The number in every account of this negotiation is two, and it has been wrong since the first week of March.
Today Iran sent a ceasefire proposal to the United States through Pakistani mediators. That sentence contains three governments, and the arrangement it describes contains at least two more: Israel, which has been conducting the strikes on Iran and is not a party to the talks, and the Gulf states, which have absorbed the retaliation for those strikes and are not a party either.
A bilateral framing makes the process legible and it misdescribes what is actually being attempted. What follows is a comparison with the last time this particular pair of governments negotiated through an intermediary, because the differences are the reason this is harder.
The channel that worked
The template everyone has in mind is Oman.
Beginning around 2011 and running through 2013, Oman hosted a series of contacts between American and Iranian officials that neither government acknowledged at the time. Those contacts established that a negotiation was possible at all, and they produced the framework that became the multilateral nuclear agreement.
Four features made Oman work as a channel, and it is worth listing them because they are a checklist.
It was trusted by both sides and threatening to neither. It was small enough to have no independent ambition in the outcome. Its involvement was deniable, which meant either side could walk away without having publicly failed. And the subject was a single technical question, nuclear enrichment, on which the terms of a deal could be written down as numbers.
The channel that exists now
Pakistan is doing something that resembles this and differs on three of the four.
Trust is the one it has. Pakistan maintains working relationships with Tehran and with Washington, and its military and intelligence leadership have standing with both, which is not a common combination and is the reason it is in the room.
The rest do not hold.
Pakistan is not a small state without interests here. It shares a long border with Iran, it has its own insurgency along that border, and it has a substantial relationship with the Gulf states whose infrastructure is being struck. It has reasons of its own to want this stopped, which is useful for energy and unhelpful for neutrality, because a mediator with preferences is a participant.
Nor is the process deniable. These contacts are being reported as they happen, by both capitals, which means each round has a public result. A failed meeting is a failure somebody has to explain at home, and that raises the cost of every meeting and lowers the willingness to hold one.
And the subject is not a single technical question. A nuclear agreement could be reduced to centrifuge counts and enrichment levels and inspection schedules: unpleasant to negotiate, but expressible. This negotiation has to address a naval blockade, a war being prosecuted by a third party that is not at the table, the position of the Gulf states, and a nuclear question underneath all of it. There is no single number that resolves any of it.
What the comparison teaches
Two things, and the second is the uncomfortable one.
The first is that mediation is not a substitute for a negotiation between the parties that matter. Oman did not settle the nuclear question; it created a channel through which the parties settled it themselves, and the settlement required direct contact eventually. A mediated process that never converts into a direct one is a mechanism for passing messages, which is valuable and is not the same as progress.
The second is about who is missing. Israel is conducting the military campaign and is not in these talks. Whatever Tehran and Washington agree, the strikes continue or stop according to a decision taken in a third capital, and a ceasefire that one belligerent has not agreed to is not a ceasefire; it is an American undertaking to try to obtain one.
That is the structural weakness in the entire arrangement, it has been present since the start, and it is why the framing of two parties keeps producing expectations that then go unmet.
Why the missing party stays missing
There is a reason Israel is not in these talks and it is not an oversight, so it is worth setting out rather than treating as a puzzle.
Iran does not recognize Israel and will not negotiate with it directly, which removes the simplest arrangement immediately. That is a position of long standing and it has survived considerably more pressure than this.
Nor would a mediated channel solve it, because a mediator can carry messages between parties who decline to meet only if both accept the mediator. Pakistan’s standing with Tehran rests partly on positions that make it unsuitable as a conduit to Jerusalem.
So the structure that has emerged is the only one available: the United States negotiates with Iran, and separately undertakes to deliver Israeli compliance with whatever it agrees. That is an arrangement in which one party is guaranteeing the conduct of another party over which it has influence rather than authority.
Every previous attempt at this shape, in this region, has turned on the same question, which is how much influence that actually is. Nobody outside a very small number of rooms knows the answer, and the answer has been different at different moments.
What the Gulf states are doing about it
Worth noting, because it is the part of this that has actually moved.
The states absorbing the retaliation have spent two months building their own channels: Oman on maritime transit, Qatar and others on prisoner and shipping questions, regional foreign ministers meeting as a group. None of that is a peace process and all of it is the construction of a set of relationships that could carry one.
If this ends through negotiation rather than exhaustion, the machinery will probably come from there, because it is the only part of the region currently building machinery.
What is unresolved
The proposal sent today will be read in Washington, and a response will be sent back through Islamabad, and that exchange will be reported as a round of talks.
What nobody can say is whether it is a negotiation or a correspondence. The difference is whether either side is prepared to give something up, and a mediated channel is specifically designed so that neither has to demonstrate that in public until very late.
Which is its virtue, and the reason it is attractive, and the reason a process can run through many rounds while the distance between the positions does not change at all. Oman’s channel took two years before it produced anything anyone could point to, and it was working the entire time.
