherwise have lasted. The recipe was a state secret entrusted to a single family, the Kallinikoi. When the family died out, the recipe died with them. We do not know what was in it. We have theories — naphtha, quicklime, sulphur, pine resin — but no theory has been able to reproduce the behaviour described in the Byzantine sources, where the substance burned on water and could not be extinguished by sand. We are still guessing.

These are the famous cases. There are many more.

Tyrian purple required a particular sea snail and a particular Phoenician industrial process. When the Phoenician industry collapsed, the dye industry collapsed with it. The exact shade of purple that the Phoenicians produced for two thousand years cannot be made today. The pigment used in some Pompeian frescoes has resisted modern chemical replication for a hundred and fifty years of scholarship. The Stradivarius varnish that gave the eighteenth-century violins their resonance was lost within two generations of Antonio Stradivari’s death, despite his sons and grandsons being violin makers themselves. The construction technique that allowed Justinian’s architects to build the Hagia Sophia dome in five years and have it stand through fifteen centuries of earthquakes that have flattened modern buildings around it is still not fully understood. The Maya Long Count calendar mathematics can be reconstructed from the few codices that survived the Spanish, but the original astronomical observations that fed the calendar — the centuries of patient sky-watching that produced the numbers — are gone.

What these losses share is a particular shape. They were workshop knowledge. They lived in the hands of specific people, in the workspaces of specific cities, in the apprentice chains that connected one generation to the next. They were rarely written down in a form that could be reconstructed without the original masters. The Romans did not publish the recipe for their concrete because everyone in the Roman concrete industry already knew it. The Damascus smiths did not publish their forging technique because it was their livelihood. The Kallinikoi family did not publish the Greek fire recipe because it was the family’s reason for existing.

When the people died and the apprentice chains broke, the knowledge went with them. It did not fade gradually. It vanished in a generation. By the time anyone realised what had been lost, the people who could have rebuilt it were already a hundred years dead.

This is why the Antikythera Mechanism feels so strange when you first see it. You expect ancient technology to be primitive — something we built better versions of. Instead you find the opposite. Someone in a Greek workshop around 150 BCE solved a problem that the rest of human history would not solve for fourteen hundred years. Then they died. And we had to start over.

There is a sentence I keep returning to.

We are not in the third millennium of human progress. We are somewhere in the second great round of it. There were others before us. The strangest thing about what we know is not what we know. It is what we have already, several times, forgotten.

Watch the full video →

Sources

• Derek de Solla Price, Gears from the Greeks: The Antikythera Mechanism — A Calendar Computer from ca. 80 B.C. — American Philosophical Society, 1974

• Tony Freeth et al., “Decoding the ancient Greek astronomical calculator known as the Antikythera Mechanism” — Nature 444 (2006)

• Jo Marchant, Decoding the Heavens: A 2,000-Year-Old Computer and the Century-long Search to Discover Its Secrets — Da Capo Press, 2009

• Alexander Jones, A Portable Cosmos: Revealing the Antikythera Mechanism, Scientific Wonder of the Ancient World — Oxford University Press, 2017

• Marie D. Jackson et al., “Phillipsite and Al-tobermorite mineral cements produced through low-temperature water-rock reactions in Roman marine concrete” — American Mineralogist 102, 1435–1450 (2017)

• John D. Verhoeven, Damascus Steel: A Continuing Mystery — Iowa State University Press, 2007

• Alex Roland, “Secrecy, Technology, and War: Greek Fire and the Defense of Byzantium” — Technology and Culture 33, 4 (1992)

• Inge Boesken Kanold and Roland Haubrichs, “Tyrian Purple Dyeing: An Experimental Approach with Fresh Murex trunculus” — in Dyes in History and Archaeology, 2008

• Joseph Nagyvary, Renald N. Guillemette, Clifford H. Spiegelman, “Mineral Preservatives in the Wood of Stradivari and Guarneri” — PLoS ONE 4, 1 (2009)

• Rowland Mainstone, Hagia Sophia: Architecture, Structure and Liturgy of Justinian’s Great Church — Thames & Hudson, 1988

Arcanum World publishes one video and one essay each week on the questions every civilization has asked. Next time — the day the wealthiest empire in the world ended in a single afternoon. Subscribe.

Smith was a former apprentice engraver who had taught himself Akkadian and Sumerian by spending his lunch breaks in the British Museum’s Assyrian Gallery. In 1872, working on a backlog of unsorted cuneiform fragments from Austen Henry Layard’s excavation at Nineveh, he came across Tablet XI of what we now call the Standard Babylonian Epic of Gilgamesh. The tablet was unprovenanced — pieces had been jumbled together for two decades. Smith was the first reader to translate it as a coherent narrative.

The text told a story Smith already knew. A god warned a chosen man of an impending deluge. The man was instructed to build a ship and load it with his family and the seed of every living thing. The flood came. The ship grounded on a mountain. The man released birds — first a dove, then a swallow, then a raven, and the raven did not return. He offered sacrifice. The gods smelled the smoke.

The cadence is in Genesis. The structure is in Genesis. The bird sequence is in Genesis. The mountain landing is in Genesis. And Tablet XI predates the Pentateuch by approximately a thousand years.

The lines themselves are worth reading. Tablet XI, in modern translation, includes this passage:

“I sent forth a dove and let her go. She flew to and fro, but there was no resting place, and she returned. I sent forth a swallow and let her go. She flew to and fro, but there was no resting place, and she returned. I sent forth a raven and let her go. She saw that the waters had abated. She did not return.”

Genesis 8:7–12 reads, in summary:

“And he sent forth a raven, which went forth to and fro, until the waters were dried up from off the earth. Also he sent forth a dove… but the dove found no rest for the sole of her foot, and she returned… he sent forth the dove out of the ark… and she returned not again.”

The bird sequence is reversed in Genesis, but the device — successive birds testing whether the world has dried — is identical. There is no serious living Assyriologist who reads these passages independently and concludes they were composed in cultural isolation from one another.

What does that mean? Three things the video could not fully unpack.

First, the Hebrew authors of Genesis almost certainly inherited the flood narrative through Mesopotamian channels — most likely during or after the Babylonian exile in the sixth century BCE, when the Judean elite spent two generations in Babylon and learned its literary canon. They did not plagiarise; they assimilated. Every ancient literary tradition assimilated the traditions that surrounded it. Originality is a Romantic-era idea, not an ancient one.

Second, the Babylonians themselves inherited the story from earlier Sumerian sources. The Epic of Ziusudra — fragmentary but real — predates Gilgamesh by perhaps another thousand years. So the chain stretches further: Sumer → Babylon → Israel → Christendom → the modern Sunday-school imagination. Four thousand years of telling and re-telling. Every generation believed it was the first.

Third — and this is the part the video gestures at but cannot dwell on — Smith’s discovery did not damage faith in the Bible. It complicated it. Many serious religious thinkers since 1872 have come to read the Babylonian parallel not as evidence that the Bible is false, but as evidence that some stories are too large for any one tradition to own. The flood is one of them. It belongs to everyone who has ever watched a river rise.

That was Smith’s real discovery, in the end. Not that the Bible was second. That the flood was older than any of us thought.

He died four years later in Aleppo, on a return expedition to Nineveh, of a fever he caught en route. He was thirty-six.

Watch the full video →

Sources

• Andrew George (trans.), The Babylonian Gilgamesh Epic: Introduction, Critical Edition and Cuneiform Texts, 2 vols. — Oxford University Press, 2003

• Irving Finkel, The Ark Before Noah: Decoding the Story of the Flood — Hodder & Stoughton, 2014

• George Smith, “The Chaldean Account of the Deluge” — Transactions of the Society of Biblical Archaeology, 1873

• Stephanie Dalley, Myths from Mesopotamia: Creation, the Flood, Gilgamesh, and Others — Oxford University Press, 1989

• David Damrosch, The Buried Book: The Loss and Rediscovery of the Great Epic of Gilgamesh — Henry Holt, 2007

• William Ryan & Walter Pitman, Noah’s Flood: The New Scientific Discoveries About the Event That Changed History — Simon & Schuster, 1998

Arcanum World publishes one essay and one video each week on the questions every civilization has asked. Next week — not everyone has a voice in their head, and the people who don’t are sitting on top of one of the oldest debates in philosophy. Subscribe.

The oldest of them is around four thousand years old. Anatomically modern humans have existed for at least three hundred thousand. That gap — the ninety-eight point seven percent of human existence that no current religion accounts for — is the most interesting question in the study of religion. Because the gap is not empty. Every cave we have excavated, every burial we have opened, every fire pit we have dated, tells us the first humans were already doing something. We just stopped recognising it as religion. The video this essay accompanies walks through the four layers we have uncovered. This piece sits underneath the video and traces what those four layers still mean in your life.

The first layer is animism. The world is alive. Not metaphorically — concretely. Rivers, mountains, storms, stones, the sun, the dead, and the unborn all contain a spirit that can be spoken to. Edward Burnett Tylor coined the word in Primitive Culture (1871), and for a long time the West treated animism as a primitive failure to understand cause and effect. The Hadza of Tanzania, the Sámi of Lapland, the Ainu of Japan, the Khoisan of southern Africa, and the first nations of the Americas have always known otherwise. Recent anthropology — Robert N. Bellah’s Religion in Human Evolution (2011) is the foundational survey — suggests animism is the oldest, most universal, and most resilient pattern of human thought we have evidence for. You still do it. You talk to your car. You blame the rain. You sense the house is watching you. The wiring did not go anywhere.

The second layer is the ancestors. Sixty thousand years of Neanderthal and sapiens burials show the same thing: ochre, flowers, tools, food placed with the body. The dead were not abandoned. The dead were furnished. Death did not end the relationship; it changed the address. The grandfather who taught you is still teaching you, if you keep speaking to him. Every modern religion still does this — Mass on All Souls’ Day, Ching Ming, Día de los Muertos, the Hindu shraddha rite. So do the secular: the cemetery visit, the framed photograph, the lit candle on a death anniversary. You think you are honouring a memory. The ancient mind knew you were maintaining a relationship.

The third layer is the totem, and the law it carried. A totem binds a clan to an animal or plant: the clan does not eat the totem, does not hunt it, does not name it casually. The clan and the creature are kin. Religion was the law before there was law — the salmon clan did not overfish, the buffalo clan policed the hunt. The forbidden foods of the modern religions are totemism wearing scripture’s clothes: not the dog, not the horse, not the cow, not the pig.

The fourth layer is the mother. Around thirty-five thousand years ago, female figurines begin appearing in the archaeological record from Spain to Siberia. The Venus of Willendorf is the famous one; there are hundreds more. We are not certain they were gods. We are certain the pattern is too widespread to be decoration. Most scholars now read these as the earliest deities we have material evidence of. She survived: Isis, Inanna, Cybele, Demeter, Durga, Mary. The mother goddess never died. She was renamed.

Then, eleven and a half thousand years ago in southeastern Anatolia, hunter-gatherers built Göbekli Tepe — limestone pillars weighing twenty tons, arranged in circles, six thousand years older than Stonehenge, seven thousand older than the Great Pyramid, older than agriculture itself. Klaus Schmidt’s excavation reports rewrote the textbook story. Religion was not invented after we settled down. The temple came first. Worship came first. The gods built the city.

So what did people worship before the big religions? They worshipped what was alive, what was dead, what was other, and what was beginning. And then they built a temple before they built a town. Every modern religion is built on top of those four layers — and so, quietly, are you, even if you believe in nothing at all.

We did not invent religion four thousand years ago. We renamed something that was already three hundred thousand years old.

Watch the full video →

Sources

· Klaus Schmidt et al., Göbekli Tepe excavation reports — German Archaeological Institute, 1995–2014

· Mircea Eliade, A History of Religious Ideas, Vol. 1 — University of Chicago Press, 1978

· Edward Burnett Tylor, Primitive Culture — John Murray, 1871

· Robert N. Bellah, Religion in Human Evolution — Harvard University Press, 2011

· Marija Gimbutas, The Language of the Goddess — Harper & Row, 1989

· Frank W. Marlowe, The Hadza: Hunter-Gatherers of Tanzania — University of California Press, 2010

· Christopher Boehm, Hierarchy in the Forest — Harvard University Press, 1999

· Brian Hayden, Shamans, Sorcerers, and Saints — Smithsonian Books, 2003

Arcanum World publishes one essay and one video each week on the questions every civilization has asked. If this resonated, subscribe.

—

What actually makes the Strait of Hormuz so important?

Roughly 20 million barrels of oil and oil products move through the Strait of Hormuz every day. That is approximately one-fifth of global oil consumption and roughly one-third of all seaborne oil trade. There is no other waterway on the planet through which this much energy flows.

The strait itself is narrow ‚21 miles wide at its tightest point, with shipping lanes only two miles wide in each direction. Every tanker carrying Saudi, Emirati, Qatari, Kuwaiti, Iraqi, or Iranian oil must pass through it on the way to refineries in Asia, Europe, and North America. There is no alternative route by sea. The geography is not a choice. It is a funnel.

And on one side of that funnel sits Iran.

Why can’t the US just build a pipeline around it?

This is the first question most people ask, and the answer reveals the structural nature of the problem. A handful of bypass pipelines exist‚ notably the UAE’s Habshan-Fujairah line and Saudi Arabia’s East-West (Petroline) pipeline but their combined capacity is only a fraction of Hormuz’s daily throughput.[^3] Building enough new pipeline capacity to replace the strait would take a decade, cost tens of billions of dollars, and still not solve the core problem: the countries that produce the oil are on the Persian Gulf side of the strait. A pipeline across the Arabian Peninsula still has to end at a port, and most of those ports are themselves vulnerable to the same Iranian missile and drone inventory that threatens the strait.

In other words: the chokepoint is not the water. The chokepoint is the geography of production itself.

Could Iran really close the Strait ‚and what would happen if they did?

Iran’s military posture in the Gulf is optimized around exactly one scenario: denying the strait to other navies and to commercial shipping. The Islamic Revolutionary Guard Corps Navy operates hundreds of fast-attack craft, sea mines, and mobile anti-ship missile batteries specifically designed to swarm larger warships and disable tankers.

Most defense analysts agree Iran cannot *permanently* close the strait- a sustained US and allied response would, over weeks, reopen it. The more dangerous question is whether Iran can close the strait *temporarily*, and what a temporary closure would do.

The International Energy Agency has modeled these scenarios. A one-week disruption of Hormuz traffic, during which roughly 70 million barrels of oil fail to reach global markets, would likely push crude above $200 per barrel within days.[^5] Gasoline prices in the United States would spike by $2-$3 per gallon. Airlines, logistics networks, petrochemical supply chains, and industrial production across Asia would absorb the shock first. Recessions have started on less.

This is why the threat of closure is, in itself, Iran’s most valuable strategic asset. They don’t need to close it. They only need everyone to believe they could.

The $50 billion price tag: what America is actually paying for:

The United States maintains the Fifth Fleet at Naval Support Activity Bahrain, rotating carrier strike groups through the Gulf, an expeditionary Marine force, and a permanent air presence across bases in Qatar, the UAE, and Kuwait. The purpose is explicit: keep the strait open.

Estimates of the annual cost vary, but research by the Securing America’s Future Energy initiative and studies from Princeton’s Woodrow Wilson School have placed the figure at roughly $50 billion per year attributable specifically to securing Gulf oil flows. Some analyses go higher. None go meaningfully lower.

The striking thing about this arrangement is that the United States imports very little Gulf oil directly anymore. Since the shale revolution, the US has been a net petroleum exporter. So who is America actually guarding the strait *for*?

The answer is: for the global oil market, which is priced in US dollars. And that is where the petrodollar comes in.

The Petrodollar: the deal that turned oil into a dollar-denominated weapon

In 1974, in the aftermath of the Nixon shock that ended the gold standard and the OPEC oil embargo that tripled crude prices, the United States and Saudi Arabia arrived at an informal arrangement. Saudi Arabia would price its oil exclusively in US dollars. In exchange, the United States would provide military protection for the Kingdom and, by extension, for the oil flowing out of the Gulf.

This arrangement later extended across OPEC had two enormous consequences. First, it created permanent global demand for US dollars, because every country that imported oil had to hold dollars to pay for it. This is what allowed the US to run persistent trade deficits without collapsing its own currency. Second, it tied the stability of the entire dollar-denominated financial system to the geography of one narrow strait.

The $50 billion a year America spends patrolling Hormuz is, in effect, the maintenance cost of the dollar’s reserve-currency status. It is also the reason why, every time Iran rattles its sabers in the Gulf, the price of US Treasury bonds and the value of the dollar move in sympathy. A threat to the strait is a threat to the dollar. And a threat to the dollar is a threat to every asset priced against it.

Where this leaves us in 2026

Three structural shifts have been reshaping this picture over the last decade. China now imports more than half of the oil that transits the strait.[^8] Saudi Arabia has begun settling some of its oil trades in Chinese yuan. And Iran’s missile and drone capabilities demonstrated in the 2019 Abqaiq attack on Saudi processing facilities and in recent regional exchanges have grown far faster than the Gulf Arab states’ ability to defend against them.

The question for the 2030s is not whether the Strait of Hormuz will remain the most important 21 miles of water on Earth. It will. The question is whether the United States will still be the country guarding it and whether the dollar’s position as the global oil currency will survive the answer.

That is the trillion-dollar question that almost nobody in the headlines is asking.

---

## Sources

US Energy Information Administration, “The Strait of Hormuz is the world’s most important oil transit chokepoint,” EIA Today in Energy. https://www.eia.gov/todayinenergy/detail.php?id=55881

International Maritime Organization, Traffic Separation Scheme for the Strait of Hormuz. https://www.imo.org/

Reuters, “Explainer: Middle East oil bypass pipelines and their capacity.” https://www.reuters.com/business/energy/

International Institute for Strategic Studies (IISS), *The Military Balance 2024* -chapter on Iran. https://www.iiss.org/publications/the-military-balance/

International Energy Agency, “Oil market scenarios under Strait of Hormuz disruption.” https://www.iea.org/reports/

Securing America’s Future Energy, “The Military Cost of Defending the Global Oil Supply” (2018, updated 2023). https://secureenergy.org/

Spiro, David E. *The Hidden Hand of American Hegemony: Petrodollar Recycling and International Markets.* Cornell University Press, 1999.

International Energy Agency, “Oil 2024: Analysis and forecast to 2030” - Asia-Pacific import flows. https://www.iea.org/reports/oil-2024

---

## Further down the rabbit hole

- **Book:** Daniel Yergin, *The Prize: The Epic Quest for Oil, Money, and Power.* The definitive history of how oil shaped the 20th century - required reading for understanding why the Gulf matters.
- **Book:** Michael Klare, *Blood and Oil.* A darker but well-sourced account of how US foreign policy has oriented itself around Gulf petroleum.
- **Paper:** Eugene Gholz, “Assessing the ‘Threat’ to International Oil Supplies,” Cato Institute, 2018 - argues the $50 billion figure may overstate the real security benefit.


---

Arcanum World is a documentary-style research channel dedicated to the stories they left out ‚the hidden machinery behind the headlines. Find more at arcanumworld.org. If you found this useful, the next dispatch lands in your inbox automatically.*