Zhu Yuanzhang's Atrocities and the Brutal Logic Behind Ming Dynasty Rule

A boy too poor to bury his own parents grew up to kill the men who built his empire, and he did it with paperwork. Zhu Yuanzhang founded the Ming dynasty in 1368, ruled for thirty years as the Hongwu Emperor, and in that time executed somewhere between thirty thousand and one hundred thousand people, most of them officials, generals, and scholars who had served him loyally. The atrocities were not incidental to the dynasty-building. They were the method of it.

Zhu Yuanzhang's Origins and the Making of a Paranoid Ruler

The plague arrived in Fengyang in 1344, when Zhu was sixteen years old. It killed his father, his mother, and at least one of his brothers inside a matter of weeks. The family owned nothing. There was no money for coffins, no cloth to wrap a body in before putting it in the ground. A teenager stood over his dead parents with no proper way to bury them, and that specific humiliation, not poverty in the abstract but the inability to perform the most basic human obligation, is the biographical fact that historians keep returning to because it explains the quality of his rage more precisely than anything else.

He walked to the Huangjue Temple in Fengyang because monks ate. When the monastery ran short of food, it pushed its newest member out the door, and Zhu spent roughly three years wandering the roads of Henan and Anhui as a beggar-monk, sleeping where the night caught him, watching which villages had grain and which officials skimmed the famine relief. He came back to the temple around 1347 and stayed another four years, and during that second stretch something happened that bent the course of Chinese history: he learned to read.

A peasant farmer's son had no road to literacy. The monastery handed him one. He worked through the Buddhist texts and absorbed their whole vocabulary of purity and corruption, clean order and rotten disorder, and he carried that vocabulary out the gate in 1352 when he joined the Red Turbans, the Buddhist millenarian rebellion against the Mongol Yuan dynasty. The Red Turbans were not merely anti-Mongol. They were apocalyptic, organized around the conviction that the corrupt old order had to come down so a clean one could rise from the rubble. Zhu stepped out of a monastery that had trained him to divide the world into the pure and the corrupt, and straight into a movement that wanted to put the corrupt half to the torch.

By 1368, after two decades of outmaneuvering rival warlords across the wreckage of the collapsing Yuan, the orphan beggar had become emperor of China. He named his reign Hongwu, meaning "vastly martial." The Red Turbans had given him the ideology. The monastery had given him the literacy to write it into law. The three years of begging had given him a permanent contempt for comfortable, educated men who had never once been hungry. All three fused into a ruler who could compose imperial edicts that dripped with moral authority while signing off on deaths by the thousands, and who genuinely believed both acts were the same sermon.

What the begging years had also done was teach him exactly one lesson about loyalty: it runs toward power, not toward principle. Every village headman who fed the relief funds into his own pocket while people starved at the roadside had been someone's loyal subordinate. Every monastery that turned out its own novices when the grain ran low had been a community bound by vows. Zhu concluded, and never revised the conclusion, that the only loyalty worth trusting was the loyalty of a man who owed his entire standing to you personally and had nothing to fall back on if you withdrew your favor. A general beloved by his troops, a scholar respected for his learning, an old comrade with his own history and his own followers: each of these was a man whose loyalty did not run entirely through Zhu, and that made each of them a threat, regardless of their actual intentions.

The Hu Weiyong Purge Against the Scale of Earlier Chinese Political Massacres

The twelfth of February, 1380, is the date that outweighs every other in the Hongwu reign. On that morning Zhu had his Grand Chancellor, Hu Weiyong, executed alongside the Censor-in-Chief Chen Ning. The official charge was a coup: Hu had allegedly conspired with Mongols and Japanese pirates to assassinate the emperor. Modern historians regard the charge as almost certainly fabricated, the sort of confession produced when the interrogator already knows what answer he needs.

The execution itself was not the point. What followed was unlike anything in the prior thousand years of Chinese dynastic politics. Zhu reopened the investigation again and again across the next decade, each reopening dragging in a fresh ring of people through a logic of guilt by association applied with recursive patience. The accused named their associates. The associates named their patrons and protégés. The patrons named their relatives. Each wave named the next, the circle widening for ten unbroken years.

By the time it burned itself out, the dead numbered between thirty and forty thousand. That figure comes from Edward Dreyer, who completed his doctorate at Harvard under John Fairbank and Lien-sheng Yang, two foundational figures in Western scholarship on imperial China. Dreyer became the leading authority on the military dimensions of the early Ming, and his work forms the basis of the Cambridge History of China's account of the entire period. Some reconstructions push the total past fifty thousand once the full decade of follow-on waves is included. That range matters because it reflects genuine scholarly uncertainty, not false precision.

No previous Chinese dynasty had produced a purge on this scale from a single accusation. The Han, Tang, and Song dynasties had all seen court bloodletting, factional massacres, and the elimination of rival power blocs. Liu Bang, the founder of the Han, had eliminated his major military allies after securing the throne. But those earlier purges tended to be concentrated, relatively swift, and aimed at specific rivals. Zhu's innovation was the decade-long reopened file, the recursive expansion of liability outward through every social network connected to the original accused. A historian associated with the China Project described the 1380s purges as the most horrendous bloodbath of civilian violence the Chinese state had yet produced. The mechanism was genuinely new: killing as a clerical operation, sustained across years, under the continuous cover of legal proceedings.

The slowness was the genius of it. A ruler who killed forty thousand in a single afternoon would have triggered a rebellion. A ruler who killed them a few hundred at a time, over a decade, under the paperwork of formal judicial process, met almost no resistance at all.

The Four Major Cases and Zhu Yuanzhang's Phased Dismantling of Ming Power Blocs

The Hu Weiyong purge was not an isolated eruption. Historians of the early Ming organize the reign's violence into the Four Major Cases, a framework that reveals the underlying strategic logic. Each purge targeted a distinct social constituency, separated by years, in a phased dismantling of every power bloc capable of challenging the throne.

The Hu Weiyong case of 1380 was the first and largest, aimed at the civil bureaucracy and the ministerial class. The Empty Seal Case, triggered by a procedural dispute over official seals, spread across thirteen provinces and into the censorial apparatus, demonstrating that the bureaucracy itself, not just individual ministers, was a target of systemic elimination. The Guo Huan case of 1385 struck at regional governors and officials who held significant local authority. The Lan Yu case of 1393 completed the sequence by destroying the military aristocracy.

The Lan Yu case deserves separate attention because of what it eliminated and who it killed. Lan Yu was one of the Ming's most decorated commanders, the general who had driven the Mongol forces deep into the steppe and won the dynasty's most significant northern victories. His execution in 1393 on treason charges triggered a purge that swept through the officer corps with the same recursive logic the Hu Weiyong case had applied to the civil administration. The precise death toll for the Lan Yu purge is harder to verify than for the Hu Weiyong case, and writers who cite a specific figure should know that solid secondary scholarship on the exact number is thinner than it appears in popular accounts.

What the Lan Yu case accomplished, beyond the killing, was the final elimination of any general with an independent military reputation. Zhu was by then in his sixties, preparing the succession. A throne passing to a young heir needed an officer corps that owed its commissions entirely to the throne and had no battlefield legend of its own to trade on. The purge delivered exactly that, at the cost of gutting the military leadership that had made the dynasty.

The fate of Li Shanchang sits at the center of this logic. Li was one of Zhu's oldest comrades, a founding strategist of the dynasty, a man who had been at the emperor's side since the rebel years. By the 1390s he had long since withdrawn from active power. He was seventy-seven years old. He posed no operational threat to anyone. In 1390 Zhu ordered him to take his own life; more than seventy members of his family were executed alongside him. The target was not a conspirator. The target was a reputational memory, a living symbol of the old coalition that Zhu needed to erase before he died. Li Shanchang's crime was that people remembered him.

The Jinyiwei, the Zhenfusi, and Zhu Yuanzhang's Architecture of Surveillance

Before the purges could run at scale, Zhu needed an instrument that operated outside the normal judicial system. The Jinyiwei, the Embroidered Uniform Guard, was that instrument: a political secret police answering solely to the emperor, with its own prison, its own interrogation apparatus, and no obligation to the ordinary courts.

On paper the Jinyiwei began as an imperial bodyguard. By 1382 Zhu had transformed it into something the Chinese state had not formally fielded before. Its agents reported on ministers, generals, and scholars. They monitored private dinners. They ran informant networks through the capital and into the provinces. The Zhenfusi, the Jinyiwei's private prison, sat entirely outside the normal judicial process: a suspect could be held there without reference to ordinary law, interrogated by methods the regular courts were not supposed to use, and kept indefinitely at the emperor's discretion.

What made the system dangerous was not the cruelty, which any jailer can supply. It was the reporting structure. Every denunciation went directly to Zhu, bypassing every institutional filter that might have slowed or questioned it. Any official could be named by any other official. Any private conversation could find its way back around. If you served the Ming under Hongwu, you governed with the working knowledge that the man at the next desk might be writing down your name tonight.

The Jinyiwei outlasted Zhu. Later Ming emperors expanded the model into the Dongchang, the Eastern Depot, a eunuch-run agency with similar powers of surveillance and arrest. The institutional template Zhu built in the 1380s survived him by nearly two centuries, which is itself a measure of how effectively he had encoded paranoid centralism into the dynasty's operating structure.

Zhu Yuanzhang's Agrarian Reforms and the Nation-Building Dimension of the Hongwu Reign

Zhu's agrarian reforms were as central to the Ming project as the purges, and they succeeded on their own terms. The same emperor who killed tens of thousands of officials also redistributed land to landless peasants, banned the private slave trade, planted fifty million trees around Nanjing, and rebuilt the irrigation networks that Mongol-era neglect had allowed to collapse.

Land redistribution after the Ming founding was systematic. Zhu confiscated estates from wealthy landowners and reallocated them to peasants who had no land. Conditional grants required recipients to personally work the plots, preventing the immediate reconsolidation of large holdings. Those who held more land than they could cultivate had the surplus confiscated. A policy exempting newly cultivated land from taxation for three years gave farmers a direct incentive to bring marginal and abandoned acreage back into production. By 1393, cultivated land had reached over 8.8 million measurement units, a figure preserved in the Veritable Records that reflects the scale of agricultural recovery under his reign. That number matters because it represents a measurable reversal of the Mongol-era agricultural collapse.

The Yellow Registers and Fish Scale Records, systems of household and land registration Zhu introduced, were simultaneously administrative tools and surveillance instruments. They gave the state detailed knowledge of who farmed what land, who owed what tax, and who belonged to which household, making the countryside legible to the imperial center in a way the Yuan had never achieved. The labor supply was expanded by abolishing private slavery outside the imperial family and banning the purchase and sale of free people, which preserved agricultural workers for the fields rather than elite households.

John Dardess, in his study of 128 collected works from the period 1340 to 1400, complicates the standard victim-and-tyrant framing. Confucian scholars, Dardess argues, were not simply the targets of Zhu's autocracy. They were active participants in constructing its intellectual machinery, helping to design the administrative structures and legitimizing ideology that made the regime function. Zhu aligned himself with the Confucian class to present the Ming as a restoration of orthodox Chinese civilization after the Mongol interregnum, and the scholars accepted that role, even as Zhu simultaneously censored Confucian texts that challenged absolute rulership, removing passages from Mencius that implied limits on imperial authority.

The agrarian program was real, and it worked. The dynasty's early decades saw genuine recovery in food security, population, and agricultural output. But the recovery came packaged with tight controls on mobility, hereditary occupational categories, and a surveillance apparatus that treated the household as a unit of governance. Peasants got land. They also got registered, monitored, grouped into mutual-responsibility units, and subjected to the same ideological imperatives that governed the court.

Revolutionary Leaders Who Destroyed Their Own Coalitions: A Cross-Cultural Pattern

Zhu Yuanzhang won the throne through coalition warfare, then spent thirty years systematically killing most of the people who had made that coalition work. This pattern, the revolutionary leader who destroys his own revolutionary cohort after consolidating power, recurs across cultures in ways that suggest it reflects a structural feature of how certain kinds of regimes stabilize rather than a uniquely Chinese or uniquely personal pathology.

Stalin's Great Terror of 1937 to 1938 eliminated the old Bolshevik generation, the men who had actually made the revolution, in favor of a younger cadre whose careers Stalin had personally built. Oliver Cromwell governed through punitive force, suppressed opposition in Ireland and Scotland with considerable violence, and purged the Parliament that had made the Civil War possible. The mechanisms differ, the scales differ, and the ideological justifications differ. But the underlying logic is consistent: a founder who came to power through a coalition of independent actors finds, once power is secured, that the same independence that made those actors useful during the struggle makes them dangerous during the peace.

How Does Zhu Yuanzhang's Purge Pattern Compare to Stalin's and Cromwell's?

Zhu is closer to Stalin in method than to Cromwell in scope. Both used serial purge logic, accusations of conspiracy, and recursive guilt-by-association to clear out the founding generation. But Zhu was more focused on erasing the meritorious founding class specifically, the men whose independent reputations predated his own rule, whereas Stalin's terror swept far more broadly through Soviet society. The Chinese historian who described Zhu's killing of meritorious officials as unprecedented in Chinese history was pointing at exactly this feature: not the scale alone, but the targeted elimination of the people whose achievements had made the dynasty possible.

Did Zhu Yuanzhang's Purges Kill More People Than Stalin's Great Terror?

Probably not, on the most commonly cited estimates. Stalin's Great Terror produced roughly 750,000 executions in 1937 and 1938 alone, a figure drawn from post-Soviet archival research. Zhu's purges across his entire thirty-year reign reach a maximum estimate of around 100,000 in the most expansive reconstructions, with the most carefully sourced figure, Dreyer's estimate for the Hu Weiyong case alone, sitting at thirty to forty thousand. Zhu's purges were extraordinary for a premodern empire but do not approach the industrial scale of Stalinist repression. What makes Zhu historically significant is not that he killed more than Stalin, but that he killed more systematically, over a longer period, with more elaborate bureaucratic cover, than any Chinese emperor before him.

Was Zhu Yuanzhang's Violence More Bureaucratic Than Stalin's?

Both were bureaucratic killers. Zhu's violence was organized through state institutions, documented with formal written accusations, and conducted under the continuous cover of legal proceedings. He kept paper trails for everything. The purges were, in a precise sense, a clerical operation. Stalin's apparatus was also heavily bureaucratized, with NKVD quotas and mass operations run through a modern party-state. Zhu's violence was intensely personalized and court-centered, while Stalin's was impersonal and industrial in scale. Zhu signed off on individual cases. Stalin issued quotas by region. Both methods produced mass death; the administrative texture was different.

What Did the Ming Shilu Record, and Can Historians Trust It?

Historians use the Ming Shilu, the Veritable Records of the Ming dynasty, with careful skepticism: it is indispensable and compromised at the same time. The records were compiled after each emperor's death from court diaries, audience records, and government documents, and they preserve a level of daily chronological precision that scholars value highly. But they were compiled by the very autocratic machinery Zhu built, meaning the sources used to measure his violence were curated by a state whose legitimacy depended on projecting his omnipotence.

Whether the numbers are inflated to magnify imperial reach or suppressed to minimize embarrassment is a question the records themselves cannot answer. Modern historians cross-check the Shilu against local gazetteers, private writings, inscriptions, and later histories. The result is that the broad scale of the purges is well established, but precise figures should be treated as approximations rather than verified counts. When this article says thirty to forty thousand died in the Hu Weiyong purge, that figure comes from a serious scholar working from the best available evidence. It does not come from a neutral ledger.

How Did Zhu Yuanzhang's Domestic Surveillance Reach Inside the Family Household?

Zhu's surveillance apparatus extended well beyond the political elite into the structure of ordinary family life, treating the household itself as a basic unit of state governance, taxation, morality, and security. The Jinyiwei watched the court. The household registration system watched everyone else.

Families were organized into mutual-responsibility units of ten households, then one hundred, with wealthier families made responsible for ensuring the others met their tax and occupational obligations. This created peer pressure and local monitoring inside domestic life without requiring the Jinyiwei to be everywhere at once. Every household was supposed to possess copies of imperial moral proclamations, meaning state ideology entered the home rather than remaining at the level of officials and magistrates. Movement was restricted at night by curfew systems that confined people to their residences during specified hours.

The Dan'elu, a legal instrument examined in scholarship on Zhu's attempts to regulate interfamilial conflicts, reveals a dimension of his authoritarianism that accounts focused on court purges tend to miss entirely. Zhu conceived of family morality as a state concern. Interfamilial disputes, conflicts between relatives, failures of filial obligation, were not private matters to be resolved within the household. They were potential violations of the moral order the emperor had defined, and they carried punitive consequences. The family was not a refuge from imperial power. It was one of its primary administrative units.

Did the Dan'elu Apply to Peasant Families or Only to the Elite?

The Dan'elu's punitive reach applied primarily to elite households rather than peasant families as a universal obligation. Zhu's state differentiated between common farming households and socially privileged families, with special controls falling more heavily on those with status, wealth, or official connections. Peasant families were the base of the tax system and were assessed as farming households under the Yellow Register framework. Elite families, landlords, local notables, and households with inherited status faced a different layer of regulatory scrutiny. The domestic surveillance architecture was not flat: it pressed hardest on those whose private behavior carried the greatest potential for political consequence.

How Did the Abolition of the Grand Chancellorship Destabilize the Late Ming?

After Hu Weiyong's execution in 1380, Zhu did not merely sack a minister. He deleted the position of prime minister from the Chinese state, dissolved the Central Secretariat that had administered Chinese civil government for nearly a thousand years, and decreed that any future emperor who tried to revive the post, or any official who proposed doing so, would be executed. In the short term this was an act of extraordinary administrative consolidation; in the long term it was the structural flaw that contributed to the dynasty's eventual collapse. Every strand of executive authority that had been distributed across a sprawling government now ran into one pair of hands.

The immediate effect was total centralization. Zhu was energetic enough to manage the workload directly, and the terror he had established ensured that no one in the bureaucracy was going to complain about the arrangement. But subsequent emperors inherited a system with no institutional buffer between the throne and the machinery of empire. The Grand Secretariat that later Ming rulers developed as an advisory body lacked the formal authority of the old chancellorship. It could not reliably discipline ministers, coordinate the central bureaucracy, or resist palace politics. Its power was unstable, sometimes strong and sometimes weak, depending entirely on the emperor's personal engagement with governance.

In the late Ming, this structural weakness became critical. The bureaucracy fractured into cliques and factions that the center could not discipline. Emperors who were disengaged, incompetent, or simply overwhelmed by the volume of decisions that now ran directly to the throne created governance vacuums that eunuch officials filled. The powerful eunuch bureaucracy that historians identify as a major factor in the dynasty's destabilization was not an accident or a cultural pathology. It was the predictable consequence of eliminating the only institutional role capable of coordinating the central government without the emperor's direct involvement.

Did Any Ming Emperor Attempt to Restore the Grand Chancellorship After 1380?

No Ming emperor restored the Grand Chancellorship across the dynasty's two hundred and seventy-six years. The prohibition Zhu encoded into law held for the entire dynasty, which is itself a remarkable fact: the decree carried enough terror that no successor, across fourteen emperors and nearly three centuries, was willing to challenge it. The Grand Secretariat developed as a functional substitute, but it was never granted the formal authority of the old chancellorship and was never described as equivalent to it in official sources. The Ming Shilu records the original abolition decree; it records no attempt to reverse it.

Were Confucian Scholars Victims of Zhu Yuanzhang's Autocracy or Its Architects?

The standard framing puts Confucian scholars on the receiving end of Zhu's violence, and the violence was real: he had officials beaten at court, executed scholars who criticized his rule, and removed passages from Mencius that implied limits on imperial authority. But John Dardess's analysis of 128 collected works from the period 1340 to 1400 demonstrates that Confucian scholars were active co-constructors of Ming autocracy, not simply its victims.

Zhu needed the scholar class. He needed their administrative competence to run the empire, their ritual knowledge to legitimize the dynasty as a restoration of orthodox Chinese civilization after the Mongol interregnum, and their literary skill to produce the documents, proclamations, and legal codes that gave the regime its intellectual architecture. The scholars, for their part, accepted the arrangement. They helped design the institutions, wrote the legitimizing texts, and staffed the bureaucracy. They retained influence precisely insofar as they served the emperor's state-building goals. The partnership was real, and it was unequal, and when a scholar's learning or status or criticism threatened imperial control, the partnership ended immediately.

Zhu was also constructing his own historical memory in real time. His self-authored proclamations, circulated throughout the empire and into every household, were deliberate instruments of ideological self-presentation. The Huang Ming Zuxun, the Ancestral Injunctions, were issued in 1373 and revised in 1395, attributed to Zhu's direct authorship in imperial cataloging. The introductory chapter, at minimum, was composed by the emperor himself, urging his descendants to govern in a strict Legalist manner. A man who understood that controlling historical memory was itself a form of power, and who built the institution, the Ming Shilu, that would record his reign, was not simply a violent ruler. He was a sophisticated propagandist who grasped that the narrative of the purges mattered as much as the purges themselves.

Did Zhu Yuanzhang Write the Huang Ming Zuxun Himself?

Partly. Zhu is credited with composing the preface and authorizing the Ancestral Injunctions, while the full text was a court-issued compilation refined across multiple editions rather than a single self-authored work. The traditional catalog attribution lists him as the author, and at least the introductory chapter reflects his direct voice. But the work was first composed between 1369 and 1373 and then revised and updated in 1395, which implies a process of court compilation and editorial development over time. The practical distinction matters: Zhu understood the ideological function of the document clearly enough to authorize and shape it, which makes him its intellectual author even if he did not draft every line himself.

Where Did Zhu Yuanzhang's Agrarian Reforms Succeed and Where Did They Fall Short?

The reforms succeeded where they targeted immediate recovery from wartime collapse and fell short where they imposed structural rigidity on the economy they had revived. Land redistribution, tax relief, and resettlement of war-devastated regions produced genuine, measurable gains in agricultural output and peasant welfare in the late fourteenth century. The same system's tight controls on mobility and hereditary occupational categories limited the long-term flexibility of the rural economy.

Cultivated land expanded substantially under the Hongwu reign. Wasteland reclamation incentives worked. The three-year tax exemption for newly cultivated land brought marginal acreage into production. The ban on private slavery preserved agricultural labor for the fields. Refugees and displaced persons were resettled on vacant land in underpopulated regions. These were not cosmetic policies. They addressed real structural damage left by decades of war and Mongol-era mismanagement, and the recovery in food security and population that followed was genuine.

Where the system broke down was in its insistence on control. Peasants gained land but lost the freedom to move, to change occupations, or to accumulate in ways the state had not sanctioned. Commerce was deliberately suppressed in favor of agriculture. The rural economy that emerged from Zhu's reforms was more productive than what the Yuan had left behind, but it was also more rigid, more surveilled, and less capable of adapting to changing conditions. The same administrative architecture that registered every household and every plot of land for equitable taxation also locked people into categories from which there was no legitimate exit.

The constructive and the coercive were not separable dimensions of the Hongwu reign. They were the same project, pursued through the same instruments, by the same man.

Zhu Yuanzhang's Legacy: What the Atrocities Built and What They Broke

Zhu Yuanzhang died in 1398 in his own bed. He had outlived his eldest son and heir, outlived most of the generals and ministers who had built the dynasty with him, outlived most of the men he had ordered killed. The Ming dynasty he founded lasted two hundred and seventy-six years. The question of what his violence built and what it broke is not a moral question. It is an institutional one, and the answer is the same for both sides of the ledger.

The purges built a court in which no official, no general, and no scholar could accumulate independent political standing. Every person who served the Ming under Hongwu governed with the knowledge that their position was entirely contingent on imperial favor, that the Jinyiwei was watching, that a single denunciation could open a file that might not close for a decade. That knowledge produced the stability Zhu wanted. It also produced a bureaucracy that had learned, at the cost of tens of thousands of lives, that initiative was dangerous and deference was survival. The dynasty that followed him was stable and, in its middle decades, capable of genuine administrative achievement. From the beginning, though, it was structurally incapable of producing the kind of independent institutional check that might have managed the crises of the late Ming more effectively.

The abolition of the Grand Chancellorship is the single decision whose long-term consequences most clearly illustrate this. Zhu made it to ensure that no minister could ever accumulate the kind of independent authority that might threaten the throne. He succeeded. No Ming emperor faced a chancellor powerful enough to challenge him. What the dynasty faced instead, across its final century, was a governance vacuum at the center that eunuch factions filled, a factionalized bureaucracy that could not coordinate policy, and a succession of emperors who were either disengaged or overwhelmed by the administrative burden Zhu had concentrated in the throne. The institutional conditions for the dynasty's collapse were written into its founding documents.

The Ming Shilu problem sits underneath all of this and should not be footnoted away. The primary sources for Zhu's violence were compiled by the machinery of the state he built, edited by officials who served the dynasty he founded, and preserved because the dynasty found them useful. The thirty to forty thousand dead in the Hu Weiyong purge is the best scholarly estimate available. It is not a neutral count. Whether the actual number is higher or lower, the figure was recorded by a state whose interest in the narrative of its own founding was not disinterested. Zhu understood that controlling memory was a form of power. He built the institution that controlled the memory.

What the atrocities built was two hundred and seventy-six years of Ming stability, a dynasty that rebuilt Chinese agrarian civilization after Mongol-era devastation, produced the Forbidden City, and held the largest state on earth together for nearly three centuries. What they broke was every institutional mechanism that might have made the dynasty adaptable when the conditions that had sustained it changed. The Ming Shilu records no attempt by any of his fourteen successors to restore the Grand Chancellorship he abolished.