The Iran War Didn’t Start Because Diplomacy Failed. There Was No Diplomacy.
What a professional mediator sees in the Geneva talks that arms control analysts can document but cannot fully explain.
On February 26, 2026, two days before the United States and Israel launched Operation Epic Fury, Steve Witkoff and Jared Kushner sat across a table from Iranian Foreign Minister Abbas Araghchi in Geneva.
They were there to prevent a war.
They did not prevent it.
Last week, the Arms Control Association published an analysis by Kelsey Davenport, the organization’s director of nonproliferation policy, based on recordings and transcripts of Witkoff’s briefings with reporters on February 28 and March 3. The analysis is methodical, sourced, and devastating. You can read it in full at armscontrol.org.
What Davenport documents is a catalog of technical failures so fundamental that they reframe the question from why did diplomacy fail to whether there was any diplomacy at all.
Witkoff called Iran’s enrichment facilities industrial reactors. They are not reactors of any kind. He referred to Iran’s IR-6 centrifuge as probably the most advanced centrifuge in the world. It is not. He expressed surprise that Iran produces centrifuges, a capability Iran has possessed and documented for decades. He believed Iran had been actively weaponizing its nuclear program since 2003, a position directly contradicted by his own government’s intelligence assessments since 2007.
He concluded that Iran’s seven-to-eight-year supply of reactor fuel was evidence of weapons stockpiling. The IAEA had been transparently tracking that fuel supply in publicly available reports for years. The reactor in question was originally supplied by the United States and has been producing medical isotopes since 1967.
These are not minor technical gaps. They are the equivalent of someone pretending to be a surgeon who cannot identify the organ he is supposed to operate on, then walking into an operating room and making consequential decisions based on what he thinks he sees.
But here is what Davenport’s analysis, for all its precision, cannot fully explain.
The technical failures are symptoms. The underlying failure is something that arms control analysts are not trained to see, but that professional mediators recognize in the first minutes of a negotiation.
Witkoff didn’t just misunderstand the nuclear file.
He misunderstood what the negotiation was actually about.
There is a distinction that every professional mediator learns early and that virtually no diplomat or politician ever fully internalizes.
The difference between a position and an interest.
A position is what someone says they want. An interest is what they actually need.
Positions are the surface of every negotiation. Interests are the foundation. And the single most consistent failure mode of every failed negotiation I have witnessed in forty-nine years of practice is a negotiator who engages the positions while missing the interests entirely.
In Geneva, Iran’s position was clear. It had the right to enrich uranium. It would not abandon its enrichment program. Free fuel from the United States was an assault on its dignity.
Witkoff heard these statements as evidence of bad faith. As obstacles. As proof that Iran was not serious about diplomacy.
A trained mediator hears something completely different.
When Araghchi said that free fuel was an assault on our dignity, he was not being irrational or theatrical. He was stating Iran’s core interest with precision.
Iran’s interest was never the fuel.
Iran’s interest was recognition. Sovereignty. The acknowledgment that it is a legitimate state actor with rights under international law, specifically under Article IV of the Nuclear Nonproliferation Treaty, which grants all signatories the right to peaceful nuclear energy.
The moment Witkoff offered free fuel as a solution, he communicated, probably without intending to, that he understood Iran’s position to be about fuel acquisition. He had not heard the interest underneath the position. He had not heard what Iran was actually saying.
Iran heard his offer as confirmation that the United States was still treating it as a rogue regime to be managed rather than a sovereign state to be negotiated with.
That is why the offer was rejected as an assault on dignity.
Not because Iran didn’t want fuel. Because the offer demonstrated that Witkoff had not listened to what Iran was telling him.
And when you have not listened, no technical proposal you offer can address the actual problem. Because the actual problem is that the other party does not feel heard.
In my third book, Elusive Peace, I argue that the fundamental failure mode of international conflict resolution is the confusion between an agreement and a resolution. Diplomats optimize for the former. Professional mediators are trained to pursue the latter.
An agreement is a document that can be signed and announced. A resolution is an outcome that addresses underlying interests, has a self-enforcing mechanism, and does not require continuous external pressure to maintain.
The Geneva talks failed to produce either. But the reason they failed was not, as Witkoff claimed, that Iran was not negotiating in good faith.
Davenport’s analysis makes clear that Iran came to Geneva with a proposal. It was an opening offer, not a bottom line, which is how every serious negotiation begins. It showed flexibility on several key points, including Iran’s offer to blend down its stockpile of uranium enriched to 60 percent, a level near weapons-grade, to lower levels. IAEA Director-General Rafael Grossi said publicly that a deal was maybe possible.
The proposal was not acceptable as presented. Opening offers never are. That is what negotiation is for.
Witkoff dismissed the proposal because he believed the Tehran Research Reactor was a weapons stockpiling lie. It was not. He dismissed Iran’s sovereignty argument because he interpreted it as evidence that Iran only wanted to enrich, rather than as a statement of the interests that any serious agreement would need to address. He agreed to exclude ballistic missiles from the nuclear negotiations and then cited Iran’s failure to discuss missiles as evidence of bad faith.
He created conditions that made agreement impossible and then concluded that Iran was the reason.
This is not incompetence at the margins. This is the complete absence of the foundational skill every professional mediator possesses: the ability to hear what the other party is actually saying beneath what they say.
I want to say something here that most commentators on this crisis are not saying.
The failure in Geneva was not primarily Witkoff’s personal failure.
It was a structural failure. A predictable consequence of sending someone into the most consequential negotiation since the Cuban Missile Crisis without the training, the technical knowledge, or the process discipline that high-conflict dispute resolution requires.
This is not about a political party or ideology. Trained negotiators are not a partisan resource. The skills required to reach durable international agreements, the ability to distinguish positions from interests, opening offers from bottom lines, and agreements from resolutions are teachable, learnable, and have been documented in the professional literature of conflict resolution for decades.
They are simply not the skills governments look for when choosing who will represent them at the negotiating table.
They selected a real estate developer and the president’s son-in-law.
And soldiers died for it.
Here is what I wrote in my Substack article, The Reckoning, three weeks ago: the current negotiating team is optimizing for the announcement. The signals are everywhere.
The signals have only multiplied since.
The stumble-bumble agreement is still coming. The announcement will provide political cover for all parties. The market will rally on the day it is announced. The implementation will fail within weeks as the underlying interests that were never addressed reassert themselves. Israel, which was never bound by the framework the U.S. reached with Iran, will continue its own military operations. Iran will use continued Israeli strikes to justify resuming Strait restrictions. The oil price that partially normalizes following the announcement will settle back at $85 to $95, rather than at pre-war levels.
I wish I were wrong about this. The human cost of continued conflict is real and is being paid by people who had no voice in the decisions that created it.
But the stumble-bumble agreement will not end the conflict. It will pause it. And the pause will seed the next escalation, as every agreement that papers over unaddressed interests always does.
What would professional mediation have looked like?
Not the outcome. The process.
All three principals are at the table simultaneously. The United States, Iran, and Israel. Not bilateral U.S.-Iran talks that Israel can torpedo at will because it was never a party to them. All three. With the explicit acknowledgment that Netanyahu’s personal legal situation, his ongoing corruption proceedings suspended under Israeli law during active conflict, creates a conflict of interest that requires special attention in any process design.
A negotiating team with technical expertise in the nuclear file, regional geopolitics, and international law, supplemented by experienced process facilitators trained in the distinction between positions and interests.
An explicit agreement at the outset on the difference between an agreement and a resolution, with the goal defined as the latter rather than the former.
And the patient, sustained, unglamorous work of naming what each party actually needs underneath what it says it wants. Not in a single session. Over weeks or months of structured dialogue.
This is not idealism. It is the documented methodology of every durable international agreement in the post-war period, from the Camp David Accords to the Good Friday Agreement to the original JCPOA itself.
None of it is happening in the current process.
I am writing this two weeks before I deliver a TEDx talk on what I believe is the most important thing I have learned in forty-nine years of practicing law, mediation, and peacemaking.
The talk is not about Iran. It is about the most fundamental human capability that every broken system, from marriages to maximum security prisons to international negotiations, is failing to apply.
The capability to hear what another person actually feels underneath what they are saying. To name it accurately. From the inside of their experience.
When Araghchi said free fuel is an assault on our dignity, he was naming Iran’s emotional reality with precision. He was telling Witkoff exactly what Iran needed in order for any agreement to be possible.
Witkoff did not hear it.
Not because he is a bad person. Because he was never taught to listen at that level. Because the selection process that put him in that room selected for everything except the one skill that the room required.
The same selection process that put him in that room has been putting the wrong people in every room that matters for the past seventy years.
David, the CEO I wrote about last week, cannot understand why his best people keep leaving.
Witkoff cannot understand why Iran would reject free fuel.
The structure of the failure is identical.
Someone is telling you exactly what they need.
You are hearing the words.
You are missing the nervous system.
And the consequences, whether they are a resignation letter or a war, follow with the same inevitability.
The full analysis of the Geneva failure and what a genuine resolution would require is at armscontrol.org. Kelsey Davenport’s piece is essential reading for anyone trying to understand what actually happened in that room.
What she cannot tell you, because she is an arms control analyst and not a mediator, is that the failure she documents so precisely was not primarily a failure of technical knowledge.
It was a failure of listening.
And listening, unlike technical knowledge, is not acquired by reading more briefing books.
It is a skill. It is trainable. It has been proven to work in the hardest human environments.
Including, once, in a mediation room where I watched a woman name what her ex-husband felt and watched twenty-five years of contempt dissolve in three and a half hours.
Two words.
You feel.
The gap between what Witkoff did in Geneva and what those two words make possible is the gap between the world we have and the world that is available to us.
It is the gap I will spend 10 minutes trying to close on April 18th at the Apex TEDx event.
The full text of everything I have written on this, including the Reckoning and the pieces on affect labeling and nervous system leadership, is at dougnoll.substack.com.
Subscribe for free. The 14 Affect Labeling Scripts, the word-for-word framework for the situations where this matters most, are yours immediately.
If this article names something you have been trying to articulate, send it to one person who needs to read it.
Not broadly. One person.
That specificity is where everything begins.