Thoughts. What I notice while watching.
In the summer of 1989, a foreign policy journal runs an essay with a question mark in the title: “The End of History?”. The author is the deputy director of policy planning at the US State Department, 36 years old, RAND before that. The Berlin Wall is still standing when the piece goes to print. It falls in November.
That is the first twist in this story. The text that supposedly cheers the fall of the Wall was printed before the Wall fell.
The second twist takes thirty years. In 1992 the essay becomes a book, The End of History and the Last Man. After that the title turns into shorthand for intellectual arrogance. Yugoslavia, Rwanda, September 11, Putin, 2016: every new rupture arrives with the same line attached. So much for the end of history. Samuel Huntington supplies the counter-thesis in 1993 with the clash of civilizations, and the matter is closed in the public filing system. Fukuyama is the man who got it wrong.
I passed that on for years. Then I read the last page of the 1989 essay.
The title has two halves
Almost nobody quotes the second one. “And the Last Man” is not decoration, it is half the thesis, and it comes from Nietzsche: the comfortable, risk-averse, self-satisfied man who no longer wants anything he would have to give something up for.
Fukuyama never claimed that nothing would happen anymore. His claim was narrower and harder: that no order was left in sight that could take legitimacy away from liberal democracy. Wars, massacres and attacks do not refute that by themselves. They refute it only when someone fights them in the name of a better order, one that wins followers across borders.
And then he writes a paragraph at the end of the essay that you read today with your mouth open:
“The end of history will be a very sad time. The struggle for recognition, the willingness to risk one's life for a purely abstract goal, the worldwide ideological struggle that called forth daring, courage, imagination, and idealism, will be replaced by economic calculation, the endless solving of technical problems, environmental concerns, and the satisfaction of sophisticated consumer demands.”
And the final sentence of the piece:
“Perhaps this very prospect of centuries of boredom at the end of history will serve to get history started once again.”
That is not a victory lap. That is a warning with a date on it. The man whose name is used as a synonym for blindness wrote down in 1989 that the condition he saw coming would produce people who cannot stand it. We quote the first half to laugh at it, and we live in the second.
The currency that cannot be divided
The concept everything in Fukuyama turns on is recognition. He takes it from Plato, thymos, the part of the soul that does not want to be fed but to be seen, and reads it through Hegel. And this is where my own argument starts, because that concept explains something the usual polarization diagnoses do not.
Politics has two currencies, and they behave nothing alike.
Interests are divisible. Four percent against six percent ends at 5.1. Both sides can go home and say they got something. A compromise here is not a defeat, it is arithmetic.
Recognition is not divisible. “We are worth the same.” “Our way of life counts too.” “Our experience is not imaginary.” There is no 5.1. Half of recognition is not half a solution, it is an insult. Negotiate in this currency and there are only two possible outcomes.
Now the sentence I am here for: a system that stops serving interests does not go quiet. It changes currency.
When procedures take years, when appointments are handed out and not kept, when every commitment comes with a reservation, people do not learn that governing is complicated. They learn that their own case does not appear down there. And once that is settled, the conflict does not go away. It moves into the only currency still available when there is nothing left to hand out: who actually counts here? Who gets asked? Who gets to be considered decent?
That explains the tone of the present better than any story about evil algorithms. We are not arguing louder than before. We are arguing about something else. You can lose a question of interest and carry on the next day. In a question of recognition you do not lose a position, you lose your claim to be a serious person. Nobody hands that over. Nobody should have to.
Why we talk about each other instead of with each other
Almost everything else follows from that one property.
The audience matters more than the opponent. Recognition is not distributed in a conversation, it is distributed in front of spectators. So you talk about the other side rather than to it. Talking about them is the actual transaction, not a side effect.
Error turns into character. The decisive step is not holding the other side to be wrong. It is holding them to be not decent. That is convenient, because it dissolves the burden of proof: against an argument I have to produce something, against a bad character I do not. The price is on the back of the receipt. An error is repairable, a character is not. Declare your opponent a character case and you have removed your own ability to ever reach a result with them. You cannot negotiate with someone whose agreement would stain you.
Milieus close. Not because people got dumber. Because in a recognition conflict, every contact with the other side is a risk with no return. What is there to win? An argument you cannot use, plus suspicion from your own people. So you stay inside, and inside, opinion becomes the entry ticket. Contradicting your own camp costs more than losing to the other one.
That is not a character flaw of one generation. That is the physics of a currency that cannot be divided.
The state nobody maintains
The second mistake is not Fukuyama's, it belongs to how he was received. It sits in the word itself: end.
A final state gets no maintenance. What is finished gets celebrated, not serviced. And that is what happened to the 1989 sentence right across the West. Liberal democracy went from being an operation to being a possession.
But it is an operation. Courts that are fast enough. Administrations that answer. Parties where more happens than staffing questions. Schools where arguing is practiced. A press that researches instead of sharpening. All of that is upkeep, and upkeep is invisible as long as it works.
The mistake is cheap to describe and expensive to pay for: whoever considers the result stable stops investing in the ability to deliver. And the ability to deliver is, see above, exactly the lever that decides which currency the fight is fought in. Both halves of my argument mesh here. The declared final state saves the upkeep, the missing upkeep produces the non-delivery, and non-delivery switches the conflict from interests to recognition. If you want to see what the last step looks like when it arrives as an election result, it is written up here.
The return to a state that never existed
Now the uncomfortable symmetry.
The extremes promise to restore a moderate normal. You hear it from the right as a return to a country that never was, with streets without fear, pensions without worry and a population without dispute. You hear it from the left as a return to a republic where decency was self-evident and nobody had to say what they were against.
And you hear it, and this is the part almost nobody concedes, from the center: back to normal. Back to a factual tone. Back to where we were before all this.
Ask all three for the year. Back to exactly when? 1990, with the baseball bat years? 1965, with the conditions that blew up in 1968? There is no year that survives the question, because the target was never a historical state. It is a feeling: the memory of belonging without having to argue for it. Which is recognition again, this time projected backwards.
The popular reading of the end of history supplied the template. If history has a natural resting state, then every deviation from it is a disturbance. And disturbances have culprits. From there it is a very short walk to the conviction that the conditions are not what is in dispute, but that somebody is deliberately messing them up.
Delay is not refutation
In July 2026 the German musician Danger Dan releases a song called “Keine Angst”. Public broadcaster ZDF drops him from the 100th episode of the satire show “Die Anstalt” over the lyrics, the director general calls the text a call to violent resistance, and a debate about artistic freedom follows. What interests me about the song is neither its program nor the row about it. What interests me is a tense. One line goes:
“It will still come for us, just a little later.”
That is the most precise sentence I have come across on this subject in years, and it is not about the far right, it is about our relationship with time. We treat “not yet” as “never”. A danger that fails to show up on the announced date counts as handled. A promise that does not break immediately counts as kept.
That is exactly how Fukuyama was “refuted”. Not by an argument, but by a calendar. The end of history did not arrive, therefore the thesis was wrong. That is not a test, that is impatience.
The knife cuts both ways, which is why I have to turn it on myself. A thesis that books every delay as a confirmation is no longer a thesis, it is a prophecy. So here is what would make mine false.
The lens turned on myself
First, the obvious objection. I am doing what Fukuyama did: taking a curve and claiming finality. If my mechanism holds, it has to be reversible. It says: non-delivery drives the conflict into the recognition currency. Then noticeable delivery has to pull it back out. Not mood, not communication, but things you can measure: processing times, permit durations, waiting times for appointments, punctuality. If an administration visibly improves on those numbers and the recognition conflicts keep running unchanged anyway, my mechanism is wrong. Then delivery is a sideshow and I have built a cause out of a correlation. I have run the numbers on how easily that happens.
Second, the objection from two sides. From the left: the recognition story depoliticizes, this is really about distribution and ownership. From the right: this is about culture and migration, not processing times. Both are partly right, and neither hits my point. I am not claiming that the ability to deliver is the only cause. I am claiming it is the only part that lies entirely in the hands of the people who currently look so concerned.
Third, and this is the interesting part: where was Fukuyama actually wrong? Not about events. In the 1989 essay he explicitly checks what could still threaten the liberal order and names two candidates: religion and nationalism. He dismisses both, nationalism with the argument that it is not one single phenomenon and has no program for organizing a society beyond the wish for independence.
That was defensible in 1989. Today it is wrong, and wrong in a precise way. The new right has a program: rebuild the judiciary, capture public broadcasting, hollow out the administration from inside. It has a method, it has an international, and its parts copy each other across borders, from Budapest via Warsaw to Washington and back. Fukuyama was looking for an alternative that writes a book. The alternative arrived as a playbook.
That is the real error, and it is far sharper than “he did not see September 11 coming”. He missed that an order does not have to be replaced by a better idea. It can simply be hollowed out by people who keep using its forms. Fukuyama filed the correction himself in 2018, in a book about dignity and resentment. It was reviewed a great deal more quietly than the sentence everyone laughs at.
What I take from it
This holds for a country, and it holds unchanged for a company, a team, a family.
- Check the currency before you negotiate. When it stops being about numbers and starts being about respect, the negotiation is lost before it begins. The first job is then not persuasion but translation back: which divisible quantity is hiding behind the respect?
- Delivery counts twice. Once as a result and once as recognition. A deadline that holds tells someone they exist. No communication concept on earth substitutes for that.
- Never say “we have that now”. In every organization that sentence is the beginning of decay, whether it refers to culture, security, quality or democracy. None of those is a state. All of them are an operation.
- Keep the conflict repairable. You are allowed to think someone is wrong. The second you think they are not decent, you have written off every future result with them.
Fukuyama's last sentence from 1989 has been read as a triumph. It was a note left on the table: the prospect of centuries of boredom at the end of history might be what gets history started again.
It started again. On time, in the described order, for the described reason. Nobody took it for a forecast, because everybody had already started laughing at the title.
Sources and status
Status: September 21, 2026. This is an opinion piece. The record:
- Francis Fukuyama, “The End of History?”, The National Interest No. 16, Summer 1989. Both quotes are from the closing section. The essay appeared before the Wall fell in November 1989.
- Francis Fukuyama, The End of History and the Last Man, Free Press 1992.
- Francis Fukuyama, Identity: The Demand for Dignity and the Politics of Resentment, 2018.
- Samuel P. Huntington, “The Clash of Civilizations?”, Foreign Affairs 1993, as a book in 1996.
- Danger Dan, “Keine Angst”, released July 2026. The removal from the 100th episode of the ZDF show “Die Anstalt” and the director general's reasoning were widely reported in July 2026. The quoted line is my translation of the German lyric.
- The assessment of religion and nationalism is in the section of the 1989 essay where Fukuyama looks for remaining contradictions in the liberal order.