1Part Five, in four paragraphs
So this part stands on its own, even if you have not read the last one.
Farming produced a surplus, and a surplus feeds people who do not farm — kings, priests, soldiers, scribes. Those are the people history is about. Every monument is stored peasant labour, moved upward. Above about a hundred and fifty people, personal trust runs out, so something else has to hold a society together. Harari calls it an imagined order: a shared story about how things ought to be arranged, believed by enough people that everybody behaves as though it were simply true. He proved it with the Code of Hammurabi and the American Declaration of Independence — two documents, both claiming eternal justice, flatly contradicting each other. An imagined order survives three ways. It never admits it is imagined. It is installed in you before you are old enough to examine it, so it shapes what you want and not merely what you do. And it lives in millions of heads at once, which is why you cannot leave it on your own. Then Chapter 7: a kingdom generates more numbers than a brain can hold, so somebody invented writing — for barley receipts, not poetry. The first name in recorded history may belong to a clerk called Kushim. And because the archive was built by administration, the past reaches us pre-sorted into the categories of whoever was doing the taxing.
Two chapters, and they zoom in and then all the way out.
Chapter 8 asks what imagined orders do to the people inside them. Every large society sorts people into higher and lower. Harari's claim is that almost none of those rankings have any basis in biology — they start as accidents and end up as nature. This is the chapter where he discusses caste, and where I have to be careful about my own position, which I do in Section Three.
Chapter 9 stands right back and asks whether any of it is going anywhere. His answer is yes, towards unity — and his explanation of why cultures move at all is one of the best ideas in the book.
Fair warning: Chapter 8 is the most politically loaded chapter so far. I am going to give each side its strongest case and I am not going to hand you a verdict where the evidence does not produce one.
2Every large society builds a ladder
And Harari's claim is that almost none of the rungs are real.
Every imagined order sorts people. Free and slave. Noble and commoner. Upper caste and lower. White and black. Men and women. Citizen and foreigner. Insider and outsider. Look for a large society that did not do this and you will struggle. Harari's claim is blunt: almost none of these hierarchies have any basis in biology. They begin in historical accidents, harden into rules and law, and then get explained by a story that makes the accident look like the natural order of things.
He is not saying the differences between people are imaginary. People genuinely differ in height, strength, health and aptitude, and always have.
He is saying the rankings are imaginary — the systems that take one visible feature, sort the entire population by it, attach different rights and different value to each rank, and then insist the arrangement was there before anybody made it.
The difference matters. Nobody disputes that human beings vary. What is being disputed is whether variation licenses a ladder.
And the machine that keeps a ladder standing is the most important idea in this chapter.
3The vicious circle
How a myth manufactures the very evidence that appears to prove it — and why decent people end up believing a hierarchy is deserved.
Here is the loop.
Read step five and step six together, because that is where the trap is. Deny a group schooling for two hundred years, and they will be unschooled. Then point at the fact that they are unschooled. The evidence is real. The system manufactured it. That is why so many decent, sincere people end up believing a hierarchy is deserved. They are not lying and they are not stupid. They are looking at genuine evidence, produced by the arrangement itself, and drawing the obvious conclusion — and after a few generations nobody remembers the accident at the beginning. All anybody can see is a group that is visibly worse off and a story explaining why.
The example Harari builds it on
There is nothing about African ancestry that suits a person to slavery. The reasons Africans were enslaved in the Americas are a stack of accidents that have nothing to do with the people involved.
ISugar and cotton plantations in the tropical New World needed enormous quantities of brutal labour. The native population had collapsed from European diseases. Europeans would not do the work, and transporting European convicts was expensive and politically awkward at home. West Africa was relatively close to the Atlantic routes. Slave-trading networks already existed on the African coast and could be bought into. And people from a tropical disease environment survived malaria and yellow fever far better than Europeans did, which made them, in the arithmetic of a plantation, a better investment. Not one of those reasons is these people are inferior. Every one of them is geography, disease and money. But once millions of people were enslaved, a justification became necessary — because Christian Europeans and later Americans held principles that slavery flatly contradicted, and human beings do not tolerate that gap for long. So the myths were built. First religious ones, including a twisted reading of the biblical curse of Ham. Then, in the nineteenth century, pseudo-scientific ones: skull measurements, racial classifications, invented tables of intelligence, all published in the vocabulary of science and looking exactly like the real science of the period. And then the circle turned. Slavery ended in the United States in 1865. But black Americans were then largely barred from good schools, decent housing, skilled work, the vote and equal treatment in court — which produced poverty and low education, which was pointed at as proof of inferiority, which justified continued exclusion. The loop kept running for a hundred years after the cause was legally abolished. That is the point of the diagram. Removing step one does not stop the machine, because by then steps four, five and six are feeding each other.
4The same machine, different parts
Harari's second example is caste. Before I explain it, I have to tell you where I am standing.
am Indian. I am writing about the Indian caste system, and I am writing about it inside a series where I have already said I would flag any place my own position could pull the writing.
This is one of those places, and it pulls in two directions at once.
There is a habit among some Indians of treating caste as a foreign slander — a thing invented by colonial administrators and exaggerated by outsiders who do not understand us. There is an opposite habit, common in Western writing, of treating caste as the single explanation for everything about India, applied with an enthusiasm nobody brings to hierarchy in their own country.
Both of those are ways of not looking at the thing.
So here is what I will do. I will explain Harari's argument. I will give the scholarly complications he leaves out, including the ones that make the story less tidy in his favour. And where the evidence does not settle something, I will say so rather than choosing the version that suits me.
The historical origin
The account most historians give runs roughly like this. Somewhere in the second millennium BC, Indo- Aryan-speaking peoples moved into the subcontinent, and over a long period a social order took shape in which they occupied the higher positions. Harari puts this at around three thousand years ago. The dating is argued over — most mainstream estimates for the migrations sit around 1500 BC, and the formation of anything resembling a settled ranked order is later and gradual rather than sudden. There are also scholars, particularly in India, who dispute the migration account itself. I am not competent to settle that and I am not going to pretend otherwise. What matters for Harari's argument is only this: the origin is a historical accident of who arrived, when, and with what advantages. It is the same category of accident as the Atlantic slave trade, in a completely different shape.
Then the justification
What followed is the part that fits his machine exactly. The Rig Veda contains the myth of Purusha — a primeval being from whose body society emerges. The priests come from the mouth. The warriors from the arms. The merchants from the thighs. The servants from the feet. Look at what that story does. It converts a political arrangement into cosmic anatomy. You are not below someone because of something that happened to your ancestors. You are below them because of which part of a divine body you came out of — which is not a historical claim at all, and therefore cannot be argued with historically. Layered on top came doctrines of ritual purity and pollution, and a system in which occupation, marriage, food and physical contact were all regulated by birth.
Then the circle
Groups placed at the bottom were, over centuries, denied education, land ownership, entry to temples, and any honourable trade. The resulting poverty and illiteracy were then read back as evidence that they belonged where they were. Same loop as Figure 1. Different rungs, different myth, identical machinery.
Three complications that make the picture less tidy and more interesting.
Varna is not jati. The four-fold scheme in the ancient texts — priest, warrior, merchant, servant — is not the same thing as the thousands of actual birth-groups that regulate marriage and occupation on the ground. The relationship between the textual scheme and the lived reality is one of the oldest arguments in Indian sociology, and Harari treats them as one system.
The colonial period reshaped it. British census-taking from the late nineteenth century required every person to be assigned to exactly one caste, in a fixed list, with a fixed rank. That was an administrative convenience applied to something that had been considerably more local, more fluid and more disputed. A number of historians argue this rigidified the system substantially. Note what that means: the archive problem from Part Five is doing damage here. The record we have of caste was largely created by an authority that needed it to be countable.
It was fought from inside, continuously. Devotional movements across many centuries rejected birth- ranking outright. Sikhism — which I have a personal stake in, so take my emphasis with that in mind — was founded partly on that rejection, and the practice of langar, where everybody sits in a line on the floor and eats the same food regardless of birth, is a deliberate mechanical attack on purity ranking rather than merely a statement against it.
None of this makes the hierarchy less real or its consequences less severe. It makes the history less like a single ancient block and more like a long argument that particular parties kept winning.
And the legal ending is worth stating plainly, because both hierarchies got one. India's constitution abolished untouchability in 1950. American civil rights legislation came in the 1960s.
And in both cases the circle kept turning afterwards, because disadvantage accumulated over centuries does not vanish when a statute does. That is not a claim about either country's virtue. It is what Figure 1 predicts.
5Why purity feels like nausea
A short section on the biological instinct that hierarchies borrow, and why arguing against them so rarely works.
Humans have a real, evolved instinct of disgust. Rotting food, faeces, open wounds, visibly sick bodies. It is automatic, it is fast, it happens before any thinking, and it is genuinely protective — it is why bad meat smells the way it does. That instinct can be pointed at people. Once a society teaches children that contact with a particular group is contaminating — that their touch, their shadow, their water, their food is polluting — the ancient disease-avoidance machinery fires at human beings.
This explains something I had never understood, and it is bleak.
When you argue with somebody about a purity-based hierarchy, you assume you are arguing with a belief. You bring evidence. The evidence does not work, and you conclude they are stupid or dishonest.
You are not arguing with a belief. You are arguing with nausea. The feeling is genuine; only the target was assigned. And a feeling that arrives before thought cannot be removed by a better argument, any more than you can be reasoned out of flinching.
What does seem to shift it is contact — ordinary, repeated, unremarkable contact, of the kind that gives the machinery a chance to stop firing. Which is exactly what these systems are built to prevent, and that is not a coincidence.
6Biology enables, culture forbids
The sharpest single sentence in the book, and the one you can use most often.
Then Harari turns to the hierarchy that shows up almost everywhere: gender. First a distinction that clears up an enormous amount of confusion.
Sex is biological. Chromosomes, hormones, anatomy. Male and female are categories of biology.
Gender is cultural. Man and woman are categories of society — bundles of expectations about clothing, work, emotion, authority and behaviour.
Why it matters: sex barely changes. Gender changes constantly. If gender were simply a readout of biology, it would look the same everywhere and in every century, and it does not come close.
What counted as manly in ancient Athens, in seventeenth-century France, in medieval Japan and in modern Mumbai are wildly different things. European noblemen wore heels, wigs, powder, lace and jewellery and were thought more masculine for it, not less. What a man is permitted to cry about, own, wear, or say has been rewritten many times, in living memory, more than once.
And then the line I would carry out of this book if I could only take one:
Biology enables, culture forbids.
Biology permits an enormous range of human behaviour. Culture takes a narrow slice of that range and calls everything outside it unnatural.
And here is the tell, which I think is the smartest single move Harari makes:
There is no law forbidding people to photosynthesise. There is no rule against breathing underwater, no commandment about not walking through walls, and no religion that prohibits living for a thousand years.
Why not? Because you do not need to forbid what cannot happen.
So every prohibition that exists — about who may work, marry, inherit, own, speak, lead, or love — exists precisely because the thing it forbids is entirely possible. A rule is a door that could otherwise be walked through.
Which means: any time somebody tells you a human behaviour is unnatural, notice that the ban is proof nature permitted it. If nature had forbidden it, the culture would not have had to.
This is a tool. It works on any rule, in any society, including the ones you agree with.
7The question he cannot answer
Patriarchy is nearly universal. Harari takes the three standard explanations apart and then admits he does not know. It is the most impressive thing in the chapter.
Almost every known complex society has put men in the dominant political and economic position. That is a striking pattern. Unlike caste or race, it is not obviously the result of one accident in one place, because it happened almost everywhere, independently. So there ought to be a good explanation. Harari examines the three that are usually given, and rejects all three.
"Men are physically stronger." The link between muscle and social power is weak to non-existent. The jobs that demand the most physical strength — labouring, mining, portering, construction — are usually the lowest-paid and lowest-status, while power sits with people who lift nothing heavier than a document. Sixty-year-old priests, lawyers and politicians rule societies full of strong young men. And in several respects women endure hardship better: they generally live longer and cope better with famine and disease. If strength produced power, the world would look nothing like it does.
"Men are more aggressive, and war makes rulers." Wars are not won by the most aggressive individual in the field. They are won by organisation, logistics, supply, alliance-building and diplomacy — cooperative skills, not brawling ones. And history is full of arrangements where lower-class men obeyed upper-class women without anybody finding it strange. Aggression does not explain who ends up giving the orders.
"Evolution selected pushy men and agreeable women." The theory runs that men evolved to compete for mates while women, dependent during pregnancy, evolved to be accommodating in order to secure support. It does not survive contact with the animal kingdom. Among bonobos — as closely related to us as chimpanzees are — females form cooperative alliances and dominate males. Elephant herds are led by matriarchs while males live at the edges. Female cooperation defeating male competition is common in nature, so there is no reason it could not have happened in us. And if women were selected for dependence, why would they consistently choose the most aggressive males rather than the most reliable ones?
Where it stands. Harari's answer is: we do not know. He lays out the three, dismantles them, and stops.
What would settle it. Honestly, nothing available. The pattern was set in societies that left no records, and every later society we can study is already inside it. I want to give this its due. Harari is a writer who will build a confident chapter on thin evidence, and I have said so in every part of this series. Here, at the point where the strongest counter-example to his own thesis sits — a near-universal hierarchy that does not look like a local accident — he does not paper over it. He says he cannot explain it.
That is worth more to me than any of his confident passages.
All three explanations above — and Harari's rejection of all three — share one unexamined premise. They assume power must be tracking some capacity.
Strength, aggression, reproductive strategy: each answer names a thing men supposedly have more of, and derives the hierarchy from it. Harari's method is to test each capacity and show the link fails. But he is still playing the same game, on the same board.
Ask instead: what if power does not track any capacity at all? What if it tracks whoever was holding it last, plus whatever story is currently available to justify that?
That is exactly what Figure 1 describes. The vicious circle needs no founding capacity. It needs a starting position and time. Harari uses that mechanism for race and for caste, and then, faced with gender, reaches back for capacity-based explanations and reports that none of them work.
He may have had the answer in his own chapter and looked past it — because for gender, unlike race or caste, there is one physical fact that genuinely differs and genuinely constrains, and it is not strength. It is that in every society before contraception and modern medicine, pregnancy and prolonged childrearing were dangerous, frequent, and immobilising. That is a real starting position. It does not require anybody to believe women were inferior. And a starting position is all the circle needs.
I am not confident this is right. I am confident it deserved a paragraph, and it did not get one.
And then the hopeful part
NHarari ends the chapter with something concrete. In roughly the last hundred years, in many societies, gender roles have changed faster and more dramatically than in the preceding several thousand. Women vote, hold office, run companies, argue cases in court, perform surgery, fly aircraft, command soldiers. Human DNA has not changed since 1920. So whatever was holding the arrangement in place, it was not biology — because biology is the one thing that did not move. It was an imagined order. And imagined orders can be rewritten. That was the promise back in Part Two, and this is the clearest evidence in the whole book that it is not just a theory.
8Every culture is broken, and that is what makes it move
One of the best ideas in the book, and it arrives so quietly that it is easy to miss.
ow we zoom all the way out. Chapter 9 opens with something you already know and have probably never said out loud: no culture makes sense. Not no culture is stupid. Every culture holds beliefs that flatly contradict its other beliefs, and carries on regardless. Take medieval European nobles. They were Christians, and Christianity told them to turn the other cheek, that the meek would inherit the earth, that killing was sin. They were also knights, and chivalry told them honour must be defended with a sword, that cowardice was worse than death, that an insult demanded blood. A knight went from mass, where he was told to forgive, to a duel, where he was told to kill. And he believed both, sincerely, on the same afternoon. Take now. Modern politics runs on two values that cannot both be maximised: liberty and equality. Push liberty all the way and some people own everything while others starve. Push equality all the way and you must take from people by force, which destroys liberty. Every democracy on Earth is an unstable compromise between two things it claims to believe absolutely.
Cognitive dissonance: holding two beliefs at once that do not fit together, and feeling the discomfort of it.
The usual assumption is that this is a fault — a sign of muddle or hypocrisy, something a well-organised mind would not do.
Harari's claim is the opposite. It is the engine.
A culture with no contradictions would be frozen. Nothing would push it anywhere. The tension is what makes people argue, reform, rebel, reinterpret and reinvent. Every culture is permanently trying to resolve a contradiction it cannot resolve, and that trying is history.
The knight arguing with the priest is the medieval world in motion.
I think this is genuinely excellent, and it changed how I read arguments — including political ones happening now. The argument inside a society is not a sign that the society is failing. It is the society working.
9Where all of it has been going
The direction of history, and why it cannot be predicted even in principle.
So culture is always moving. Does it move anywhere in particular? Harari says yes, and it is visible if you stand far enough back. Around 10,000 BC the planet held thousands of separate human worlds. Not thousands of countries — thousands of self-contained universes. A band in the Amazon and a band in Siberia had no idea the other existed. There was no such thing as "humanity" as a category anybody could have used. By 1450 AD, most of Africa, Europe and Asia was a single interconnected zone. Goods, diseases, religions and ideas moved from Morocco to China. By 2000 AD there is essentially one world. One financial network. One scientific community. One information web. One set of problems that no single country can solve. The direction is towards unity. Slowly, messily, violently, but consistently. And what made it possible were three new kinds of order that appeared in the first millennium BC, all of which were universal in a way nothing before them had been.
| Word | What it means |
|---|---|
| Money | An economic order. Anyone can use it, whoever they are. Part 7. |
| Empire | A political order. Anyone can be conquered and absorbed. Part 7. |
| Universalreligion | Buddhism first, later Christianity and Islam. Anyone can convert. Part 8. |
Older orders were local. Your gods belonged to your valley. Your rules applied to your tribe. Outsiders were simply outside. For the first time, people started thinking in terms of one humankind — meaning everybody, including people they would never meet.
Two warnings that make the chapter more than a story about progress
First: the direction tells you nothing about whether it is good. Unity is where we have been heading. That is a description, not a compliment. Getting there involved conquest, slavery, epidemics and the destruction of thousands of ways of living. Second: history is not made for our benefit. We assume good ideas win because they are good, and that arrangements which make people happier out-compete arrangements that do not. Harari says there is no evidence for this whatsoever. History selects for whatever spreads. Whether it is good for the people inside it is a separate question, and nothing in the process is checking.
And why nobody can predict it
Harari uses a distinction here that is worth learning properly, because it applies well beyond history.
Level one chaos does not react to predictions about it. Weather is level one. It is fiendishly complicated, but a forecast does not change what the clouds do. So with better data and bigger computers, forecasts genuinely improve.
Level two chaos reacts to predictions about it. Markets are level two. If a perfect model predicts oil will hit two hundred dollars tomorrow, traders buy today, the price moves now, and the prediction has destroyed itself by being made.
History is level two. Predict a revolution and the government arrests the leaders. Predict a crash and everybody sells early. Predict an election and the campaigns change.
Why it matters: this is not a limitation waiting for better technology. It is structural. History cannot be predicted, ever, in principle — and neither can anything else that reacts to being predicted.
Which raises a fair question. If history cannot be predicted, why study it? Harari's answer is the best defence of history I have read in one sentence, and it is why I keep doing this. We study history not to know the future, but to see that our present situation is neither natural nor inevitable — and that we therefore have far more possibilities in front of us than we imagine. You read the past to stop mistaking the present for reality.
10Where I think he is wrong in these two chapters
One outstanding chapter, one that contradicts the rest of the book and never notices.
1. He contradicts himself about direction, and does not resolve it
This is the biggest problem in the book and Chapter 9 is where it becomes visible. In this chapter he argues history has a clear direction — towards unity — and spends the whole of the next two parts describing the mechanisms delivering it. In the very next chapter he argues history is chaotic, undirected, unpredictable in principle, and that nothing is selecting for anything except contagiousness. Both arguments are good. They cannot both be doing the work he asks of them. If history is undirected chaos, an arrow is a pattern you have drawn on it after the fact, from the endpoint you happen to be standing on. He never reconciles the two and, as far as I can find, never acknowledges the tension.
2. The arrow depends heavily on writing in 2011
From 2011, globalisation looks like the end point of a long convergence, and unity is the obvious shape of history. Look from 1350, mid-plague, with the Mongol trade network collapsing and Eurasia fragmenting, and history is obviously going the other way. Empires have broken apart at least as often as they have merged. And the years since 2011 have produced a serious resurgence of nationalism, trade barriers, contested borders and open great-power rivalry. The arrow may still hold over centuries. It has not held over the last decade, and a claim about direction that reverses within fifteen years of publication is worth handling carefully.
3. He treats caste and race as one system running twice
The parallel is illuminating and it is also flattening. Caste is not simply race with different labels — it involves ritual purity, occupational inheritance, endogamy and a religious cosmology, and the relationship between the textual four-fold scheme and thousands of actual birth-groups is genuinely disputed among people who spend their lives on it. He also does not mention that the system we describe today was substantially reshaped by colonial administration, which is a serious omission — not because it excuses anything, but because it is precisely the archive problem his own Chapter 7 sets up.
4. The near-universality of patriarchy is a real hole
I said in Section Six that his admission is impressive. It is. It is also a genuine problem for his thesis. If hierarchies are historical accidents, why did nearly every independent society land on approximately the same one? Accident is a hard word to defend against that pattern, and the answer I sketched in the hidden assumption box — that pregnancy and childrearing without medicine or contraception supply a starting position the circle can work on — is at least worth engaging with. He does not.
5. And where he is simply right
The vicious circle is the best tool in the book. It explains how a hierarchy manufactures its own evidence, why removing the original cause does not stop it, and why sincere people believe it. "Biology enables, culture forbids" is the sharpest sentence he has written, and the observation that nobody bans the impossible is a genuine piece of reasoning you can use for the rest of your life. And "contradictions are the engine of culture" is worth the chapter on its own.
Hierarchies begin as accidents — who arrived first, which crop needed picking, who survived which disease — and are then explained by a story that makes the accident look like nature.
The vicious circle is the machine. The myth causes the disadvantage, the disadvantage looks like evidence for the myth, and the evidence hardens the myth. The evidence is real. The system made it.
Purity taboos borrow a real instinct. Disgust evolved for rot and disease and can be aimed at people, which is why argument bounces off and only contact seems to shift it.
Biology enables, culture forbids. Nobody bans the impossible, so every prohibition is proof that the thing it forbids can be done.
And every culture is full of contradictions — which is not a fault but the engine. History is societies trying to resolve tensions they cannot resolve. It cannot be predicted, because predicting it changes it.
11Words from this part
Everything I stopped to explain, gathered up.
| Word | What it means |
|---|---|
| Viciouscircle | Myth causes discrimination, discrimination causes the conditions, the conditions appear to prove the myth. |
| Sex | Biological. Chromosomes, hormones, anatomy. |
| Gender | Cultural. The bundle of expectations a society attaches to men and to women. It has been rewritten many times. |
| Patriarchy | A social order in which men hold the dominant political and economic positions. Nearly universal, and unexplained. |
| Varna | The four-fold scheme in the ancient Indian texts — priest, warrior, merchant, servant. |
| Jati | The thousands of actual birth-groups that regulate marriage and occupation. Not the same thing as varna, and the relationship is disputed. |
| Purusha | The primeval being of the Rig Veda, from whose body the varnas are said to emerge. A political arrangement stated as anatomy. |
| Cognitive Dissonance | Holding two incompatible beliefs at once. Harari says this is the engine of cultural change, not a fault. |
| Levelonechaos | Chaos that ignores predictions about it. Weather. |
| Leveltwochaos | Chaos that reacts to predictions about it. Markets, and history. |
12The timeline so far
Carried forward, with these two chapters added.
| When | What |
|---|---|
| 10,000 BC | Thousands of separate human worlds, most unaware of each other. |
| c. 1500 BC | Indo-Aryan migrations into the subcontinent; over a long period, a ranked order takes shape. |
| 1st millennium BC | Money, empire and universal religion appear — the three glues. Part 7 starts here. |
| 1450 AD | Most of Africa, Europe and Asia is one interconnected zone. |
| 1865 | Slavery abolished in the United States. The vicious circle keeps running for a century. |
| 1950 | India's constitution abolishes untouchability. |
| 1920–2020 | Gender roles change faster than in the previous several thousand years, with no change in human DNA. |