On a Wednesday this June, two presidents — one American, one Iranian — signed their names to a single piece of paper that ended a war. They never shared a room. The signatures arrived remotely, from separate countries. The document ran to fourteen points — it reopened the Strait of Hormuz, unfroze Iranian assets, promised “investment” money, and started a clock ticking toward a final deal. On paper.
That same week, under the dark skies over southern Lebanon, the event that would test the signing was already in motion. Airstrikes, in a conflict that neither signatory fully controlled, were carried out by a party that hadn’t signed the agreement and didn’t see itself as bound by it. Within days, Iran declared the strait closed. Ships kept moving through it anyway, as the tracking data showed. A peace had been signed by two countries, while what could shatter it was already at work elsewhere.
This isn’t a column about who is right. It’s about a common mistake that’s easy to overlook and ultimately has nothing to do with the Middle East. The mistake is believing your system ends where your map does.
* * *
Every steward of a complex system—your family, for example—starts by drawing a boundary. You have to. You can’t hold everything in your mind at once, so you decide what is inside the problem and what is outside it. For negotiators, the problem was a pair: two states, a nuclear program, a column of sanctions, a strait. Draw that line, and the work becomes manageable. Objectives align. Leverage becomes clear. The plan gains the clean lines that make it feel like rigor.
But the boundary is also the blind spot. The same line that makes a problem solvable determines, in advance, what you will never look at. And complex systems have a habit of hiding the most important thing just beyond that line.
Look closely at the signatures themselves. They are cheap. They cost an afternoon and a press release. But what they promise — to halt a program, withdraw a navy, deliver Israel’s restraint — is expensive, and the price is paid later by multiple nations over months and years. A signature shows what a party is willing to do. It does not reveal what the party is capable of doing.
You could see the gap the day it was announced. The same agreement produced two statements that did not describe the same thing. One president sold closure — it is signed, the oil is flowing, the other side can never have the weapon. The other sold dignity — we did not trade our pride, we kept our right to enrich. A victor’s announcement, pitched to the markets and the home crowd; a survivor’s, pitched to a nation that had just buried its leader and could not call this a win. Neither was speaking to the other. Each was telling the people who could hurt him that the open question had broken his way — and it was the same question, left open in the document precisely so that both of them could claim it.
Here, the machinery reveals itself. The agreement front-loads its rewards — the money, the relief — to the moment of signing, before any of the costly work is done. Watch what that selects for. A system rewarded for committing before actually acting learns to perform. The reward flows to the display, and the display is what you get more of. The gap between what was signed and what gets done does not close. It widens — and the structure of the agreement is what causes that widening.
This is the difference between naming a symptom and understanding a mechanism. “The deal is fragile” is a symptom. “The deal pays for the performance and not the enactment, so it manufactures its own fragility” is the mechanism. A steward who stops at the symptom treats the wrong thing, confidently, for as long as the budget lasts.
* * *
Now, the turn. Ask what the boundary was built to prevent: what truly determines whether this holds? Follow it honestly, and the answer isn’t within the US-Iran dyad at all. It’s the neighboring conflict — the strikes in Lebanon, by the party that never signed. That conflict connects to the agreement like a live wire across a circuit it was never wired into. A single night’s strikes in Lebanon can produce a global chokepoint in the Persian Gulf within a day — and the entire settlement quakes.
The critical factor lies outside the boundaries that negotiators established. Here is the part that should stop a steward cold: the link most capable of breaking the system is the one the most powerful party sits closest to — and has publicly promised not to touch. The lever that controls the outcome isn’t in the rival’s hands. It’s in your own, placed on the table next to a vow not to pick it up.
That is the error in its simplest form. We search for the source of a system’s behavior within the system we have defined and within the rival we have identified as our problem. The source is often an edge we left off the map, running through a relationship we see as ours to protect rather than ours to govern.
There is a subtler trap beneath it, and it is the one that flatters us most. When you cannot see inside a system that does not want to be seen, you read it through whoever is visible — and the visible people are often the ones your own actions have left standing. A war removes some leaders and raises others. A negotiation draws out the pragmatists willing to sit at your table and leaves the hardliners in the room next door. Read the system through the survivors and you will mistake the part that is talking to you for the whole that is deciding. Even worse, the survivors learn which lens you are using and hold it up to you. Their reasonableness confirms your belief that they are reasonable. From the inside, the loop feels like mounting evidence. It is the sound of your own assumptions coming back with interest.
And then there is the comfort of the long average. This isn’t the first ceasefire in this conflict. It’s the latest in a series — announced and broken, announced and renewed. Each time, the announcement carries the same quiet hope: ceasefires usually lead somewhere. Perhaps they do — taken across all the ceasefires in all the conflicts there have ever been. But you are not living across all of them. You walk a single path, and so far the sequence reads: truce, break, war, truce, break. The average of all the histories that could have been is not the history you are standing in. A steward who relies on the cheerful average is preparing for a system different from the one beneath their feet.
* * *
What, then, is the move? Not a smarter plan aimed at the rival. The discipline is older and quieter than that, and it cuts against everything our instincts tell us.
You cannot close the system. You cannot command the weather or order the sea to lie flat. What you can do is stop pretending the boundary you drew is the boundary that actually exists. Redraw the map to include the edge that governs you — even, especially, when that edge runs through a relationship you would rather not examine. Before you reach for leverage against the other party, check whether the binding constraint is a lever already resting in your own hand. And accept that in a system this tightly coupled, the most disciplined move is sometimes to do less rather than more: to protect what is actually working, to take the smallest real step and observe how the system responds, and to refuse the grand, all-at-once gesture that a system with no capacity to absorb it will simply break against.
None of this is passivity. It is the difference between the engineer, who blueprints an outcome and builds to it, and the gardener, who shapes the conditions and tends what grows. The engineer is certain. The gardener is patient and a little humble because they know the soil gets a vote. In a complex system, you do not control the result. You can only nudge the odds of a better one — and you forfeit even that the moment you mistake your map for the territory.
* * *
The negotiation will resolve in some way none of us can yet see. The reading is never complete, and any account of it, including this one, is a hypothesis with a timestamp. But the lesson doesn’t wait for the outcome, and it doesn’t stay in the Middle East.
You run a team, a company, a coalition, a household. You have drawn a boundary around your problem, and the boundary feels like clarity. But feelings aren’t facts. Somewhere just outside it is an edge you chose not to look at — a relationship you treat as fixed, a party you decided was not your concern, a lever resting in your own hand. The system is not sealed. It never was. The only question is whether you will redraw the map before the edge you left off it redraws the map for you.