Who Is Allowed to Coordinate in the Middle East Crisis

In late May, with the talks to end the Iran war stalled, the US president threatened to blow up Oman if it did not, in his words, behave like everybody else. Oman had done nothing an enemy does. It had spent the war as the quiet back-channel between Washington and Tehran, the role it has played for decades, and it had tried to turn its standing over the Strait of Hormuz, whose navigable lanes run as close to Omani waters as to Iran’s, into a priced, co-administered arrangement. That was the offence. A small state was threatened with destruction, by the power it was serving, for trying to charge for a passage it helped manage.

Set that beside Pakistan. Over the same weeks another mediator sat in almost the same position, a working line to Tehran and a working line to Washington at once, and was rewarded for it, its channel to Washington warmer than before and its leaders praised from the top even when others there wanted them cut loose. Two intermediaries, nearly the same position, opposite fates. That is the puzzle, and raw connection does not solve it. A state does not own its proximity to the adversary. The hegemon lends it, and can call it back.

That the two sat in the same place is not an impression. In April I proposed a framework to map the regional security complex as a set of directed ties, scoring each relationship on a scale of operational connectivity from zero to one and reading the crisis through who was connected to whom rather than through the balance of power. On that map Oman and Pakistan were near-twins. Almost every other state leaned to a single pole. These two, and only these two, held real ties to both at once, Oman at 0.4 to Iran and 0.5 to Washington, Pakistan a notch stronger at 0.5 and 0.6. Here is that map.

April 2026

This is a risk-analysis tool more than a measurement. The point is to approach connectivity as something that can be mapped, one relationship at a time, and then read for the behavioural patterns the whole system produces. The weights are analytical judgements, deliberately approximate. A weight at or above 0.6 is coordination, a tie that carries load without constant tending. Around 0.3 is friction. At or below 0.2 the tie has effectively stopped transmitting.

Look wider than those two, and the rest of the map has a simple shape. Almost every other partner is adjacent to a single pole. The Gulf states cluster around the American coalition and hold no live operational relationship with Tehran of the kind a mediator would. When Iran struck Saudi Arabia, Qatar, the UAE, Kuwait and Bahrain in its opening barrages, it was reaching through them at the American bases they host and the energy assets that tie them to the Western economy, striking them as leverage over the coalition rather than as adversaries in their own right. Iran’s own column is near zero everywhere the coalition stands, its one tie of real strength running to its proxy network of aligned militias at 0.7. Only Oman and Pakistan held meaningful adjacency to both poles at once, and a node connected to both sides can carry what neither can carry directly.

Three months on, the same map looks like this.

July 2026

These figures freeze the system on 9 July, the day after Trump declared the memorandum over at the Ankara summit, and should be read as one fixed moment. The war has openly resumed in the days since.

The moved weights, shown in bold, cluster in five places. Pakistan rose on both of its edges, to 0.7 with Washington and 0.6 with Tehran. Oman fell on both, its reach toward Washington down from 0.5 to 0.4 and its Iran tie down from 0.4 to 0.2, while Washington’s directional weight over Oman held at 0.5. The Saudi-Iran tie lifted off fragmentation to a wary 0.3. The US-Israel bond cooled asymmetrically, Washington’s weight over Israel easing from 0.8 to 0.7 while Israel’s reach back fell from 0.8 to 0.6. The principal edge between Washington and Tehran reads 0.2, having made a round trip the two grids bracket rather than show. The aligned Gulf otherwise held its shape exactly, unmoved while the conduits shifted around it.

The war had opened on 28 February with a strike that killed Iran’s Supreme Leader, Ali Khamenei, and his son Mojtaba assumed the office in March. Through the late spring the proximity between the two governments nonetheless climbed off the floor. A ceasefire took hold in April and was extended. By 17 June the two sides had a signed text, a fourteen-point memorandum of understanding covering the ceasefire, an end to hostilities in Lebanon, the nuclear file, and the reopening of the Strait of Hormuz toll-free for sixty days, which Trump signed during a victory appearance in France while Iran’s president signed in Tehran.

Washington and Tehran had no direct line to run it through. It ran instead from Washington through Islamabad to Tehran, and through Doha and Muscat alongside, even the Lebanon front handled through a de-confliction cell facilitated by Qatar and Pakistan, without Israel, the party actually fighting there. The direct edge between the principals barely moved. The proximity that ended the war was manufactured through the conduits, and over the same weeks the two conduits that mattered split completely.

Pakistan should, on a reading of raw magnitude, have been the most exposed node of all. It held a nodal position, a place between two networks that will not speak directly, whose value is that the strong side needs it and cannot build it for itself. It held the strongest dual adjacency in the system, and it was the one significant regional state Iran never struck. Instead it spent its position rather than owning it.

When a report surfaced alleging that Pakistan was sheltering Iranian military aircraft on its airfields, and one of the most influential hawks in Trump’s own circle used it to argue Islamabad should be replaced, the response from the top was to reaffirm Pakistan without hesitation. The mediator was defended at precisely the moment its adjacency looked most compromising, because that adjacency was understood to be running in service. Its tie to Washington rose to 0.7 and its tie to Iran to 0.6, as it absorbed the high-level channel Oman could no longer carry. It rose on the very edges that should have condemned it, because it never tried to hold them for itself.

Oman held the same position and travelled the opposite arc. Two movements degraded it. The first was not of its own making. As the war widened, Iran struck Omani ports, ships and towns, and its Iran-adjacency, the very thing that had made it valuable, fell from 0.4 to 0.2 as the high-level mediation role passed to Pakistan. The second movement was Oman’s own. Having lost the service role, Oman reached for the standing that remained, its position over the Strait of Hormuz, and moved, from May onward, to convert that into something priced and co-administered. On 30 June it carried a Hormuz proposal into the Doha talks. The response was immediate. The line from Washington was that Iran cannot toll the strait, and the proposal proceeded only on condition that no toll attach. The node was permitted to carry a proposal. It was not permitted to bank one.

The obvious objection is that Oman simply became less useful after Iran struck it, and the fall needs no talk of ownership. That explains the first movement. It leaves the second untouched, because Washington’s threat to destroy Oman came months after the strikes, at the moment it tried to price the strait, and it was the pricing that drew the threat.

Through the war Trump had treated the blockade of Iran’s ports as his own lever, boasting it cost Iran 500 million dollars a day, and had spoken of seizing control of the strait himself. Oman’s reach toward Washington fell, in a way that suggests its sovereignty over that adjacency was conditional on never using it. It fell only as far as 0.4, because in early July France and Britain stepped in, co-signing a statement with Muscat to keep its waters safe, receiving the Sultan in Paris and London, the French press describing Paris as having rediscovered Oman’s mediation capacity. Even the punished node’s landing was cushioned, this time by a European conduit.

Two nodes, near-twins on the April map, inverted by July. What sorted them was not the size of the adjacency, larger for Pakistan on both edges, nor its direction, the same for both. On the reading proposed here, it was whose account the adjacency was spent on. The system lets a conduit stand close to the enemy. What it will not allow is the conduit turning a channel it was lent into an asset it can price. Call it a conduit ceiling. A mediator may carry almost anything for the strong side, but it may not own what it carries. The line is crossed when a conduit tries to turn its access or geography into an entitlement of its own, a toll it sets, a veto it holds, a role the principal can no longer revoke.

A third cross-pole tie moved more quietly, from the opposite end. Saudi Arabia is not a conduit. It brokered no round, and Iran bombed Riyadh and Saudi oil facilities during the war. But it holds, since the China-brokered normalisation of 2023, a live channel to Tehran no other Gulf capital maintains, and in July it used it in the most visible way available. When the state funeral for Ali Khamenei ran across the first week of July, a delegation from Riyadh led by a deputy foreign minister came to Tehran, received with a recitation invoking an ancient battle fought on Arabian soil, a message rather than a courtesy. And yet it came, and the rapprochement held, lifting the Saudi-Iran edge to a wary 0.3.

This drew no correction. Saudi holds the loosest tie to Washington any coalition partner keeps by choice and an open line to the adversary at once, and on the reading that condemned Oman it should be the most exposed hedge in the system. It is left alone, because Saudi neither spends its Iran channel in the hegemon’s service nor tries to price it. It holds it in silence and claims nothing from it. Adjacency kept in silence is tolerated exactly as adjacency kept in service is.

Then, this week, the manufactured edge collapsed. On 6 and 7 July, in the middle of that same funeral week, Iran attacked three commercial ships in the Strait of Hormuz, among them the Saudi-flagged Wedyan. The attack was the enforcement edge of an apparatus Iran had been assembling since the war began, from transit fees in March to a twelve-article sovereignty law over the strait in April and, on 5 May, the Persian Gulf Strait Authority, a body formalising discretionary, sanctions-resilient pricing of a global chokepoint. Its architect had called control of the strait an opportunity as precious as an atomic bomb. Iran was running its own rationing system through the chokepoint, reaching, as a principal, for exactly the thing Oman had been forbidden to touch as a conduit, and it met the same prohibition, enforced this time by fire.

The United States answered with what its own officials called punishment strikes, explicitly not proportional, more than eighty targets, and reimposed the oil sanctions it had waived. Iran then struck Kuwait and Bahrain, the aligned hosts of American bases. On 8 July, from the NATO summit in Ankara, Trump declared the memorandum over. A second wave of strikes followed that night, and a third by 9 July, as Washington moved to reimpose the naval blockade the memorandum had lifted.

The collapse states plainly what the ceasefire’s construction had implied. The proximity between Washington and Tehran was never jointly held, so only one party could end it, and did. As the strikes ran on, Qatar was still urging both sides to implement the memorandum, the conduit working to hold a line it had never been allowed to own, while the principal who could end it already had. The direct principal edge fell back almost to the war’s floor. What did not revert was the conduit structure. Oman was down and Pakistan up, those shifts holding through the collapse. The memorandum’s durable mark was the re-sorting of which conduit is licensed.

The same summit supplied one more face of the pattern. Trump spent the week berating the European allies over Iran and the strait, a rebuke that was not new. Back in March, when Europe declined his call to help force the strait open, he had called it a foolish mistake. Europe had been engaged on the defensive side throughout, and by early July France and Britain were deploying minehunters and standing up a multinational mission to keep Hormuz open. What it declined was offensive facilitation. Where Oman was punished for trying to own its adjacency, Europe was rebuked for withholding the part the hegemon most wanted spent, two faces of one demand. Adjacency in this system is to be spent on the hegemon’s account, neither owned nor withheld.

There is a limit to the pattern, and it is narrower than it looks. Israel spent the whole period acting on its own account and kept its freedom to do so. Washington could veto its most escalatory moves, instructing it in June to cancel a planned strike on Beirut and warning Netanyahu he would fight on his own if he resumed the war with Iran, but it could not stop the baseline campaign. Israel went on striking southern Lebanon and hit Iranian air defences and a petrochemical plant after being asked directly not to. A conduit that behaved a fraction as autonomously would have met the vassal clause within days. Israel met a phone call. The reason is that the ceiling governs conduits, and Israel is a principal. It is the coalition’s second belligerent, one of the parties that waged the war, and a principal owns its adjacency by right.

But owning your operations is not owning the outcome. Israel kept its military freedom and lost any say in the settlement. It was shut out of the Pakistan and Qatar mediated talks entirely, Netanyahu refused a meeting with Trump for weeks, and the memorandum, written without Israel, contained Iran’s missiles and proxies rather than eliminating them and left Hezbollah’s disarmament to a separate track. By July the apex ally was reduced to petitioning Washington for approval to finish a job the deal had left undone. On the grid, Washington’s weight over Israel holds at 0.7 while Israel’s reach back falls to 0.6. Being a principal buys operational latitude and nothing more. The proximity that ended the war ran through licensed intermediaries, and Israel, for all its adjacency, was not one.

The April map got two things right in advance. It held that the two dual-adjacent nodes would matter out of proportion to their size, and that Iran would press through Hormuz rather than through open battle. What it did not forecast is the pattern read backward out of the divergence between Pakistan and Oman. Power alone could not read this crisis, and adjacency alone cannot either. It is read through what each node is permitted to own, and on which side it owns it. Pakistan spent its adjacency as directed and rose. Oman tried to bank it and fell. Saudi Arabia held its in silence and drew no penalty. Europe withheld what was wanted and was rebuked. Israel owned its operations and kept them, and still got no share of an outcome the hegemon settled without it.