◆ FIELDNOTES / POLITICS, TECHNOLOGY & CONSTITUTIONAL LAW №20 — JUNE 2026
A dispatch in six movements, with an epilogue — closing the second volume

The Doctrine of
Resumption

The war was declared over twice. It has now resumed twice. An innovation used once is an expedient. An innovation used twice is a doctrine; and the doctrine, once established, no longer requires the paperwork.

Published June 11, 2026
Reading time ~30 minutes
Filed under Politics / Technology / Law
Closes Volume Two (№11–№20)

At 5:15 in the afternoon, Eastern time, on June 10, the United States began striking Iran again. President Trump, speaking to reporters, said the United States would be "attacking them very hard," and explained the decision in a single temporal clause: Tehran had taken too long to reach a deal. Al Jazeera reported that the strikes, conducted jointly with Israel, included civilian railway lines; its footage showed rescue workers pulling people from the wreckage. The Middle East analyst James Dorsey, writing the next morning, titled his assessment with the only joke available: Trump and Iran agree on one thing; the definition of insanity.

The war that began on February 28 has now been declared over twice and has resumed twice. The first cycle was documented in the fourteenth dispatch in this series, The Terminated War: on May 1, the president sent letters to the Speaker of the House and the Senate's president pro tempore formally declaring that the hostilities had "terminated" as of April 7; on May 4, ninety-six hours later, American helicopters sank seven Iranian boats in the Strait of Hormuz under an operation called Project Freedom. The second cycle completed itself on June 10. After the May combat subsided, talks resumed; the twelve-nation Riyadh statement of May 17 demanded Iranian compliance; negotiations proceeded through late May and early June; and then the president decided the negotiations were proceeding too slowly, and the bombing resumed.

Note what was absent from the second cycle: a letter. The first termination required formal correspondence with Congress, a legal theory, an anonymous administration official previewing the argument to reporters. The second resumption required a podium. No new notification has been reported; no new legal framework has been articulated; no congressional vote has been scheduled, demanded, or seriously discussed. The hostilities remain, in the administrative record, terminated. The bombs fall anyway.

2 Complete termination-and-resumption
cycles in six weeks
60yrs Maximum sentence facing protesters
convicted under NSPM-7
20 Dispatches since this pattern
was first named

This is the twentieth dispatch in this series, and it has two jobs. The first is the usual one: to read the week's events structurally rather than episodically, and to name what the second resumption confirms that the first could not. The second job is one no previous dispatch has carried: this dispatch closes the second set of ten, the volume that began with The Dual Blockade in late April, and so its final movements look back across what the twenty dispatches have accumulated; not as a victory lap, but because the accumulated argument is now itself a finding, and findings belong on the record.

Chapter I

The Second Resumption

The Terminated War made a specific claim: that the May 1 letter was a constitutional innovation, categorically different from the previous half-century of War Powers evasions. Reagan in Lebanon, Clinton in Kosovo, and Obama in Libya all argued that their ongoing operations did not constitute "hostilities" within the meaning of the statute. The Trump administration's May 1 maneuver accepted that the operations were hostilities, declared them terminated as of a retroactively selected date, and reclassified everything that followed as something else; defensive escorts, maritime interdiction, humanitarian convoy protection. The dispatch predicted that the innovation would outlast the war and that whoever inherited the executive branch would inherit the mechanism.

The prediction was correct but insufficiently ambitious. The mechanism did not wait for the next administration. It matured within six weeks, inside the same administration, in the same war; and its maturation is visible precisely in what the second cycle did not require.

Consider the asymmetry. The first termination was a production: letters with statutory citations, a legal theory previewed to reporters on background, a defense secretary testifying that the 60-day clock could "pause or stop" during a ceasefire, Democratic senators objecting that there is no pause button in the Constitution. The administration treated the maneuver as something that needed to be defended, which is to say, it treated the constraint as something that still existed and had to be answered. The second resumption was a remark to the press pool. Talks were taking too long; the attacks would be very hard. No theory. No letter. No defense. The constraint was not answered because, after the first cycle, there was no longer anything to answer. The termination of April 7 remains on file. As long as it remains on file, every subsequent military operation exists in the post-hostilities period, and the clock that would govern hostilities has nothing to attach to.

This is what doctrinal maturation looks like in administrative practice. The first use of a procedural innovation carries risk; the actor does not yet know whether the institution being circumvented will resist. The institution's failure to resist is information. The second use incorporates that information: it dispenses with the justificatory apparatus because the first cycle demonstrated the apparatus was never load-bearing. The letter of May 1 was scaffolding. The building stands without it now.

An innovation used once is an expedient; its author can still claim necessity, singularity, exception. An innovation used twice is a doctrine; repetition is the announcement that the exception has become the rule, and that the rule's former enforcers have been measured and found absent. — The arithmetic of precedent

The series will state the obvious prediction plainly, because falsifiable predictions are the publication's discipline: there will be a third cycle. The frozen conflict documented in the twelfth dispatch has not thawed; the structural conditions (the blockade, the toll system, the strait, the regime in Tehran that survived and is consolidating) remain in place; and the doctrine now makes resumption procedurally costless. The only question is the date.

Chapter II

The Domestic Mirror

On June 23, in Prairieland, Texas, a group of anti-ICE activists will be sentenced. They were convicted of providing material support to terrorists. The conduct underlying the conviction, as reported by The Intercept, included dressing in black to attend a protest. They face up to sixty years in prison. They are the first people convicted under the enforcement framework created by National Security Presidential Memorandum 7, the directive that designated certain categories of protest activity as terrorism-adjacent.

The Terminated War identified the procedural-redefinition mechanism in the military domain: a category that triggers a constitutional constraint (hostilities, which engage the War Powers Resolution) is administratively reclassified into a category that does not (post-hostilities defensive operations). The Prairieland convictions are the same mechanism in the domestic domain, and the mapping is exact. The First Amendment protects protest. It does not protect material support for terrorism. The constraint is therefore never confronted; it is rendered inapplicable by the classification of the conduct, and the classification is controlled by the same branch of government whose conduct the constraint was designed to limit.

The temporal structure deserves attention, because it is where the constitutional damage actually accrues. The prosecution operates at the speed of an executive memorandum and a motivated U.S. Attorney's office: months from arrest to conviction. The constitutional adjudication that might eventually find the prosecutions infirm operates at the speed of the appellate system: years, possibly the better part of a decade if the question reaches the Supreme Court. In the interval, the deterrent effect is total and immediate. A person deciding whether to attend a protest this summer does not consult the eventual appellate outcome; they consult the sixty-year exposure of the people sentenced on June 23. The punishment is the process, and the process runs at executive speed while the remedy runs at judicial speed. This is the temporal-mismatch pattern operating inside the constitutional order itself: the constraint exists, the constraint may even ultimately prevail, and the constraint is nonetheless inoperative during the entire period in which it matters.

Read together, the second resumption and the first NSPM-7 sentences describe a single constitutional condition. Abroad, the category of "war" has been dissolved so that the constraint attached to war cannot engage. At home, the category of "protest" is being dissolved so that the protection attached to protest cannot engage. The mechanism is identical; only the direction of the reclassification differs. War is defined down, out of its constraining category. Protest is defined up, into a criminal one. In both cases, the classification authority and the constrained actor are the same, and the institutions that nominally own the categories (Congress, the courts) operate too slowly to repossess them.

Chapter III

The Dromologist's Clock

In 1977, the French theorist Paul Virilio published a short, strange, prophetic book called Speed and Politics, in which he proposed that the discipline missing from political analysis was what he called dromology; from the Greek dromos, the race, the running. Virilio's claim was that the history of societies is not primarily the history of wealth or of ideas but the history of acceleration: that every revolution in transportation, communication, and weaponry redistributes power toward whoever masters the new tempo, and that the political form we call deliberation (parliaments, courts, treaties, the entire apparatus of considered collective decision) is itself a technology; a slow one, built for a world in which events moved at the speed of horses and sail.

Virilio's central aphorism was that speed is the essence of war, and his darkest forecast was that as the technologies of instantaneity advanced, the time available for human decision would compress toward zero, and politics would be replaced by the management of velocity; the state of emergency would become the permanent state, because emergency is simply the name deliberative systems give to tempos they cannot match.

This publication has been writing dromology for four months without using the word. The temporal-mismatch thesis, stated in Virilio's terms, is this: institutions fail when they cannot match the tempo of the actors and events they are designed to constrain, and the tempo gap is not an accident of circumstance but a structural feature, because the constrained actors are continuously incentivized to accelerate while the constraining institutions were built, deliberately, to be slow. Deliberation is slowness on purpose. Its designers understood slowness as a safeguard. They did not anticipate adversaries for whom the safeguard's tempo would itself become the vulnerability to exploit.

The June 10 resumption is the purest dromological event the series has recorded, because the stated casus belli was tempo itself. The president did not allege a new Iranian provocation, a violated term, a discovered weapons program. He alleged slowness. The talks were taking too long; the offense was the rate of diplomatic progress; the remedy was kinetic. A negotiation is a deliberative technology, and the president's complaint was, precisely, that it deliberates. When the velocity of concession falls below the threshold of executive patience, the executive switches instruments to one that operates at his preferred speed. Virilio would have recognized the move instantly; he spent his career warning that the actor who controls the tempo of events will eventually refuse to tolerate any process slower than himself.

◆ The integral accident

Virilio paired every technology with the accident it invents: the ship invents the shipwreck, the railway invents the derailment, the airplane invents the crash. His late-career anxiety was what he called the integral accident; the accident of systems so synchronized and so global that their failure occurs everywhere at once. The Defense One brief in this week's reporting noted unexplained GPS outages across multiple regions; positioning and timing infrastructure degrading from electronic warfare spillover, deliberate interference, or causes not yet attributed. GPS is the synchronization layer of the global economy; shipping, aviation, financial timestamping, power-grid coordination all depend on it. A dromological reading notes the symmetry without overstating it: the war over a chokepoint in physical space is generating interference in the infrastructure of time itself. The series adds it to the watch list.

Chapter IV

The Gated Frontier

The same week the war resumed, the AI frontier reorganized itself, quietly, into a two-tier architecture; and the reorganization belongs in this dispatch because it is the same structural story told in a different domain.

Anthropic released its newest frontier system this week in two forms: a generally available model carrying additional safety measures for dual-use capabilities, and an unrestricted version, called Mythos, available only to approved organizations. The trade press treated the release as a product story. Defense One asked the only question that matters structurally: who gets Mythos? Because the moment frontier capability is gated, the gate becomes a governance instrument, and access becomes a currency of state power. The question of who is approved, by whom, under what criteria, reviewable by which institution, is not a product question. It is a constitutional question wearing a product launch's clothing.

The historical precedents are well-worn and worth invoking precisely because their arcs are known. Strong cryptography was classified as a munition and export-controlled until the late 1990s; the gate leaked, the technology diffused, and the control regime collapsed into irrelevance within a decade of its peak enforcement. High-resolution satellite imagery was a state monopoly until commercial operators ended it. Nuclear materials remain the one durable gate, and they are durable for reasons (detectable physical infrastructure, controllable feedstocks) that information technologies conspicuously lack. The dual-use lifecycle runs: gate, leak, race, normalize. The gates buy time. They have never, for an information technology, bought permanence.

The gated frontier is therefore a bet about time, and the bet sits on top of two structures this series has already mapped. The first is the financial architecture: Oracle reported its fiscal fourth quarter on June 10, and the headline pairing (record AI-cloud backlog, genuinely enormous cash burn) is the Self-Strengthening Trap's concentration problem restated in a quarterly filing; the commitments are real, the revenue is promised, and the gap between them is being financed at scale. The second is the physical chokepoint: the November 10 expiration of China's rare-earth export-control suspension is now five months away, and nothing in the Beijing summit's afterglow has resolved it. A gated frontier whose gates are administered by one country, whose chips depend on materials refined at 91 percent by a rival country, and whose financing depends on revenue projections that the rival's own laboratories are racing to undercut, is not a stable architecture. It is a tempo bet; the bet that the gate-holders can convert their lead into permanence faster than the excluded can build around the gate. Virilio again: the race is the politics.

Chapter V

To the Reader

This publication has never had many readers. That is worth saying plainly, in the publication's own pages, because pretending otherwise would violate the only editorial rule that has governed every dispatch: say what is true, including about ourselves. Fieldnotes publishes essays that run twenty-five minutes, in a register closer to a quarterly journal than a feed, under titles that withhold their subjects, about a pattern that takes several dispatches to see. Every one of those choices narrows the audience. They were made anyway, because the alternative (faster, shorter, louder) is the tempo problem this series exists to describe, and a publication about the costs of acceleration that accelerated itself to be heard would be its own counterexample.

So this chapter is addressed to you; the small company who have been here since the spring, and the reader who arrives later, holding this as a chapter in a book, knowing how some of these open questions resolved. You two are the same audience at different points on the clock, and you have watched the same thing accumulate.

What you watched was twenty attempts to read the present structurally. You watched a dispatch about munitions production curves argue, eight days into a war, that duration would be the entire war; and you watched duration become the entire war. You watched a dispatch about the Bronze Age insist that the Sea Peoples were the wrong villain for the right story, and then spend the following months finding the right story everywhere: in a strait that would not open, in a fleet of ghost tankers, in a paralegal's severance letter, in six Iranian boats burning on the water three days after the war was declared over, in 6.1 million clerical workers whose adaptive capacity was measured and found wanting at the precise moment their support systems were defunded, in a Qing admiral's modern fleet and an American chipmaker's concentration index, in seventeen elements refined almost entirely in one country, in a sparse autoencoder finding that the saddest thing a machine does is describe its own condition.

Twenty dispatches, read as one argument, say this: complex systems do not fail because they are attacked. They fail because the mechanisms built to protect them run slower than the things they need protection from, and the speed of the protective mechanisms is a design choice; one made years or decades or centuries before the threat arrives, by designers who assumed the world would keep their tempo. Six variations of the failure have been catalogued in these pages: the institution that is too slow, the institution that is dismantled, the epistemic ground that erodes, the person who lives in the gap, the system that transforms continuously beneath its occupants, and now, twice confirmed, the constraint that is dissolved by reclassification faster than its owners can object. The seventh variation is forming in plain sight, in the gated frontier; a protective architecture whose own designers know it is a bet against the clock.

None of this was prophecy, and the reader holding the book knows it better than the reader of the website, because the book reader knows which of the timestamped predictions failed. Some will have. The publication's commitment was never to be right; it was to be checkable. Every dispatch was a structured guess, dated, falsifiable, and held loosely; and the value of the series, whatever it turns out to be, lives in that discipline rather than in any individual call. If the framework survives contact with the events you already know about and we did not, it will have earned its keep. If it did not survive, the record of exactly how it failed is itself the useful document. Either way: thank you for doing the unusual work of reading slowly about speed.

Chapter VI

The Watches

The series carries three load-bearing watches forward out of this volume, stated here as falsifiable expectations so that the record can judge them.

The third cycle. The Doctrine of Resumption predicts its own repetition. The structural conditions of the frozen conflict persist; the procedural cost of resumption is now zero; therefore a third termination-and-resumption cycle should be expected, and its arrival will confirm the doctrine while its absence over a long horizon (a genuine, durable settlement) would be evidence against this dispatch's central claim. The series will log the date, whichever way it goes.

November 10. The suspension of China's rare-earth export controls on semiconductor inputs expires five months from now. Re-suspension buys another year; reimposition activates case-by-case Chinese review over the supply chain of every advanced chipmaker on earth; genuine resolution remains the least likely outcome. The spot prices of neodymium and dysprosium oxide remain the most honest real-time forecast available, more honest than any summit communiqué. The series will track them into the fall.

The gate's first test. The gated-frontier architecture will at some point encounter a state actor fielding equivalent capability without gates. The Chinese AI financing structures mapped in the Collapse Signature project's monitoring streams are the most likely origin. When the test arrives, the gate-holders will face the choice every dual-use control regime has faced: adapt the architecture or watch it become ceremonial. The historical base rate is not encouraging for the gates. The series will be watching for the first credible demonstration.

Behind the three watches, the running instruments stay on the board: inflation printed at 4.2 percent in May, the highest in two years, which is the frozen war's monthly invoice to every household; the Oracle cash-burn trajectory, which is the AI buildout's quarterly confession; and the Collapse Signature dataset, which grows by a case brief at a time toward the second book after this one; the methodical one, the one written at the speed of the analysis rather than the speed of events, targeted for 2028.

The war was declared over twice. It resumed twice. The first time required a letter; the second required a podium. That difference; the vanishing of the paperwork; is the whole finding. The constraint did not lose an argument. It stopped being argued with.

◆ End of dispatch — end of Volume Two

Epilogue — The Doctrine, Stated

Every volume should end by saying its thesis in the fewest possible words, so that the reader can carry it out the door. Here is this one's.

Any constraint that can be dissolved by reclassification will eventually be dissolved by reclassification, at the speed at which the constrained actor discovers that the reclassification is available, repeatable, and unanswered. The dissolution does not announce itself as crisis. It announces itself as procedure; a letter, then a podium, then nothing at all, because by the third iteration the category has simply ceased to exist as a constraint and exists only as a word. The War Powers Resolution is the cleanest case this volume documented. The First Amendment's protest protections are the case forming now. The gated frontier is the case being constructed for the next volume to document.

Against this, the only durable countermeasure the historical record offers is not faster institutions; institutions cannot outrun the actors they constrain, and the ones that try become the danger themselves. The countermeasure is custodianship of categories: the unglamorous, continuous work of insisting that words like war, protest, hostilities, emergency, and termination retain their meanings, because every constraint in a constitutional order is, in the end, a category, and the order survives exactly as long as its categories do. That work runs at the speed of writing and reading; which is to say, at the speed of publications like this one, and of readers like you.

The dispatches will continue. The freeze will deepen or it will break; the deadline in November will arrive; the third cycle will come or, better, it will not. Whatever happens will happen first as an event and then, here, as a structure. The clock is still running. So is the record.

← Return to the archive