The Strait of Hormuz is closed. Tankers have burned in a neighbor’s territorial waters. A crew member is dead, another missing, and a waterway that carried a fifth of the world’s oil now passes a handful of ships a day under escort and threat. Into this, last week, came an essay in Foreign Affairs from Javad Zarif — the diplomat who negotiated the 2015 nuclear agreement, twice the public face of Iranian statecraft — laying out his country’s vision for ending the war and rebuilding the region.
You should read it. But not the way most people will read it — for its argument, to agree or object. Read it as a diagnostic artifact. A document like this is what I call a displayed path: a staircase the author shows you, landing by landing, toward a future he wants you to believe is already under construction. Complexity science gives us the tools to grade that staircase — to separate the landings the system actually holds from the landings the author needs it to hold. Zarif’s essay is unusually valuable because it contains both, and the line between them runs exactly where the science says it should.
Start with the author, because he is the system’s first node. Zarif built his career on one asset: the existence of a diplomatic path. Every negotiation that opens restores his currency; every war that resumes debases it. And he writes now from outside a decision apparatus that has been decapitated and rebuilt around him — the leader he served for decades killed in the war’s opening strikes, succeeded by a son consolidating power through the security services, not the foreign ministry. A man in that position does not simply describe the future. He describes the future in which men like him still matter. That’s not an accusation. It’s a boundary condition — the initial condition of the document itself — and it tells you where to press.
* * *
Begin with where he gets it right, because his strongest passage is the one that costs him the most to write.
Midway through the essay, Zarif concedes that Iran’s leverage in the strait is fragile — that a sustained closure will drive the world to build pipelines, ports, and land corridors that bypass the waterway altogether, and that once markets diversify away from Hormuz, the threat of closing it deters no one. He is arguing against his own country’s strongest card. In signaling terms, that is what makes the passage credible: costly admissions are the ones worth believing.
And the science behind his worry is exactly right, though it is moving faster than his essay admits. This is non-ergodicity — the property of systems whose history forecloses states rather than revisiting them. An ergodic system wanders its whole landscape given time; averages over many parallel worlds tell you what your world will do. Measure the temperature at one spot in the room you’re sitting in for one hour, and it’s the same as measuring every spot simultaneously for one second. Interchangeable averages – that’s the signature.
A non-ergodic system, on the other hand, passes one-way doors. The strait’s value as leverage is behind one of those doors now. The Emirates have a second bypass pipeline, roughly half built. Saudi Arabia is pushing millions of barrels a day across the peninsula to the Red Sea. Gulf producers are asking Japan to help finance additional routes, and regional executives describe bypass investment as common sense rather than a contingency. Capital of that scale, once sunk, does not un-sink. Every week the closure persists, it finances the infrastructure that makes the closure irrelevant — a feedback loop in which the exercise of leverage consumes the leverage. An American energy official put it in a single sentence: this is a card you can play once. Zarif knows it. It is the most honest sentence-length idea in his essay, and it is the one his own government is spending down as he wrote it.
He’s right about a second thing, too, and the science behind it is the geometry of attractor basins. The war with Iran was premised on the idea that enough pressure would tip the Iranian state into collapse or capitulation — that the system sat near the rim of its basin and a hard shove would push it over the edge into a new basin, one where the Iranian regime would become tractable. It did not. The state absorbed the assassination of its supreme leader, ran its succession under bombardment, and held its territory. When a system takes that magnitude of shock and does not change basins, the honest inference is that the basin is deeper than the planners’ model assumed — and further pressure of the same kind mostly deepens it, because siege consolidates the besieged. Even inside the American government, a quieter assessment has surfaced publicly: you can destroy radars and missiles, but it’s simply too easy to fire at ships in a strait, and a problem with that shape is not solved by force alone.
Zarif’s larger structural claim follows: security architectures built on excluding a major regional node have repeatedly failed, and they fail the way networks with severed hubs do — the excluded node, Iran, does not disappear; it reroutes, and the rerouting is the instability. The record since 1979 is clear on this. Shut out of the Gulf’s formal security order, Iran built an informal one: Hezbollah in Lebanon, the militias that filled post-2003 Iraq, the Houthi missile arsenal now reaching shipping lanes that Iran’s navy never could. Cut off from the dollar system, it built a shadow economy — a ghost tanker fleet with transponders dark, discounted crude moving east, barter chains that no sanctions office can fully map. Each exclusion produced not an absence but a shadow parallel network running outside the channels where states can see it, price it, or negotiate with it. That is the network scientist’s point exactly: sever a hub and the traffic doesn’t stop — it moves to routes with fewer rules and even less visibility. On this, Zarif is mostly right, and forty-plus years of rerouting is the evidence.
* * *
Now the other half — and the failures are not rhetorical excesses. Each one is a specific, nameable error in complex-systems reasoning.
The first is a survivorship bias, elevated to a national narrative. Zarif reads the state’s survival as proof of the resilience of the nation — the system endured, therefore the system is strong and its people united. But survival is an outcome, not a mechanism, and a lone survivor tells you almost nothing about the process that produced it. The counterfactual worlds where the succession fractured or the security forces split are invisible precisely because they didn’t happen here. What the observable record actually shows is a state holding itself together by suppression: an uprising in January put down at terrible cost, prison protests running as his essay went to press, factions maneuvering under a new and untested leader. Complexity science draws a hard line between two kinds of persistence. Adaptive systems persist by reconfiguring — absorbing the shock and becoming something that fits the new environment. Brittle systems persist by clamping — spending energy to hold a configuration the environment no longer supports (something the USMIL and USG arguable do as well). The second looks identical to the first right up until it doesn’t, because clamping stores the stress it refuses to metabolize. Zarif’s essay cannot tell these apart, because telling them apart would require auditing his own principal — and the essay audits everyone in the region except the government it speaks for. That reflexive silence is the document’s deepest tell.
The second error is a decoupling of the display from the mechanism. Zarif frames the closure of the strait as defensive screening: Iran cannot distinguish commercial cargo from war materiel, so it restricts the flow until the war ends. But watch the mechanism, not the message. Ships are being struck not for carrying weapons but for declining Iran’s route-authorization scheme; tankers were hit in a neighbor’s territorial waters; the operating logic on the water is tolls, permissions, and enforcement — a control-and-revenue architecture, not a checkpoint. In any complex system, when the displayed rationale and the operating mechanism diverge, the mechanism is the truth of the system. Behavior is expensive; explanation is free. The configuration Zarif shows — responsible co-administration of an international waterway under law — is not one his state currently operates. It’s one his essay stages.
The third error is a missing staircase. Herbert Simon observed sixty years ago that complex systems assemble through stable intermediate forms — configurations that can stand on their own while the next level builds on top of them. A cell had to be viable before tissue, tissue before organ; a village had to function before a kingdom, a kingdom before an empire. Nothing durable jumps straight to its final form because the half-built version has to survive in the meantime. Evolution cannot jump; neither can regional order. Zarif’s endpoint is a West Asian security community with a shared development fund and a joint nuclear-enrichment consortium — pooled fissile-fuel infrastructure among states currently shooting at each other’s shipping. Between here and there, his essay offers no intermediate form that could survive on its own: no configuration in which Gulf states extend trust to a neighbor that just killed their sailors, and every observable move those states are making — the pipelines, the alternative ports, the hedging capital — runs in the opposite direction. A destination without stable intermediates is not a path. It is a painting of one.
The fourth error — or perhaps the fourth design feature — is an asymmetry in the deal's irreversibility. Look at what each side is asked to spend. The American concessions are absorbing: permanently lifted sanctions, removed terrorism designations, a three-hundred-billion-dollar fund, recognition of Iranian administration over a global chokepoint. Each is a one-way door — legally, politically, precedentially. Once the world accepts that a state can close the planet’s most important waterway and be paid to reopen it, that lesson is learned everywhere, by every state that sits on a chokepoint. The Iranian concessions are the opposite: ratifying an inspections protocol a future government can exit, codifying the anti-nuclear fatwa of a leader who is now dead — a religious ruling his successor did not issue and does not obviously own. Cheap to display, cheap to revoke. In systems language: one side exchanges irreversible states for the other side’s reversible signals. That is not a criticism of negotiating with adversaries. It is a criticism of this exchange rate.
* * *
So what should a decision-maker do with a document like this? The tempting answers are the two lazy ones: credit it, because its author is a proven negotiator and half of it is true; or discard it, because its author is an adversary’s advocate and half of it is staged. Both are errors, and the second is subtler than it looks — throw out the essay, and you throw out the true half, the half that says the state endures, exclusion keeps failing, and the strait problem will not be closed by bombs.
The deeper risk is what engineers of complex systems call a functional failure — the kind that leaves everything downstream looking healthy. Systems are built function-first: purpose shapes structure, structure produces behavior. Get the function wrong, and you can execute the structure flawlessly and still fail, because every well-built component now serves the wrong purpose. If Washington accepts Zarif’s framing of what the strait dispute is — a security dilemma awaiting the right legal formula — then every downstream structure of a deal, however well negotiated, inherits the error. The behavior on the water says the dispute is something else: a coercion mechanism racing its own obsolescence, wielded by a state whose display and machinery have come apart.
Read the essay, then, the way you would read any strategy document that crosses your desk — a competitor’s roadmap, a partner’s five-year plan, a subordinate’s turnaround pitch. Grade the staircase landing by landing. Ask which configurations the author’s system actually operates and which the author needs you to believe it operates. Ask whether the persistence you’re shown is adaptation or clamping. Ask what each side’s concessions cost to reverse. And ask, always, what the author’s own position in the system requires the future to look like — because that requirement is written into every paragraph, whether the author knows it or not.
Zarif told the truth about one thing above all: the strait is a card you can play once. He wrote it as a warning to his own government. It doubles, unintended, as the essay’s verdict on itself — a displayed path built on leverage that is depreciating in the hand that holds it, argued by a man whose relevance depreciates with it. The clock he describes is real. It’s just not running in the direction his staircase needs.