When word reached Tenochtitlan that an Aztec regional commander had attacked the Spaniards left behind on the coast, and that some of them had been killed, Hernán Cortés did not treat it as a setback. He used it as a pretext. He walked into the palace of Moctezuma II, the Aztec emperor, and told him that for his own safety he was now in Spanish "protective custody" and should come along or there would be violence.
Si Sheppard, a military historian, said this was the moment Moctezuma's "entire worldview just collapses." Nobody had spoken to the emperor like that before. He let himself be led away and told his people not to resist. In Sheppard's telling, once Cortés had the emperor, "he already has the empire."
Sheppard went through the conquests of the Aztec and Inca empires in an episode of the Dwarkesh Podcast published October 1, 2026. The episode page puts the scale this way: Cortés and a few hundred conquistadors conquered an Aztec empire of about 6 million people in two and a half years, and a decade later Francisco Pizarro did the same to an Inca empire of some 10 million. The page also links the subject to AI, saying talk of AI takeover overlooks that history already has examples of a little-understood outside force using better technology and diplomacy to take over a much larger established order. A large part of the conversation asked how fragile those outcomes were. Sheppard's answer had several layers. The two emperors' personal decisions mattered. Centralized states were unusually easy to decapitate. And underneath both, he argued, were slow forces that would have favored Europe in the long run anyway.
Moctezuma's choice, reconsidered
Moctezuma knew Cortés was coming. Sheppard described an empire full of envoys, scouts and spies reporting that "something has washed up on the shores." The emperor first bargained, urging Cortés to leave and sending him precious gifts. Sheppard said the gifts only inflamed Cortés's appetite. Cortés reached the capital with Tlaxcalan allies, enemies of the Aztecs who had been cut off from the coast and from salt. Talks with Moctezuma ran through Cortés's interpreter and mistress, Doña Marina, known as Malinche. The chronicler Bernal Díaz described her translation as indispensable, according to an American Historical Association resource.
Moctezuma's council opposed letting Cortés into the city. Sheppard recounted the warning from the emperor's brother Cuitláhuac: if you allow this man inside your house, he will take your house away from you. Generations of historians have condemned Moctezuma for ignoring that advice. Sheppard disagrees. "I actually think that was his best option, whether he was aware of it or not," he said. A guest inside Tenochtitlan was easier to watch than "this random element at loose within his empire," talking to subject peoples who might turn against the capital.
In Sheppard's view, Moctezuma's real error was staying in the city, in a palace next to the one where the Spanish were lodged. If the emperor had found a reason to be elsewhere, such as a pilgrimage or a campaign, and had ruled through subordinates, Cortés "would have been baffled," Sheppard said. Chasing after the emperor would have been too provocative, and without him Cortés's options would have been very limited. "Because Moctezuma stayed in his capital, he made himself the target."
Cortés's own account fits the logic Sheppard describes. In his second letter to Charles V, dated October 30, 1520, Cortés wrote that he wanted Moctezuma held in Spanish quarters because holding the ruler would make the Spanish safer and help bring dependent territories under the king's authority. He also described using the killings on the coast to demand Moctezuma's move.
What followed was rule through a captive. Sheppard said Cortés had the coastal officers responsible for the attack burned alive in front of the emperor and the whole court, a punishment the Aztecs had never seen. Cortés then governed through Moctezuma, who was becoming "increasingly" his puppet. Sheppard called it classic Stockholm syndrome. Among Cortés's displays of power, he had small sailing boats called brigantines built and took the emperor cruising on Lake Texcoco. Sheppard said Moctezuma was astonished at using the wind to move across water. Cortés's letter likewise describes Moctezuma continuing to give orders while confined, with Cortés insisting that subjects obey their ruler, who was now under the Spanish king.
Why the emperors cooperated
The conversation came back to a puzzle: once captured, what did these rulers think would happen next? Did they expect the Spanish to leave if paid enough gold?
For Moctezuma, Sheppard suggested a psychological break. The records, he said, show a "pitiless autocrat" who had closed the aristocracy to outsiders. Sheppard called him "almost Stalinist" and said he had turned the ritual "flower wars" against the Tlaxcalans into campaigns of encirclement and annihilation. "Perhaps it's for the first time in his life, being told what to do just led to a complete psychological break," Sheppard said, presenting it as a guess. After that, he said, Moctezuma tried to serve his people without challenging Cortés. In the end, Sheppard said, it was not Cortés who got rid of him but his own people, who concluded he had sold them out. Sheppard called him the first Aztec emperor, or huey tlatoani, to be repudiated by his own people and replaced.
Atahualpa, the Inca ruler captured by Pizarro at Cajamarca, is harder to explain because he seemed to have his wits about him. Sheppard noted that Spanish accounts describe a regal, clearly intelligent man who adapted fast: he learned Spanish very quickly and picked up cards and chess. Sheppard thinks Atahualpa was fitting in while reaching out to his own people and waiting for a chance to mobilize them, a chance that never came. He had also struck a deal: fill a room with gold and silver, and he would be released. A November 1533 letter by Hernando Pizarro, Francisco's brother, describes Atahualpa arranging treasure shipments from the provinces to pay that ransom, and then his execution.
Sheppard's reading is that Atahualpa "misinterpreted Pizarro as someone who was on the same level, that kings don't do this to other kings." Instead, he said, Atahualpa was dealing with "a freebooting, entrepreneurial pirate, essentially, not someone operating in good faith." Sheppard said Atahualpa's critical mistake came earlier, when he walked into Pizarro's trap. Had he kept negotiations at a distance through envoys, Sheppard thinks, he might have frustrated Pizarro's ambitions.
Centralized empires and dispersed peoples
The discussion raised a contrast. Other Indigenous groups held off European powers for far longer. Into the late 19th century, after the American Civil War, US forces were still fighting tribes such as the Comanche. The suggestion was that the Aztecs and Incas fell because they were centralized: capture the emperor and everything falls, while more dispersed peoples could not be beaten that way.
Sheppard agreed, calling "dispersal" the key term. The Aztec and Inca systems invested absolute loyalty in the person at the top, the huey tlatoani for the Aztecs and the Sapa Inca for the Inca. Taking that one person destabilized the whole system and could let outsiders govern it through intermediaries or puppets. The Maya were different. Their city-states jealously guarded their independence from one another, so the Spanish had to take them one city at a time. Sheppard said that process took almost until the end of the 17th century, and even then resistance smoldered. In parts of the Central American highlands, he said, people still speak their ancestral languages before Spanish.
His North American example was Custer's Last Stand at the Little Bighorn in 1876, where he said "the entire 7th cavalry" was wiped out by warriors who gathered temporarily on horseback with guns and then dispersed. The National Park Service gives a narrower account. Custer's separate battalion of 210 men was destroyed, while the commands under Reno and Benteen survived fighting elsewhere. The Park Service also records the aftermath Sheppard pointed to: more US military pressure, seizure of the Black Hills and, within a year, the surrender of most remaining groups living outside reservations. Sheppard described the wider campaign as "what amounts to really total war," including barbed wire and the slaughter of the buffalo that Plains cultures depended on.
The Aztecs ran into a similar limit themselves, Sheppard said. They never managed to defeat the Tarascans to their west, whose first encounter with an invading Aztec army was a disaster for the Aztecs. He credited technology: the Tarascans had some exchange with South America and, like the Inca, could smelt copper, which they used for arrowheads. An archaeological report by Blanca Maldonado on one Tarascan-region site supports skilled copper smelting there, though it could not settle whether the furnaces predate the Spanish. Sheppard argued that wider exchange between North and South America could have evened the odds. Llamas would have helped the Aztecs a great deal, he said, and so would Inca bronze. "Bronze can't stand up against iron, but it's better than stone."
The conversation also pointed to a lopsided way of learning. Cortés had learned from the conquest of Hispaniola, and Pizarro knew the details of what Cortés had done a decade earlier and followed the same playbook. On the other side, it was unclear whether the Incas even knew the Aztecs existed. Sheppard called the American empires "closed worlds" and "very insular."
Re-running history
The central question was put as a thought experiment. If history were re-run many times with a few changes, such as different leaders on either side, how often would the Spanish still end up in control within a few years?
Sheppard answered with what he called the wheels of history. The large wheels, geography and climate, turn slowly. The small wheels in the foreground are individuals. On the small wheels, he said, much could have gone differently. If Cortés had been killed early, which he said could have happened several times, the leaderless conquistadors might have fallen apart and the conquest might have ended there.
The large wheels pointed elsewhere. Eurasia had many animals that could be domesticated: sheep, cows, goats, mules, pigs and, above all, horses. Sheppard argued that living alongside them created the conditions for contagious diseases that spread back and forth across Eurasia. That never happened in the Americas, so people there had no acquired immunity to smallpox or the many other diseases now found everywhere. The argument echoes Jared Diamond's Guns, Germs, and Steel, which PBS's background to the 2005 documentary describes as tracing the conquerors' advantages to what had built up and spread across Eurasia, not to differences between peoples. An Indigenous account of the conquest preserved in the Florentine Codex also gives smallpox a role. The Getty Research Institute's introduction to its digital edition says the account describes the Spaniards' first expulsion from Tenochtitlan and then Indigenous coalition warfare and smallpox before the city's final surrender.
So Sheppard's verdict was that even if the Aztecs or Inca had won the short-term battles, the demographic crisis would have eroded their ability to hold together and resist. Subject peoples would have "broken for freedom" as their overlords weakened, following what he called the overarching cycle of pre-Columbian societies. He pointed to Teotihuacan, the earlier great city of Mesoamerica, which rose, drew a huge population, overstressed its environment and collapsed. He thinks the Aztecs would have gone the same way: if Cortés had arrived in 1619 rather than 1519, Sheppard said, he would have found a swamp with moss-covered mounds where the great city had been. Sheppard did not frame this as a criticism. He called these societies more impressive given their disadvantages, building pyramids with human labor alone, without draft animals or the wheel.
Europeans would then have settled on the coasts and pushed inland more slowly, he said. Relations with Indigenous peoples might have been more balanced, but "in the longer term, they all would have been folded into the grand European imperial framework." This is Sheppard's judgment about an unrun experiment, not an established finding.
The Viking exception
Sheppard named one change that he thinks could have given pre-Columbian peoples a real long-term chance, and it lies about 500 years before Cortés. The Norse in Greenland under Leif Eriksson knew of North America, explored lands they called Helluland, Markland and Vinland, and tried to settle. They were too few and were driven out by local resistance. A 2021 study in Nature used tree rings in wood cut with metal tools to date Norse activity at L'Anse aux Meadows in Newfoundland to the year 1021, about five centuries before Cortés landed. The study describes the occupation as brief.
If the Norse had stayed, Sheppard suggested, they might have slowly passed European technology, pack animals and, above all, European diseases into the Americas. Then when the conquistadors arrived, they would have met people who had some resistance to European diseases, knew how to work iron and rode horses. "If that was the case," Sheppard said, "all bets are off."