In the summer of 1918, Rosa Luxemburg was in the women's prison at Breslau, Germany, serving a sentence for opposing the war. The Russian Revolution had happened the previous year — the soviets taking power in October, the Bolsheviks consolidating control, the first workers' state in history declared.
Her letters from Breslau are some of the most extraordinary documents of the period. She wrote about the white anemones pushing through the prison courtyard in April. The colors of starlings in the afternoon light. The birds she had named, that she fed through the bars of her window. She was paying ferocious attention to the world from a room she could not leave.
And between the letters, in a manuscript she knew would not be published while the people it criticized held power, she was writing The Russian Revolution.
It praised the revolution. It praised the Bolsheviks' courage, their decisive action at a moment when other socialist parties had failed. It said the October Revolution was one of the most significant events in history. And in the same pages, written with the same clarity, it described exactly how the revolution would fail.
Not might fail. Would fail.
She predicted the suppression of the soviets — the workers' and soldiers' councils that were the revolution's actual democratic content — by the party that claimed to speak for them. She predicted that the Bolsheviks' contempt for formal democratic rights, their willingness to suppress dissent in the name of revolutionary necessity, would produce a government indistinguishable in its operations from the one it had replaced. She predicted that a revolution won by a disciplined vanguard, without the genuine self-governing capacity of the working class, would end in the domination of the masses by the party elite.
She wrote this in 1918. It was published posthumously in 1922. Stalin demonstrated its accuracy for thirty-five years.
The essay contains a sentence that has outlasted every Soviet apologist who denied it: Freedom only for supporters of the government, only for members of one party — however numerous they may be — is no freedom at all. Freedom is always and exclusively freedom for the one who thinks differently.
Read that again in light of Wynter. The overrepresented Man — the universal human that every classification system builds itself around — cannot tolerate the one who thinks differently, because the one who thinks differently is proof that the universality is a fiction. Luxemburg is not making an abstract liberal argument about dissent. She is naming singularity as the political condition of freedom. The system that cannot hold the one who thinks differently is the same system that cannot hold the person whose humanity falls outside the recognized category. Same logic. Same exclusion. Different register.
She had been arguing with Lenin since at least 1904. He was writing from exile in London; she was in Berlin; they were debating across the European left the question that would define the next century: what does a revolution actually require?
Lenin's answer, laid out in What Is To Be Done?: the working class, left to its own devices, develops only trade union consciousness — the desire for better wages within capitalism, not the desire to abolish it. Revolutionary consciousness has to be brought to workers from outside, by a disciplined party of professional revolutionaries who have developed the correct theory. The party knows what the class needs better than the class knows itself.
Luxemburg's response, published the same year: this is backwards, and the backwardness is catastrophic. If you build a revolutionary organization on the premise that leadership knows better than members, you don't get an instrument of liberation. You get a bureaucracy that suppresses precisely the spontaneous, distributed, locally-rooted political creativity of workers — the creativity that produces genuine revolutionary capacity — in favor of party discipline. The vanguard doesn't lead the working class. It substitutes for it. And substitution is not liberation. Substitution is domination with good intentions.
She was not arguing against organization. She was arguing against a specific architecture: one that routes all knowledge and authority to the top, that treats the distributed intelligence of millions of people as raw material to be processed by a party headquarters, that cannot learn from the people it claims to represent because it has defined itself as already knowing more than they do.
Her alternative was harder to describe and easier to dismiss. She was pointing at the soviets themselves — the factory councils, the neighborhood assemblies, the organic self-governing structures that workers had built without being told to — and saying: that's it. That's what socialism looks like. Don't capture it. Don't direct it. Trust it.
Lenin won the argument in the short run because his model produced a revolution. Luxemburg won the argument in the long run because his model then ate it.
The soviets were the revolution. Not the Bolshevik Party, not Lenin, not the red flag over the Winter Palace — the councils of workers, soldiers, and peasants that had formed organically across Russia, governing factories and neighborhoods and military units, coordinating across cities, representing the actual self-organizing capacity of the Russian working class.
The Bolshevik slogan had been All Power to the Soviets. It was right. What followed dismantled it, step by step: through a series of decisions, each with a pragmatic justification — the party needed to coordinate across soviets, needed to ensure consistency during the civil war, needed to manage military threats — more and more authority was routed through the party hierarchy and away from the councils. By the early 1920s, the soviets were rubber stamps — formally present, functionally irrelevant.
This is the capture mechanism from Chapter 2, operating at revolutionary scale. The soviets provided legitimacy. The party absorbed it. The armor was made from the workers' participation.
Alexandra Kollontai named what was happening from inside the government: the party had substituted itself for the working class. She argued it in print — The Workers' Opposition — and the party answered by banning the argument and dissolving her faction in the same stroke. Then it sent her abroad — kept useful, kept visible, and kept permanently away from the rooms where what she had named was completing itself.
She made Gramsci's diagnosis in real time, eight years before the prison notebooks — not from a cell but from inside the apparatus, while it could still have been heard. She outlived him by fifteen years, watching from embassy windows while the prediction ran its course. The silencing did not need a prison. A sufficiently distant embassy would do.
In March 1921, the sailors and workers at Kronstadt — the most militantly revolutionary force of the October insurrection — rose up demanding what the revolution had promised: free soviets, freedom of speech, an end to forced grain requisition. They were holding the revolution accountable to its own stated principles. The Bolshevik government crushed them militarily. Trotsky commanded the assault. Thousands were killed or sent to camps.
This was not Stalin. This was Lenin at the height of his legitimacy, killing the people who had made the revolution in order to protect the party's monopoly on power. The soviets had been the revolution. The party ate the soviets. Stalin came after, not instead.
In July 1936, Francisco Franco launched a fascist military coup against the Spanish Republic. In Catalonia and Aragon, workers and peasants didn't wait for a party directive.
In Barcelona, workers took over the trams and kept them running — under workers' control, with elected management committees, fares reduced to a flat rate anyone could afford. The precision engineering workshops that had made Barcelona one of Europe's industrial centers were collectivized and converted for war production. In the Aragon countryside, hundreds of villages collectivized their land — not under compulsion but by assembly, by the communities that had worked it deciding collectively how it would be governed. The CNT, the anarcho-syndicalist labor union with over a million members, provided the organizational framework.
At its peak, the anarchist collectives involved hundreds of thousands of people, managed major industries, demonstrated that complex economies could be coordinated without either capitalist markets or Stalinist central planning. For roughly a year, the most industrially advanced region of Spain operated as a functioning commons.
It was not primarily Franco who destroyed it. The Spanish Communist Party — a Comintern affiliate, directed by Moscow, literally the same international organization that had dismantled the Russian soviets fifteen years earlier — systematically dismantled the revolutionary collectives in the name of building a unified front against fascism. The pragmatic argument: you can't fight a modern war with anarchist militias. The collectives were dissolved. Anarchist leaders were arrested and in some cases killed — by their ostensible allies.
George Orwell was there. He had gone to fight fascism and found himself watching the revolution being destroyed by the people claiming to defend it. Homage to Catalonia is his account. It changed him permanently — into the writer who would spend the rest of his life trying to name the specific mechanism by which revolutionary movements destroy themselves from within, and the language that makes the destruction invisible as destruction.
The Catalan collectives were the Turin councils. Not a rhyme — the same form, the same workers governing the same kind of production, destroyed by the same organization operating under a different national name. Fifteen years and a thousand miles separated them. The logic was identical. The people who built both knew how to govern themselves. The party knew this, which is exactly why it had to stop them. The distributed commons had demonstrated — in Turin, in Barcelona, in the Aragon countryside — that the vanguard was unnecessary. That demonstration was the threat.
Centralized revolutionary organizations are capture mechanisms.
This is not a moral claim about bad leaders. It is an architectural claim: the organizational form of the movement determines the organizational form of what the movement builds. A movement organized through a vanguard party builds a state that looks like a vanguard party. A movement organized through distributed self-governance builds institutions of distributed self-governance.
The means are the ends. Every time.
I tried to build a distributed institution and learned why this is hard in a specific way that the theory doesn't quite prepare you for.
The restaurant was organized, in principle, around distributed self-governance. Everyone in the room, decisions made by the people doing the work. In practice, only we had skin in the game — financial risk, personal liability, the consequences of failure landing on specific bodies. Everyone else had a labor contract. About halfway through, I understood why early syndicalists viewed labor contracts with such suspicion: a contract is an enclosure. It specifies what is owed and what is not owed, and everything outside the specification — the initiative, the investment, the genuine ownership of the project — falls outside it. The contract is a boundary that produces exactly the relationship it describes. You want distributed intelligence; you create the conditions for transactional compliance. You want people who treat the institution as theirs; you hire people who have every reason to treat it as yours.
What worked operationally was real. What we couldn't cultivate was the culture — the shared ownership of the project that would have made the architecture function. I didn't have the bandwidth to do unpaid mentorship and counseling on top of leadership and management, and there isn't enough slack in the restaurant industry to make that possible. If we had started with a larger founding cadre, people with skin in the game from the beginning, it might have been different. What undermined us wasn't the idea. It was sexism and American hyperindividualism — the ambient assumption that the individual unit is the family, that labor is transactional, that collective investment in a shared project is naïve. The hegemony Gramsci named, operating at the level of the kitchen.
Luxemburg understood this. The soviets worked because the workers had skin in the game — their lives, their neighborhoods, their actual conditions. When the party substituted itself, it removed the people with skin in the game and replaced them with people executing directives. The intelligence drained out. Coercion filled the gap.
Luxemburg understood this in 1904. Her 1918 essay understood it again — not as theory now, but as observation of a revolution she could see was already dying. She wrote it in prison, unable to intervene, knowing it wouldn't be published in her lifetime.
She was shot in January 1919, five months after writing the essay. Her body was thrown into the Landwehr Canal. They found her nearly five months later.
She had been right about Lenin and right about the revolution. She did not live to be wrong about anything.
The tradition she was pointing at — not her economic analysis but her organizational insight, the claim that distributed self-governance is not idealism but architecture — didn't die with her. It ran in parallel with the dominant tradition, was suppressed by it, kept being right, and kept building.
The council communists, who argued that the factory council is the basic unit of socialist organization because it's where workers actually hold the knowledge to govern — not a party headquarters, not a central committee, but the people doing the work. The Catalan collectives, running for a year. The cooperative movements building across a century on six continents. The Zapatistas, refusing state power because they understood what taking it does. Cooperation Jackson, patient about the timeline, building from the most marginalized.
The tradition wasn't buried because it was wrong. It was buried because it was threatening in the specific way that the party couldn't absorb: it demonstrated that the party was unnecessary.
And look at who built it. Not the theorists debating in exile. The workers who occupied the Fiat factory on a Tuesday morning. The Aragonese peasants who collectivized their villages by assembly. The sailors at Kronstadt demanding what the revolution had promised. The people the Marxist debate had always treated as the object of liberation rather than its subject. They didn't need the theory. They had the practice. They had always had the practice — in the zanjera, in Palmares, in the Valencia tribunal, in every commons that had been governing itself for centuries before anyone wrote a manifesto about it. The knowledge produced on the outside, as Robinson named it, doesn't stay on the outside. It builds.
Luxemburg in her prison cell in 1918, feeding birds she had named, writing the analysis that has outlasted the people who imprisoned her — she was doing the war of position. Not stopping the Bolsheviks from eating the soviets. Making sure the next generation understood why it had happened. Building something durable enough to survive her death, the suppression of her manuscript, and the thirty-five years of Soviet history that proved her right.
It wasn't buried. We're still reading it. And in reading it, we're in the tradition.
A book that argues against vanguardism — against the party that knows better than the people — is still a person standing up to explain the thing. Luxemburg in her cell was also doing something the party structure couldn't do: making the thinking available, rather than the direction. I am an open source developer by training — which is to say, I come from a tradition that understood this before I had the political vocabulary for it. The GPL, the Linux kernel, the open protocols of the web: all of them are Luxemburg's argument applied to code. Not a party controlling the fork. The commons governing itself through contribution. This is source, not instructions. Whether it's received that way is bigger than me. I can't control what you build with it, and that's the point. The tradition Luxemburg was pointing at has never needed a party. It has needed people who understood the architecture well enough to recognize it when they saw it, and to build it when the conditions allowed. That's what this is for.
The next chapter is who has been building.