1Part Six, in four paragraphs
So this part stands on its own, even if you have not read the last one.
Part Six was about ladders and arrows. Every large society sorts people into higher and lower, and Harari's claim is that almost none of those rankings have any basis in biology. They begin in historical accidents — who arrived first, which crop needed picking, who survived which disease — and then get explained by a story that makes the accident look like nature. The machine that keeps them standing is the vicious circle. A myth causes discrimination, the discrimination produces poverty and ignorance, and the poverty and ignorance look like proof of the myth. The evidence people point at is real. The system manufactured it. That is why removing the original cause does not stop the machine, and why sincere people believe a hierarchy is deserved. He also gave us the sharpest sentence in the book: biology enables, culture forbids. Nobody bans the impossible, so every prohibition that exists is proof the thing it forbids can be done. And he admitted, honestly, that the one near-universal hierarchy — patriarchy — defeats all three of the standard explanations, and that he does not know why it is there. Then Chapter 9 stood back. Every culture is full of contradictions, and the discomfort is the engine that makes cultures move at all. And the direction they have been moving, over ten thousand years, is towards unity — driven by three new orders that could in principle include everybody: money, empire, and universal religion.
Two of the three glues, and in my view the two strongest chapters in the book.
Chapter 10 is money, and it is where I stopped resisting Harari's whole argument and admitted he had me.
Chapter 11 is empire, and it is the chapter he gets attacked over most — some of it fairly, some of it by people who have not read what he actually wrote. It is also the chapter where he uses India as his main example, which means I have to declare myself again. I do that in Section Seven.
2The problem that stops a village becoming a city
Barter has a flaw that sounds trivial and is actually fatal.
You grow apples. You want shoes. So you find the shoemaker. But he does not want apples — he wants a goat. So you find the goat farmer, who does not want apples either; she wants cloth. So now you need a weaver who wants apples, to get cloth, to get a goat, to get shoes. And every one of those people has to want what you are holding at the exact moment you want what they are holding.
The double coincidence of wants: for a swap to happen, each side must want what the other has, at the same time.
In a village of forty people trading six kinds of thing, that is manageable. You know everybody, you can wait, and you can owe each other favours.
Why it matters: it does not scale, and the reason is arithmetic rather than inconvenience.
Money solves it in one move. Everybody converts what they have into one thing that everybody accepts, and converts it back into what they want, whenever they want. The chain of four desperate people becomes two ordinary transactions.
3Money is not coins
The first thing to unlearn, and I had not unlearned it.
Harari's definition is deliberately wide: money is anything people are willing to use to represent, systematically, the value of other things, for the purpose of exchange. Anything. And across history it has been almost everything — shells, cattle, salt, cloth, beads, cigarettes in prisons and prison camps, and one of the earliest, barley. In Sumer around 3000 BC, the basic monetary unit was the sila, roughly one litre of barley. Wages, rents and debts were counted in it. And barley has one advantage nothing since has matched: if the whole system collapses, you can eat your money. It also has obvious problems. It rots, it is heavy, and its quality varies. So Mesopotamians moved to the silver shekel — and here is the detail that surprised me.
It was a weight: 8.33 grams of silver.
There was no stamp, no face, no denomination. You carried lumps of silver and you weighed them out on a balance, which is why merchants carried scales and why an honest set of scales appears as a moral image in so many ancient texts.
And silver has no nutritional value, no practical use for most people, and cannot be eaten, worn or burned for warmth.
Its entire worth came from everybody agreeing it had worth. That is the moment money becomes purely a story, and it is much earlier than I had assumed.
The first true coins came around 640 BC, minted by King Alyattes of Lydia, in what is now Turkey.
A coin is a lump of metal with two guarantees stamped into it: this is exactly this much, and it is exactly this pure. The stamp is an authority saying I vouch for this, so that you no longer need scales and no longer need to trust the person in front of you.
4What money actually is
Trust. And once you see that, several strange things about history stop being strange.
Money is trust. Harari calls it the most universal and most efficient system of mutual trust ever devised. That one line explains something I had never thought to ask about.
Across many societies and many centuries, counterfeiting was treated as one of the gravest crimes available — frequently as treason, frequently punished by death, at times when ordinary theft was punished with a fine or a whipping.
That looks absurd. Stealing a coin and making a coin both take one coin's worth of value from somebody.
But they are not the same act. Steal from a person and you have harmed a person. Counterfeit the currency and you have attacked the belief that holds the entire economy together.
You are not stealing money. You are damaging the story, and the story is what everybody's savings, wages and debts are made of.
Once you see money as an imagined reality, the punishment stops being disproportionate and starts being logical.
And how strong is that trust? Consider the Roman denarius.
It circulated so widely and was trusted so completely that rulers well outside Roman control — in Arabia, in India — minted imitation denarii stamped with the face of a Roman emperor who had no authority over them whatsoever.
Think about what that means. Their own coinage, issued by their own legitimate authority, was less trusted by their own merchants than a copy of a foreigner's. So they copied the foreigner.
That is the power of a story that has spread far enough, and I find it genuinely astonishing.
Everybody in this chapter — Harari, the textbooks, most economists — assumes the thing that came before money was barter.
A great many anthropologists think that is a myth. The evidence for societies actually running on barter before money is thin. What the record more often shows is credit: informal, remembered obligations inside a community. You take what you need from your neighbour, everybody remembers, and it evens out over years or over a lifetime. Barter, where it appears, often shows up after a monetary system has broken down — which is the opposite of the standard story.
Underneath the barter story is a deeper assumption nobody states: that the basic unit of economic life is a transaction — a discrete, simultaneous, settled swap between two people who then owe each other nothing.
For most of human history the basic unit was probably the opposite. An ongoing obligation between people who expected to still be there next year, and who never fully settled up, because not settling up is what a relationship is.
If that is right, then money did not solve the problem of exchange. It solved the problem of exchange between strangers — and in doing so it made the stranger the normal counterparty, which is a much bigger and stranger claim than the chapter makes.
5Everything into everything
Money's real magic and its real poison are the same property.
Money is a universal translator for value. Land into education. Time into food. Labour into a house. Your skill as a farmer into your child's schooling — which without money requires an impossible chain of favours across a decade, and with money is two transactions. But universal convertibility does not stop at the boundary you would like to draw around it. Once everything has a price, everything becomes purchasable, including things most societies say should not be for sale. Land a family held for six generations. Honour. Political influence. Justice. Medical care. And people — Part Ten is about a trade in human beings that was organised, insured and financed exactly like a shipping business, because that is what it was. Money does not respect any boundary you draw. It cannot; that is the whole design. And communities that ran for centuries on obligation, kinship and reputation get quietly dissolved when everything becomes a transaction — because a transaction ends when it is finished, and a relationship does not.
And the thing that surprised me most
Here is Harari's most unexpected claim in this chapter, and I think it is true. Money is the most tolerant thing human beings have ever created. Money does not care about your religion, your caste, your colour, your language, your gender, or whether the person who grew your food would be permitted to eat at your table. A coin from an enemy nation buys exactly as much bread as a coin from your own. He puts it memorably: Christians and Muslims who could agree on nothing whatsoever about God could agree completely about gold. Two people who would cheerfully kill each other over doctrine will trade all afternoon, honestly, and both walk away satisfied. That is why money spreads faster than religion or empire. It asks nothing of you except that you also want it.
I want to hold two things here at once, because Harari does and most people quoting him do not.
The achievement is real. Money has connected people across every hostile boundary our species invented, and a great deal of ordinary peace between strangers rests on it.
But notice what kind of tolerance it is. Money is not tolerant the way a person is tolerant. It does not respect you, understand you, or accept you. It is indifferent to you. It would treat you exactly the same way if you were a saint, a stranger or a slave trader, and it has.
Indifference does a lot of the work of respect and none of the work of solidarity. That distinction matters enormously in Part Ten.
And then the twist that stops the chapter being a celebration.
Money creates trust between strangers, and the trust is not in the strangers.
You hand cash to a shopkeeper you have never met and will never see again. You are not trusting him. You are trusting the note, and behind it a central bank, a government, a legal system, and millions of other people's continued belief. The human being in front of you is almost irrelevant to the transaction, which is precisely why it works with anybody.
So money built a bridge between strangers by removing the need for them to know each other at all. That is the achievement and it is also the loss, and Harari's summary is exact: it is the most universal system of mutual trust we have ever built, and it is not a trust in humans, communities or sacred values. It is a trust in money itself.
6What an empire actually is
Get the definition right first, because almost everybody uses the word loosely and Harari does not.
An empire has exactly two features. One: it rules over a significant number of distinct peoples, each with its own culture, identity and territory. Not one people with one language — many, held under a single authority. Two: it has flexible borders and a potentially unlimited appetite. It can swallow more without becoming a different kind of thing. A nation-state has a shape. An empire has an edge that keeps moving.
Nothing about size. The Athenian empire was small.
Nothing about having an emperor. Plenty did not.
Nothing about how it is governed at home. The British Empire was a parliamentary democracy in London while being something else entirely everywhere it ruled.
Nothing about cruelty. Cruel and less cruel empires are both empires.
Why it matters: it is a precise tool, and it lets you recognise empires that do not call themselves empires — which is most of them, most of the time.
And here is the fact that reframes everything: empire has been the world's most common form of political organisation for the last two and a half thousand years.
Not an aberration. Not a dark chapter between the normal ones. The default. For most of recorded history, most human beings alive were the subjects of an empire.
Which means the nation-state — the thing that feels eternal, the thing on your passport — is the recent experiment. It is a couple of centuries old in most of the world and has no guarantee of permanence at all.
7Empires absorb, and they always say it is for your good
Two features that sound like praise and are not.
They are culturally inclusive
This sounds wrong, because we associate empires with racism and brutality, and there was plenty of both. But the logic of empire pushes towards absorption. An empire that accepts only one ethnic group cannot expand past that group, and expansion is the entire point of the thing. So empires assimilate. They take conquered peoples, teach them the imperial language, give them imperial law, let them into the army, promote the useful ones — and after a few generations the descendants of the conquered are running the place. Rome did this to a remarkable degree. Peoples the Romans had fought as barbarians became Roman citizens, then Roman generals, then Roman emperors. "Roman" stopped meaning from Rome and started meaning participating in Roman civilisation.
And they always claim a universal mission
The first clear example is Cyrus the Great of Persia, in the sixth century BC. Earlier conquerors said something straightforward: I took your land because I wanted it and I was stronger. Cyrus said something new — that he ruled for the benefit of all peoples, that he brought order and justice, that his subjects were fortunate. He allowed conquered peoples their own customs and gods, and presented conquest as liberation. Every major empire since has said a version of it. Alexander spread Greek civilisation. Rome brought law and peace. Islamic empires brought the true faith. China civilised. Britain brought order, commerce and Christianity to people it described as requiring all three. The Soviet Union liberated workers. And all of them believed it, at least partly, which is the uncomfortable bit.
Harari's point is not that these claims were lies. It is that an empire needs a universal story, because ruling dozens of peoples requires a reason larger than we won.
So here is the tool. Anybody claiming to act for the good of all humanity is making an imperial claim, whatever they call themselves and however sincerely they mean it.
That does not make them wrong. Some things genuinely are good for everybody. But the claim itself is a specific move with a very long history, and it is worth noticing when it is being made — including when it is being made by people you agree with.
8The hard part
Harari's main example is India. I have to declare myself before I explain it.
am Indian, writing about the British Empire in India, and this is the second place in this series where my own position could pull the writing.
There are two well-worn ways of getting this wrong, and both are comfortable.
One is the ledger — listing railways against famines as though they cancel, which is a way of not looking at either. The other is treating the whole argument as an insult to be repelled, which is a way of not reading what was actually written.
So I will lay out Harari's argument as he makes it, name the atrocities at full weight rather than as a balancing item, and then tell you exactly which step of his reasoning I think does not hold. That last part is in Section Nine, and it is the sharpest disagreement I have with him in the whole book.
The argument
ILook at almost any culture today and ask where it came from. Consider what a modern Indian might name as part of national life: parliamentary democracy, the legal system, the civil service, the railway network, English as a language shared across states that share no other, cricket, tea as a mass drink — and the one that really lands, the idea of "India" as a single political unit at all. Every item on that list arrived in its present form through the British Empire. The subcontinent before that was a shifting field of kingdoms, sultanates, empires and principalities, with enormous shared civilisational inheritance and no single political identity. And the same empire caused catastrophes. Famines in which millions died under policies that prioritised revenue and export over relief. The deliberate destruction of Indian textile manufacturing, which had been among the world's largest. The extraction of enormous wealth over two centuries. Racial hierarchy written directly into law. And a partition in 1947 whose death toll runs from the high hundreds of thousands to over a million, with something like fifteen million people driven from their homes. Both lists are true. At the same time. About the same empire.
So Harari asks the uncomfortable question
What would it actually mean to reject the imperial legacy? Should Indians stop playing cricket? Abandon the parliamentary system? Stop using English between states that cannot otherwise talk to each other? And if they did — go back to what? To the Mughal Empire, which was an empire. To the Delhi Sultanate, which was an empire. To the Gupta Empire, which was an empire. Keep peeling and you never reach a layer that is not somebody's conquest. There is no authentic, pure culture underneath. Every culture is sediment left by empires, and the thing being defended as traditional is almost always the residue of an earlier conquest that has been there long enough to feel native. I want to be very careful about what this argument does and does not do, because it is misused constantly by people who have read the summary and not the chapter. It does not make empires benign. It does not weigh famine against railways. Harari's claim is narrower and harder to escape: you cannot cleanly separate what an empire did to people from what it made people into, because the same process did both. That is a claim about entanglement. It is not a claim about accounting. He says so. Readers routinely ignore that he says so, and in Section Nine I argue that the way he builds the chapter invites exactly that misreading.
The imperial cycle
And here is the best idea in the chapter. Empires create the people who destroy them, and are destroyed with their own equipment.
HIndia is the textbook case, and it is not a criticism of the independence movement to say so — it is how every anti-imperial movement in history has worked, because there is nothing else to work with. The movement was led substantially by men educated in British-style institutions, several of them trained as lawyers in London. They argued in English. They used British constitutional principles, the vocabulary of rights, and the idea of national self-determination that Europe had been broadcasting across the world. And they organised across a subcontinent that was a single unit because the British had made it one. The empire built the tools, taught the tools, and was taken apart with the tools.
9The empire that belongs to nobody
Harari's most testable prediction, and the one that has aged least well.
e ends by arguing that a new empire is forming, and that it belongs to no nation. A global scientific community with shared standards and shared journals. A global economic system. International law and institutions. A worldwide network of officials, executives, academics and experts who often have more in common with their counterparts abroad than with people two streets away at home. And a set of problems — climate, pandemics, financial contagion, nuclear risk, artificial intelligence — that no single state can solve, which pushes states towards common authority whether or not anybody wants it. It has no emperor and no capital. But by his own two-part definition — rule over many peoples, with flexible and expanding borders — something imperial is being assembled. And it may turn out to be the last one, because it is the first with nowhere left to expand into.
He wrote this in 2011, near the high-water mark of the idea that globalisation was a one-way process.
What has happened since: a substantial resurgence of nationalism in many countries, new trade barriers, contested and violently redrawn borders, open great-power rivalry, and international institutions that struggled visibly during a pandemic that was exactly the kind of problem they were supposed to be for.
That does not prove him wrong over the timescale he is working on. Empires take centuries and he is describing a direction, not a schedule.
But it is worth naming plainly: this is the most falsifiable claim in the book, and the last decade and a half has not gone his way.
10Where I think he is wrong in these two chapters
One near-perfect chapter and one with a genuine logical hole in it.
1. The barter story is contested and he does not say so
Covered in the hidden assumption box above, so briefly here: the "barter came first, then money" sequence is standard economics and a great many anthropologists regard it as a myth, with credit and remembered obligation as the older arrangement. Harari presents the tidy version without mentioning that there is another one. Given how central money is to his whole argument, that is a real omission.
2. "Money corrodes community" is asserted, not shown
It is a plausible and very old worry — Aristotle had it, Marx had it. But in this chapter it arrives as a claim rather than as evidence, and it sits awkwardly beside the celebration of money as the great tolerant connector three pages earlier. Both can be true. He never says when each applies, and the switch does a lot of rhetorical work.
3. And here is my sharpest disagreement in the whole book
The empire chapter contains a step that does not follow, and it is the load-bearing one. "There is no pure culture to return to." True. Demonstrated well. I accept it completely. "Therefore judging empires is difficult." This does not follow, and he never argues for it — he arranges the chapter so that it feels like it follows. You can accept entirely that every culture is a hybrid of conquests, and still say clearly: famine policy that let millions die was wrong. The destruction of an industry to protect another country's mills was wrong. Racial hierarchy written into law was wrong. The people who suffered those things were wronged, and the wrong does not become complicated because their great-grandchildren now play cricket. Impurity of origin is not a reason to suspend judgement. Where you came from and whether you were wronged are separate questions, and the chapter quietly lets the first dissolve the second. Harari does state that he is not excusing conquest. I believe he means it. But if you build a chapter whose structure invites a misreading, and the misreading is the one that gets quoted for fifteen years, at some point the structure is the problem.
4. Where he is completely right
The definition of empire is precise and genuinely useful, and I have found myself applying it to things that are not countries. The observation that empire, not the nation-state, is the historical default is the kind of fact that rearranges your sense of what is normal. And the imperial cycle is one of the most reliable patterns in political history. It applies to companies, institutions, universities and movements as much as to empires. Systems teach their own critique, because they have to teach something and anything teachable can be turned around.
Barter cannot scale, because both people must want what the other has at the same moment. Money replaces every-good-against-every-good with everything-against-one-thing.
Money is not coins. It is anything people agree represents value — barley, silver by weight, cigarettes, a number in a database. Its entire worth is the agreement.
Money is trust, which is why forging it was punished as treason: you are not stealing a coin, you are damaging the story everybody's savings are made of.
It is the most tolerant thing we have built and the tolerance is indifference. It works between enemies because it does not care who you are, and it does not care who you are because it does not care at all.
Empire is the historical default, not the exception. Empires absorb people, which is why almost every culture is imperial sediment — and they are eventually dismantled by the people they educated, using the ideas they exported.
11Words from this part
Everything I stopped to explain, gathered up.
| Word | What it means |
|---|---|
| Doublecoincidenceofwants | For a swap to work, each side must want what the other has, at the same time. The reason barter cannot scale. |
| Sila | The Sumerian barley measure, roughly a litre, used as money around 3000 BC. |
| Shekel | 8.33 grams of silver. A weight, not a coin. |
| Denarius | The Roman silver coin, trusted so widely that rulers outside the empire minted copies of it. |
| Credit | An obligation remembered rather than settled. Many anthropologists think this, not barter, is what came before money. |
| Convertibility | Money's ability to turn any value into any other value. Its greatest strength and its greatest problem, and they are the same property. |
| Empire | Rule over many distinct peoples, with flexible and expandable borders. Nothing to do with size or having an emperor. |
| Imperialcycle | Empires teach their subjects the imperial culture, and are then dismantled by those subjects using it. |
| Nation-State | One people, one territory, one government, fixed borders. Feels eternal. Is about two centuries old in most of the world. |
12The timeline so far
Carried forward, with these two chapters added.
| When | What |
|---|---|
| 3000 BC | Barley money in Sumer. Later, the silver shekel — a weight, not a coin. |
| 640BC | The first coins, minted in Lydia by King Alyattes. |
| 6th century BC | Cyrus the Great: the first empire to claim it ruled for the good of all peoples. |
| 1st century AD | The Roman denarius, copied by rulers outside Roman authority. |
| 1757–1858 | A private company governs much of India. Part 10 returns to this. |
| 1947 | Partition. Up to a million dead, around fifteen million displaced. |
| 2011 | Harari argues a global empire is forming that belongs to no nation. |