On 27 May, asked at a Cabinet meeting whether he would accept a short-term arrangement under which Iran and Oman would share control of the Strait of Hormuz, Donald Trump declined and added a sentence that has since been read as a slip. Oman, he said, would behave just like everybody else, or the United States would have to blow them up. The White House was asked to clarify whether the President had meant Iran, the state he has threatened for months, rather than a country Washington has counted among its partners in the Gulf for half a century. The question assumes the line was an error. It is more useful to assume it was not.
To understand why the line is not an error, one has to understand what the Strait has become over the course of this war. The status of the waterway is the central unresolved knot of the US–Iran endgame. For weeks Washington’s position has been absolute. The Strait must be open, free, and clear, and any attempt to close it, condition it, or charge for it has been met with threats to bomb Iran back to the Stone Ages. What is at issue is not merely passage but the principle that passage is uncharged, that the chokepoint cannot be converted into a revenue instrument by anyone.
On 20 May, Iran’s newly created Persian Gulf Strait Authority published a graphic depicting a “controlled maritime zone” spanning the Strait, even as Tehran’s foreign ministry insisted it was not seeking to collect tolls. A reported draft agreement would have Iran retain management duties over the waterway, sharing those duties with Oman. The Trump administration ruled the arrangement out entirely, with senior officials stating there was no scenario under which Iran would impose tolls as part of any deal. Trump’s own formulation at the Cabinet meeting was sharper still. Nobody, he said, was going to control the Strait. The United States would watch over it, but nobody would control it.
This was not said from a position of detached diplomacy. It was said by a power already enforcing that principle by fire, in the same days as fresh US strikes on Iranian positions near the waterway, framed as defensive measures to hold a fragile ceasefire. The threat to Oman is issued from inside an active military campaign to keep the Strait open and uncharged, which is what makes it less a rhetorical lapse than an extension of a force already in use. It is against that line that the warning to Oman has to be read, and against that line it is far less of a slip than it appears.
A threat to destroy an ally is not, on its face, coherent. Allies are not the objects of annihilation threats. Adversaries are. When the grammar of a sentence assigns to a partner the treatment reserved for an enemy, the conventional response is to treat the sentence as a malfunction. But a sentence can be malfunctioning at the level of diplomacy and perfectly accurate at the level of structure. The line exposes no confusion about who Oman is. It discloses what Oman has been all along, under a description Washington does not say out loud and has had no occasion to say until now.
The description is the difference between an ally and a vassal. The two are easy to conflate because, for long stretches, they are observationally identical. An ally is a sovereign whose interests you coordinate with. A vassal is a node whose behaviour you license. Vassalage in this sense is not total dependency, and the Gulf monarchies clearly possess substantial bargaining capacity, financial weight, and room to hedge. It denotes a conditional sovereignty within a security architecture, where considerable agency is permitted in any direction except those the architecture forecloses.
As long as the sovereign’s interest and the suzerain’s interest point the same way, the two categories generate exactly the same conduct, and nothing in the behaviour alone reveals which one is actually operative. Oman kept the Strait open, carried messages between Washington and Tehran, hosted the quiet channels, and asked for little in return. That conduct is consistent with alliance. It is equally consistent with vassalage. The behaviour does not discriminate between the two readings. Only divergence does.
The co-management proposal is the divergence. The relevant act is not Oman pricing a waterway along its own coast in isolation. It is Oman agreeing to share administrative control of the chokepoint with Iran, inside a structure Tehran has built and named. The reporting on the draft is provisional. The administration itself has called the underlying account a fabrication, and the arrangement may yet collapse, contract, or never have been seriously entertained in Muscat. None of which weakens the inference. If anything it sharpens it. The destruction warning was issued in response to a narrative possibility the administration says is not real, which means the load-bearing fact is not the proposal itself but Washington’s reaction to its mere imagining, a reaction that treats even the prospect of Omani co-management as ground for a threat of force.
The aftermath confirmed the grammar. Within a day the Treasury Secretary warned that Washington would aggressively target anyone facilitating a toll, and Oman used its ambassador to disavow the plan and reaffirm two centuries of good relations, hurrying to show it had not meant to misbehave. That is the assertion of a sovereign prerogative in the plainest sense. A state that agrees to co-administer a strait, to hold a say over the passage of a fifth of the world’s seaborne oil, is reaching for the kind of authority sovereigns reach for. It is treating its geography as a position it may dispose of on its own terms, including in partnership with whomever it chooses. Whether it is entitled to do so is a separate question, and a contested one. It is acting as though it were, and that is what matters. And it is precisely at the moment Oman moves toward this that the relationship reveals its real category. The move is not tolerated as the exercise of a partner’s prerogative. It is treated as misbehaviour to be corrected, on pain of destruction. That response is unintelligible if Oman is an ally. It is exactly correct if Oman is a vassal. The vassal is the ally whose sovereignty was only ever conditional on never being exercised.
To see why the co-management arrangement in particular triggers this, it helps to recall what made Oman valuable in the first place, and here the contrast with the rest of the Gulf is the whole point. In a structural mapping of the Middle East security complex I set out on Modern Diplomacy last April, I assigned each significant relationship in the system an operational weight and found that the Gulf partners cluster in a single pattern. They are adjacent to one pole. Saudi Arabia, the United Arab Emirates, Bahrain, Qatar, and Kuwait are all connected, at varying weights, to the American coalition and effectively disconnected from Iran. Their value to Washington is their alignment, and alignment is by definition one-directional. Whatever friction runs through any of these ties, none of them holds an operational relationship with Tehran of its own. Oman is the single exception. It holds meaningful adjacency to the American coalition and to Iran at the same time, a dual connectivity no other state in the matrix replicates. That is its anomaly, and the anomaly is the source of its worth.
The reason is straightforward once the pattern is made visible. A node connected to both poles can carry what neither pole can carry directly, because it is the only point in the system with a live channel to each. An aligned state may broker at the margins, but it cannot mediate across the central divide on its own account, because it has no standing relationship with the pole it is structurally set against. Its single-pole adjacency makes it useful for coordination and limited for contact across the main fault line. Oman’s two-sidedness is what allowed the back-channel to run through Muscat in the early phases of the conflict, as it allowed the secret nuclear contacts Oman hosted more than a decade earlier. That role was not a favour Oman did the system. It was the functional expression of a structural position no aligned partner could occupy. Oman is valuable for the very reason it is unlike the others. It touches the enemy.
But dual adjacency is a double-edged asset, and the co-management proposal reveals which edge Washington was counting on. A node with a genuine tie to the adversary is, by definition, not a pure conduit. It has a relationship of its own to the other side, and a relationship of one’s own is the raw material of independent action. Washington’s tolerance of Oman’s Iran tie is conditional on that tie remaining in service. A channel Washington can use, not an asset Oman can capitalise. The back-channel role is acceptable because it spends Oman’s adjacency on the coalition’s behalf. Co-administration does something categorically different. It attempts to convert that same adjacency into a standing institutional position, to capitalise the structural relationship on Oman’s own account rather than expend it in the suzerain’s, and to do so by sitting down as co-manager with the very state the guarantee is structured against. The dual adjacency that made Oman useful as a servant is the very thing that makes the arrangement legible as insubordination. Oman is not defecting to Iran. It is doing something the vassal framework finds more intolerable than defection. It is behaving as the owner of its position rather than its custodian. The aligned states cannot commit this particular offence, because they have no second adjacency to monetise. Oman can, and that is exactly why the threat falls on Oman and not on them.
This is what the word “behave” is doing in the sentence. One does not tell a peer to behave. The verb belongs to the register of a superior addressing a subordinate whose conduct is subject to correction. It presupposes a standard of acceptable behaviour that the superior sets and the subordinate is expected to meet.
Diplomacy among sovereigns does not run on that verb. It runs on interest, reciprocity, and the management of divergence. The appearance of “behave” in a sentence about an ally is the tell, and Trump’s accompanying formula only sharpens it. The Strait will be watched over, but nobody will control it. To watch over a thing is the posture of a custodian who permits use without conceding ownership. That is the relationship Washington is asserting not only over the waterway but over the states arrayed around it.
What the sentence articulates, without meaning to, is the operative definition of alliance the guarantee has always run on. An ally, under that definition, is a party free to act in many directions but never to shadow, undermine, or reprice American power projection, will, or political objective. An ally so bounded is a vassal under another name, and the definition was in force long before it was spoken aloud. Stepping beyond the permitted range is not a policy difference but a disciplinary matter, and the threat that follows, destruction, is not disproportionate once the relationship is named correctly. It is the sanction the structure reserves for a node that stops conducting and starts owning.
None of this requires the United States to have decided, at some identifiable moment, to demote Oman from ally to vassal. One could read the episode as degradation, alliance hardening into suzerainty under the war’s pressure. But that reading needs an event, a moment of conversion, and there is none to point to. What the divergence reveals is more economically explained as a structure that was always present and merely untested. There was no demotion because there was no prior moment of genuine alliance to fall from. The relationship was suzerainty in an ally’s vocabulary from the start.
What changed in May was not Oman’s status but its visibility. For decades the vassal arrangement was indistinguishable from partnership because Oman’s sovereign interest and Washington’s never seriously diverged, and a category that is never tested looks like whatever the conduct of the moment suggests. The co-management proposal is the first real test, and the test returns the answer the structure always contained. This is the more unsettling reading precisely because it does not depend on any actor’s intention. It is a fact about position, not about will. The slip did not change anything. It made something already true available to be seen.
The implication runs well beyond Oman, and Washington has stated it almost in so many words. As the contours of a settlement have taken shape, Trump has made clear that any deal with Iran should be accompanied by Gulf and Muslim states joining the Abraham Accords, and that these states, Saudi Arabia, the United Arab Emirates, Qatar, Bahrain, Pakistan, Turkey, Egypt, and Jordan among them, owe their accession to the United States.
Owe it to us is the vassal clause spoken aloud, addressed this time not to the anomalous dual-adjacent node but to the single-pole states whose alignment has so far made the clause invisible. The clause appears here not as a sanction for crossing the gate, as with Oman, but as a standing expectation under alignment, the same hierarchy showing its other face. The adjacency that gives each of them value to Washington is licensed, not owned. Useful as service, owed as tribute, forbidden as sovereignty.
Oman discovered the clause written into all of these arrangements in invisible ink. It does not activate under ordinary alignment, where the two categories remain indistinguishable, and it does not collapse every asymmetric alliance into vassalage by definition. It surfaces only at the point where a partner attempts to reposition the node itself, to convert the geography that lent it value into an asset disposed of on its own account. That is the gate the Oman case crossed, and the gate the others have not yet been asked to approach. Those who read the Oman threat as an anomaly, a slip, a sentence the White House will walk back, are looking at the standard terms of their own guarantee, recited out loud for once. The question each of them now faces is not whether Washington meant it about Oman. It is what their own position is worth the moment they try to price it themselves.
