For a little over a decade, a region in northern Syria ran itself through elected councils instead of a president or a monarch. Every governing position, from the smallest neighborhood commune up through the regional executive councils, was held by two co-chairs, one man and one woman, elected separately, deciding jointly, and wherever possible drawn from different ethnic communities so the office itself reflected who actually lived there. A fifty percent quota set the floor for women’s representation in every council that existed. None of it was symbolic. Women fought for the co-chair system against real resistance from men in their own ranks, because they had already learned what happens to equality that never gets written down.

It’s worth being specific about what that actually looked like day to day, because “gender parity” can sound like a talking point until you see the structure underneath it. Alongside the co-chair councils ran a second, fully autonomous women’s political body, Kongra Star, tracing back to an underground women’s union founded in 2004 under Baathist repression. Kongra Star wasn’t a women’s caucus inside the main political structure. It was its own parallel confederation, organizing women separately across communes, the economy, education, and self-defense, answerable to women rather than folded into a mixed body where it could be outvoted. At the neighborhood level, that structure showed up as Mala Jin, women’s houses, present in every city and most towns, staffed by local women, often the mothers and grandmothers of the neighborhood, who mediated domestic violence, forced marriage, and family disputes directly, with a case only escalating to the formal courts if the women’s house couldn’t resolve it first. The laws those courts actually enforced had changed too: forced marriage, polygamy, and child marriage were banned, honor killing lost its legal cover as a mitigating circumstance, and divorce rights were rewritten to no longer assume a woman needed her husband’s consent to leave him. None of this was a separate track that ran alongside the real politics. Kongra Star operated as the women’s movement’s counterpart to TEV-DEM, the umbrella body that organized everything else, which meant it had standing to actually shape policy, not just comment on it from the side.

I want to be precise about what that structure actually was before I say anything else about it, because getting this wrong is how these arguments fall apart. This wasn’t a leaderless experiment in the strict anarchist sense, and I’m not going to pretend it was for the sake of a cleaner story. The Democratic Union Party, the Syrian branch of the Kurdistan Workers’ Party, organized and directed the councils from the start, and Abdullah Ocalan, the PKK’s imprisoned founder, remains close to a mandatory ideological reference point across the entire structure. Ocalan converted to a version of Murray Bookchin’s libertarian municipalism after his imprisonment, and what got built in northern Syria was that conversion put into practice: democratic confederalism, a real and specific tradition, related to anarchism, indebted to it, but not identical to it. A single party stood behind the councils the whole time. Calling that anarchism without qualification oversells it. Calling it an honest, high-stakes experiment in Bookchin’s ideas, run by people who had every reason to keep enough central coordination to survive a war, undersells nothing and gets the history right.

As of this year, that project is being dismantled. Not by its own failure alone. By an army, and by weaknesses in its own structure that made the army’s job easier.

I want to sit with that fact instead of rushing past it, because it’s the kind of fact that gets used two ways, and both uses are lies.

The first lie says Rojava proves anarchism doesn’t work. Something held for over a decade, under war conditions, at the scale of an actual region with actual courts and actual schools and actual harvests to bring in, and then a bigger army arrived and it stopped existing. Case closed, the argument goes. Anarchism is a nice theory that dies the moment reality shows up.

Watch what that argument actually does. It takes the single hardest test any project inspired by these ideas has ever faced, invasion by a state’s standing military, and treats losing that test as proof the underlying idea was never sound. Nobody applies this standard anywhere else. The state of California didn’t stop being a legitimate government the one time it lost a war it wasn’t fighting. Nations get invaded. Nations get occupied. We don’t conclude from an occupation that nationhood itself was a bad theory. We conclude that a smaller, less-armed party lost to a larger, more-armed one, which is a fact about relative military capacity, not a verdict on the political arrangement that got overrun.

France in 1940 didn’t stop being a functioning republic because the Wehrmacht occupied it in six weeks. Kuwait in 1990 didn’t stop being a legitimate state because Iraq’s army crossed the border in a night. Historians don’t spend their time arguing that constitutional monarchy or parliamentary democracy is a discredited system because both of those governments briefly ceased to control their own territory. They ask what happened militarily, and separately, whether the occupied government’s internal arrangements had anything to do with how it fared. Rojava deserves the same two separate questions, not one collapsed into the other for the sake of a tidier argument.

The second lie is worse, because I’ve caught myself reaching for it. It says Rojava was purely a victim, an innocent experiment crushed by outside forces it never provoked, its failure entirely explained by geopolitics that had nothing to do with anything the movement itself did. This version is comforting. It’s also not fully true, and a movement that only tells itself comforting versions of its own history is a movement that stops learning.

Here’s what actually happened, as close as I can get to it. Assad’s regime fell in December 2024. Syria’s new government moved fast to reassert control over the northeast, and by January 2026, after talks that repeatedly stalled, government forces retook most of the territory the Kurdish-led administration had governed, including cities that had been under its control for years. What’s left of the Kurdish forces has been pushed back into their historical heartland, and the broader governing structure has been formally folded into the Syrian state.

That’s the invasion story, and it’s real. But most of the territory that fell first wasn’t Kurdish. It was Arab-majority, and the record there was genuinely uneven. Family-law reforms and curriculum changes that got implemented fully in Kurdish areas stayed partial, or absent, in Arab ones. There were protests over years against what residents called discrimination and forced conscription, well before any tank moved. When the offensive came, tribal fighters who had fought alongside the Kurdish-led forces for years defected instead of resisting. Some residents welcomed the army that was, on paper, dismantling the more democratic option.

Abdulmajid Tuma watched Syrian soldiers take the streets of Raqqa in January and told a reporter from New Lines Magazine, “We felt like fourth-class citizens.” He’s forty-five, born there, and he wasn’t talking about the war. He was talking about seven years of living under an administration that never gave his own city real power over itself. “Raqqa is 95% Arab,” he said. “Yet we had no power.” That’s not a man mourning the fall of a democratic experiment. That’s a man who never experienced it as one, in the place he was born, under the years of governance that were supposed to be the proof of concept.

I keep turning that fact over because it’s not a footnote. It’s the same argument I’ve made about horizontal groups my whole career, showing up at the scale of a region instead of a meeting room. A structure can be genuinely democratic for the people it was built around and still fail to extend real representation to the people it later came to govern. That gap doesn’t cause an invasion. But it’s exactly the kind of internal weakness that makes an invasion cheaper, because a population that doesn’t feel it has a stake in defending a government won’t defend it very hard.

There’s a second weakness sitting right next to the first one, and it’s the harder one to say out loud, because it points back at the structure I just spent two paragraphs describing with real admiration. The co-chair system was real. Kongra Star’s autonomy from men’s councils was real. And all of it operated, the entire time, inside a wider structure the Democratic Union Party organized and directed from the top. Kongra Star being answerable to women rather than to a mixed council is a genuine achievement. It is not the same claim as Kongra Star being independent of the party, and I don’t think it was. Formal gender parity, even parity with its own autonomous confederation behind it, isn’t the same thing as the absence of one-party dominance, and pretending those are the same claim is how admiration curdles into propaganda. I don’t think the parity was fake. I think it was a genuine, hard-won achievement that happened to sit inside a system that never seriously tested what would happen if a council, or Kongra Star itself, decided against what the party wanted. Both things are true. A movement that can only hold one of them in its head at a time isn’t being honest about what it built.

So here’s what I actually think Rojava proves, and it isn’t either of the comfortable stories.

It proves horizontal governance can outperform a bureaucracy at the actual work of governing: coordinating agriculture, running schools, adjudicating disputes, building something close to gender equality from a standing start, for over a decade, under conditions that would have broken most institutions I’ve ever worked inside. That’s not nothing. That’s the hardest kind of proof there is, because it wasn’t a thought experiment. People lived their whole adult lives inside it.

It also proves that horizontal governance, like every other kind of governance humans have ever built, still has to solve the problem of who gets to belong and who gets a real voice, and that unresolved problem doesn’t go away just because the people at the center of a movement have their hearts in the right place. Freeman said the same thing about small groups fifty years ago: refusing to name a hierarchy doesn’t dissolve it, it just makes it harder to see and harder to fix. Rojava’s version of that failure shows up twice, not once. Once at its edges, with the populations it governed but hadn’t been built around. And once at its own center, where a real and celebrated structure of parity operated inside a hierarchy, the party, that never got named as one.

And it proves the hardest thing of all, the one I don’t have a comfortable answer for. Mutual aid can out-organize a welfare office. Horizontal accountability can out-repair a courtroom. A cooperative can out-survive a shareholder-owned company through a bankruptcy that would have gutted a normal firm. None of that has an answer yet for a hostile army with nothing stopping it. Rojava survived as long as it did partly because outside powers needed it against ISIS and a fractured Syria couldn’t move against it. The moment both of those conditions lifted, the region’s own defensive capacity, real as it was, could not hold. I’ve spent a career arguing that people don’t need someone standing over them to cooperate well. I have not yet found the argument that answers what happens when someone standing over a much bigger army decides the cost of ending your experiment has finally dropped low enough.

I’m not going to manufacture one here. Pretending I had it would be worse than admitting I don’t.

I’ve spent twenty years inside institutions that ran on the same failure at a much smaller scale. Every workplace I’ve worked in has had an org chart, and every one of those org charts left out the actual channels power moved through, who really got heard in a meeting, whose objection got treated as a delay and whose got treated as a veto. I’ve watched a team call itself flat and collaborative while one person’s informal buy-in decided everything that mattered, and I’ve watched that same team act baffled when someone who wasn’t in on the informal channel walked out. Rojava’s fracture is that pattern at the scale of a region instead of a department, with stakes that were never abstract. The lesson isn’t different because the scale changed. It’s just harder to unsee.

What I will say is this. An idea that can run a functioning, comparatively humane, gender-equal regional government for a decade under siege conditions has already answered the only question that matters about whether horizontal organization can work. It can. The remaining questions, how it extends real inclusion to people it didn’t build itself around, and what it does about power that answers to nothing but its own size, are not reasons to abandon the idea. They’re the actual unfinished work. Every political arrangement humans have ever built, including the ones currently occupying Rojava, has unfinished work exactly like it. The difference is which ones are honest about that, and which ones just have better public relations.

I’ll say one more thing, because it’s easy to write everything above from a desk and forget who actually paid for it. People built this under an active military threat, with weapons taken from whoever would supply them, the same choice Spanish anarchists made in 1936 and nobody retroactively disqualifies them for it either. Naming a structure’s limits isn’t the same as telling the people who built it, at that cost, that they should have done something cleaner. I don’t think they should have. I think they built something real, with real flaws, under conditions I’ve never had to organize inside of, and both halves of that sentence are true at once.

I go into all of this in more detail, along with a decade’s worth of other cases where this same pattern shows up, in Anarchist’s React: Mutual Aid, Individual Responsibility, and the Way We Actually Live, TBD, hopefully out in 2026 from CtrlAltDelusion Press.