Definition
An empire is a political order with two defining traits: it rules a substantial diversity of peoples — each with its own language, customs, and territory — and it has flexible borders capable of indefinite expansion. Empires are not simply large kingdoms. They are projects of integration that absorb new peoples into a single administrative, legal, and often cultural frame.
Yuval Noah Harari treats empire as one of three universal orders — alongside money and religion — that pulled scattered Sapiens into a single human network. Tom Head and Workman's Big Fat Notebook trace the same form across five thousand years of concrete cases: from Sargon of Akkad through Rome, Persia, Tang China, the Mongols, the Mughals, the Ottomans, Napoleon, and the European overseas empires. Read across all three books, empire is not a Western artefact, a modern aberration, or a fixed institution. It is the most common large-scale political form humans have ever built, and the cultural pools we call "national traditions" are mostly sediment left behind by past empires.
Why it matters
How it works
The universal-order frame (Harari)
In Harari's reading, empire is the second of three "universal orders" — patterns of organisation big enough to absorb strangers under one frame. Money is the first such order, religion the third, and empire sits between them as the political machinery that physically integrates territory. An empire combines three ingredients: military superiority to open the territory, bureaucracy to hold it, and an imagined order — religious, dynastic, civilisational, or ideological — to explain to ruler and ruled alike why the arrangement is legitimate. The Spanish patriots who venerate the Numantian rebels do so in Spanish, a Latin-derived language imposed by the very conquerors they celebrate resisting. Indian nationalists drafted independence in English. The pattern Harari names is that the imperial idea wins the deeper battle even when specific empires lose theirs — the language in which you formulate your protest is itself evidence of what already shapes you.
From city-state to imperial machine
The earliest documented imperial transition is the move that has repeated everywhere since. Sumer's loose federation of city-states (Ur, Uruk, Lagash, Eridu, Kish) lasted roughly three thousand years — longer than any single empire that has ever existed — until Sargon I of Akkad consolidated them by force around 2300 BCE into what is generally counted as the first empire. The same pattern recurs in ancient China's Warring States, ending with Qin Shi Huang's unification in 221 BCE; in Greece, where Philip II and Alexander broke the autonomy of the poleis; and in the Italian peninsula, where Rome absorbed its city-state neighbours before turning outward. Decentralised cooperation is more durable than centralised rule but more vulnerable to a single defector with the right ambition. Once a Sargon or a Qin or a Philip exists, the federation cannot hold.
Tolerance vs. terror — two operating models
Empires diverge sharply on how to manage diversity, and the divergence tracks closely with how long they last. The Neo-Assyrian model maximised short-term extraction: deportation of conquered populations, public flaying, ritual mutilation, and the brutal-uniformity approach that Ashurnasirpal II openly advertised in royal inscriptions. The Achaemenid Persian model, organised on Zoroastrian principles that treated truth and not orthodoxy as the cosmic test, used the satrapy system to leave local elites, laws, and faiths in place — and it scaled across more territory than any previous state. Cyrus the Great's release of the Israelite captives in 539 BCE was so unusual that the Hebrew Bible names him a messiah. The Ottoman millet system organised subjects by religious community rather than ethnicity, giving Orthodox Christians, Armenians, and Jews their own courts under a Muslim sovereign. The Mughal Empire ran the experiment at maximum contrast within a single dynasty: Akbar's inclusion produced a golden age; his great-grandson Aurangzeb's forced conformity broke the empire within two generations. Across all these cases the lesson scales: empires that accept lower short-term extraction in exchange for cultural autonomy outlast empires that try to homogenise their subjects.
Bureaucracy and standardisation
No empire scales beyond a single charismatic founder without administrative machinery. Qin Shi Huang standardised script, currency, axle widths, weights, and measures — making centralised administration possible and local identity weaker at one stroke. Hammurabi posted his code in stone so that citizens could read the law in advance, replacing rule-by-decree with rule-by-published-text. The Persians built the Royal Road and a postal system. The Romans built aqueducts, military roads, and the legal vocabulary of senate, republic, and citizen the Western world still uses. The Mongols ran the yam — relay stations every twenty to thirty miles where couriers swapped horses — and turned a pastoral confederation into a continent-spanning communication network that moved messages at two hundred miles a day. The Byzantines preserved Justinian's Corpus Juris Civilis and codified Roman law into a form that shaped later European legal systems. In every case, the bureaucracy outlives the dynasty: Han China kept the Qin playbook, every later Chinese unification borrowed it, and the modern Chinese state still uses administrative concepts traceable to Qin reforms.
The marriage of empire and science
Harari's argument in The Marriage of Science and Empire is that European empire and modern science were not parallel developments but a single geared mechanism. Captain Cook's 1769 voyage to observe the transit of Venus from Tahiti was simultaneously an astronomical expedition and a colonial reconnaissance: the Royal Society paid the astronomer, the Royal Navy paid the warship, and the crew measured the heavens, mapped the coastline, claimed the land for Britain, and brought back the data that powered both the next paper and the next conquest. Empire funded expeditions; science delivered maps, longitudinal navigation, vaccination, ballistics, and an ideological vocabulary in which conquest could be told as discovery; conquest refilled the science pipeline with specimens, data, and revenue; revenue funded more expeditions. Jenner's smallpox vaccine, invented in 1796, was administered — often forcibly — across the British Empire within a century: a humanitarian gift in one frame and a tool of soft conquest in another. The frames are not contradictory; they are how the engines geared together. A purely scientific Europe could not have funded the Endeavour; a purely militaristic Europe could not have crossed the Pacific and survived.
Cultural exhaust outlives political collapse
Empires fall, but their cultural output tends to outlast their political life by centuries. Alexander's empire fragmented the moment he died in Babylon in 323 BCE at age 32; the Hellenistic state lasted barely a year, the Hellenistic culture — Greek as the lingua franca from Egypt to Bactria, the city-grid template, the Library of Alexandria's research model — outlasted him by half a millennium. The Western Roman Empire ended formally in 476, but the Eastern Empire kept the brand for another thousand years as Byzantium and only fell to the Ottomans in 1453. The Byzantine diaspora then carried Greek manuscripts of Plato, Aristotle, and Euclid into Italy in the decades before and after the fall, pouring classical learning into the Renaissance just as it could be absorbed. Napoleon's empire collapsed in a decade but the Napoleonic Code, the metric system, the abolition of feudal privileges, and the administrative model survived him across most of continental Europe and eventually shaped about a quarter of the world's legal systems. The Mongol political union fragmented within a century, but the Golden Horde shaped Russian taxation for two hundred years and the breakup khanates still outline the political map of modern Eurasia.
Why empires fall — succession, overstretch, exhaustion
Few empires fall to a single battle. Most fail because of cumulative structural pressures, and the most common one is succession. Alexander died at 32 without a clear heir; his Diadochi generals split the empire within a year. Genghis Khan trained his sons and daughters, but his grandsons fought a civil war and the unified Mongol empire fractured into four khanates. The Roman pattern — strong emperor, weak heir, civil war — repeated across the Imperial Crisis of 235–284 and then again across the slow Western collapse. The Mughal succession turned the Akbar tolerance experiment into the Aurangzeb conformity experiment. Beyond succession, empires exhaust their own material base: constant border defence, infrastructure maintenance, and a revenue model originally built on plunder fail when expansion stops, as Rome discovered when it could no longer pay its bills in real money. And the troops an empire trains to suppress its subjects often defect — late Rome's Germanic legionaries took their Roman training and equipment with them when they overthrew the state that trained them.
The European overseas empires and decolonization
The European maritime empires from 1500 to 1900 represent the imperial form at its most ambitious and its most asymmetric. The European Big Fat Notebook topics on Colonies in the Americas, European Imperialism, and the Scramble for Africa name the cost honestly: every acre claimed was already home to Indigenous nations; the Spanish conquest of the Aztecs and Inca succeeded less through superior soldiers than through epidemic disease and timely alliances with local rivals; the colonial system rested on seized land, mass death, and forced labour. In Asia, by contrast, Europe was the weaker party — Ming and Qing China dictated terms at Macau and Canton; Tokugawa Japan admitted only the Dutch through a single watched window at Deshima. Then in a single century, between 1900 and 2000, that whole imperial geography collapsed. The British Empire shrank from 13.7 million square miles to the 122,000-square-mile British Isles. The combined pressures — two world wars bankrupting the colonial powers, industrialisation and nationalism spreading to subject populations, US and Soviet opposition to the old order — made independence over-determined, but the path varied wildly: India's negotiated partition, Algeria's eight-year war, the Congo's assassinations and proxy wars. The borders the departing administrators drew (Sykes–Picot, the British Mandate for Palestine, the Berlin Conference's African lines) still organise most of the world's frozen conflicts today.
Nationalism — the empire's unintended successor
The political form that replaced empire across most of the world — the nation-state — is itself partly an imperial inheritance. Napoleon's most enduring legacy may be the one he least intended: by dissolving the Holy Roman Empire, consolidating petty German principalities, and forcing the peoples he conquered to fight French wars, he gave subject populations the political vocabulary to organise against him as nations. Spanish, Italian, and especially German national consciousness sharpened in opposition to French occupation. The Ottoman millet system worked for centuries as a pragmatic pluralism — and then broke apart in the late nineteenth century when European-style nationalism reached the empire's minorities and insisted on a single people, a single language, a single state. The post-1945 wave of decolonization produced more than a hundred new UN member states, each one a nation in the European sense — borrowing the imperial form's administrative units, written law, and lingua francas to organise resistance to it. The nation-state, in this reading, is what an empire becomes when its centre collapses but its administrative geography survives.