The War Nobody Wins

i · the genre runs on schedule

Run the tape back and the recursion is almost funny, if you like your humor cold. In 1979, sixty-six Americans were taken hostage in Tehran, and even while the crisis dragged on, the Carter White House was working a quiet diplomatic channel through Iran's own foreign minister — a channel serious enough that the hostages eventually came home, and dead-ended enough that the minister running it was executed by his own government within the year. The lesson America took from that wasn't "our counterpart got killed for talking to us." It was "you cannot negotiate with Iran." First draft of the essay, filed.

Redraft it for 2002. After September 11, Iran — no friend of the Taliban, and no fan of al-Qaeda — quietly cooperated with Washington in Afghanistan and floated the outline of a genuinely broad settlement: nuclear transparency, recognition of Israel's right to exist, a normalization track. The Bush administration's actual response was to put Iran in the same paragraph as Iraq and North Korea and call it the Axis of Evil. The diplomatic opening didn't fail. It was never opened. Second draft, filed under the same headline, the falsification treated as further proof of the thesis rather than as its cause.

Redraft it again for 2015. The Joint Comprehensive Plan of Action got signed, got Iran's centrifuges packed in storage under inspectors' cameras, and held — technically — for three years, which by regional standards is a long engagement. It ended not because Iran cheated first but because a new American administration walked out of it unilaterally in 2018 on the theory that the deal was, per the recurring essay, a trap. Third draft. Same conclusion, one more instance of the conclusion causing its own evidence.

And now the fourth draft, live this month, pegged to a war that has actually been running since February — a shorter fuse than any of the prior three, and worth looking at directly instead of through the essay's lens.

ii · what actually collapsed

Strip the rhetoric and the sequence is legible. Following the collapse of the Geneva track and a brief war in 2025, coordinated American and Israeli strikes at the end of February hit Iranian military, nuclear, and leadership targets. Iran answered by closing the Strait of Hormuz to Western-flagged shipping — mines, drone strikes, speedboat harassment, GPS jamming, over fifty vessels attacked and more than a dozen seafarers dead by summer — while quietly waving through tankers from China, Russia, India, and a short list of other capitals it still does business with. Washington answered the closure with a blockade of its own in April, intercepting anything that had paid Tehran's toll to pass. Two blockades, running the strait in opposite directions at once, each one justified by the other.

In June, both sides did the thing the recurring essay insists they never do: they made a deal. On the 17th, Washington and Tehran signed a fourteen-point memorandum ending the shooting, lifting both blockades, reopening the strait, and setting a sixty-day clock to negotiate the harder problem — the nuclear file — with sanctions relief on the table as the incentive to get there. For about three weeks, tankers moved and the essay had nothing to write about.

Then, on the 8th of July, Iran struck several commercial vessels in the strait. By the 12th, the IRGC navy had declared the strait closed again, citing Israeli strikes elsewhere as the provocation — a claim Washington disputes. On the 14th, cruise missiles hit three more tankers and killed a sailor. The United States reimposed its blockade the same week. Brent crude, which had eased under $80 during the MOU, spiked back over $120 — the kind of energy shock the world hasn't priced in since the 1970s. The sixty-day clock is still nominally running. Nobody is pretending the memorandum is not on life support.

iii · the prophecy that writes itself

Here is the actual mechanism, and it has nothing to do with Iranian national character or American bad faith specifically — it is what happens whenever two parties stop sharing a frame for what the other side's actions mean. Every Iranian strike gets read in Washington as confirmation that the regime was always going to break the deal, was in fact only using the ceasefire to rearm and reposition — never mind that the same reading would apply to any nation's forces regrouping during a lull. Every American blockade and every allied strike gets read in Tehran as confirmation that the West only offers deals to buy time for the next attack — never mind that a memorandum lifting a mutual blockade is, on its face, a de-escalation. Both readings have some evidence behind them. Both readings also guarantee their own future evidence, because a government that expects betrayal moves to pre-empt it, and a government moving to pre-empt betrayal looks, to the other side, exactly like betrayal arriving on schedule.

This is the distortion doing the work that a shared reality would otherwise do. Neither Washington nor Tehran is lying, exactly, when each says the other started it — they are each accurately reporting what they saw from inside a frame the other side cannot see into. The war isn't a disagreement about facts. It's a disagreement about which facts get to count, and that gap is what the missiles are actually settling, because nothing else is available to settle it. Call it what it is: an irrepressible conflict, in the old sense — not a dispute that compromise can split down the middle, because there is no shared frame left to compromise inside of. The distortion isn't a side effect of the war. It is the war, wearing armor.

None of which excuses the immediate facts on the water. Iran struck civilian tankers and killed sailors who had nothing to do with any of this; that is not a rhetorical frame, it's a body count, and the regime that ordered it owns it. But "Iran did this, therefore Iran always does this, therefore nothing but pressure will ever work" is the exact rhetorical move that has been available, and wrong, every single time this essay has run since 1979 — because it forecloses the one thing that occasionally has worked: an agreement neither side immediately sabotages long enough to see if it holds. The JCPOA held for three years. The Islamabad memorandum held for three weeks. The pattern isn't that engagement fails. It's that engagement gets a shorter and shorter leash each time before someone — often, not always, on the American side — decides the leash was never real.

iv · which draft comes next

The safe prediction: this MOU does not survive its own sixty-day clock. Either it gets renegotiated into something thinner and gets signed again in the fall, or it dies quietly and the dual blockade becomes the new baseline until the next crisis forces a new ceasefire, which will also get called historic, and will also collapse, on a timeline shorter than the last one. Somewhere in that cycle, the essay runs again. It will not be written by a bad person. It will be written by someone who noticed, correctly, that this deal broke — and who will not notice, because the genre doesn't ask him to, that every deal before it broke the same way, for the same structural reason, regardless of who signed it or what year it was.

The op-ed calls it a character flaw in one country. The pattern says otherwise: it's what any two powers do when they've stopped being able to read each other's evidence the same way. Iran will still be Iran the next time someone signs a deal with it. The only variable that has ever changed the outcome is whether either side was willing to keep believing the deal past the first provocation. So far, across four decades and four drafts, the answer keeps arriving early.