Section XXIX: XXIXThe Question Mark

Engels, in the formulation Section XXI named, later compressed by Luxemburg into “socialism or barbarism,” specified that the alternative to a successful socialist transition is regression into barbarism. Both outcomes remain available. Preserved capitalism in its current form is what the tests below would have to show, and this work’s claim is that they will not. The method that dated 1848 and 1899 too early has to be turned on this work as well, which is why its claims are written as the tests below, with dates: if the room is not spent, the tests will say so before the argument does.

Arrighi closed the introduction of The Long Twentieth Century on the same exhaustive partition from the other tradition. His question, paraphrased, was whether the East Asian rise was a new kind of superiority of force or the beginning of the end of the superiority on which the West’s fortunes had rested for five centuries, and whether capitalist history would end through a world empire founded on the West’s enduring force, as Weber seemed to expect, or through a world market economy in which that force withered, as Smith did.³⁶⁴ The Weber ending is the one Washington attempted and this work has documented failing: the de facto monopoly of legitimate violence assembled after 1991, spent down from Indochina’s precedent to a chokepoint where the underwriters now perform the interdiction the fleet cannot.

The Smith ending, force equalized by mutual fear until the injustice of nations is overawed into respect, is recognizable as the multipolar deterrence structure the nuclear solvent enforces and the reserve asymmetry administers. Smith’s equalizer and Engels’s fork are the same observation made a century apart. The architecture ends. The character of what follows is the only variable.

The partition has a second statement at the other end of the same book, and this work adapts it. The epilogue lays out three endings rather than two: a world empire imposed by the old centers’ still-superior force. A world market society centered on East Asia in which Western force withers into one power among several, or systemic chaos, humanity consumed in the escalating violence of the transition itself. His sentence for the third, paraphrased, is that before humanity chokes or basks in the dungeon or paradise of a Western-centered empire or an East Asian-centered market society, it may burn in the escalating violence that has accompanied the liquidation of the Cold War order.³⁶⁵ Against Section XXI, the world empire is the ending Washington attempted and Hormuz foreclosed. Systemic chaos is Branch 3. The world market society is the terrain the rest of the scheme divides on the two axes Arrighi left open, whether the imperial core reorganizes toward a Marxist politics or an Orbanized consolidation, which Branches 1 and 2 separate, and whether the successor stays capitalist, which the recomposition case contests. The branches are his partition adapted, with the one question he left open, the class character of what follows, specified as the thing the falsification structure adjudicates.

In the third week of September the incumbent’s own press listed the four in one paragraph, war warnings from the European capitals, the alarm over AI, the energy crunch and the bond market’s signal of a debt crisis, as a fraught moment wanting far-sighted men. This work has read them as one event under four names, and the men it wants are the ones the remedies above have shown no incumbent can supply.³⁶⁶ Lenin’s pamphlet of September 1917 opened on the same scene: everybody says this, everybody admits it, everybody has decided it is so, and nothing is being done.³⁶⁷ The existing American imperial architecture is in clear liquidation. The AI fictitious capital expansion has no validation in sight, and the listing test below is where one would appear. The petrodollar enforcement at Hormuz is materially failing. The political articulation contest in the imperial core is being conducted at social media speed below the level of effective institutional intervention or mediation. The Communist Party of China has absorbed 150 years of revolutionary practice and counter-revolutionary failure as operational doctrine and is conducting the strategic stalemate transition into counter-offensive with limits that the prior phases of On Protracted War were designed to enable.

The substrate is no longer where the imperial framework’s institutional memory says it is. The recovery mechanisms of prior crises were available because the productive center sat inside the architecture being rescued. The productive center now sits outside it. Recovery on the prior pattern requires reabsorbing what has migrated, on a timeline the cascade does not provide. The form of the supersession is the only live question.³⁶⁸

Which branch the rupture phase resolves into is not yet determined. Marx refused to answer the political articulation question in his published work, and the material conditions don’t answer it either. It depends on the work done within the rupture phase, on the frameworks pre-positioned to absorb the recognition, on the organizational and intellectual infrastructures that exist when the air pocket closes, the lag between the structural break and the moment the population feels it. Beijing’s institutional capacity to absorb the transition is one variable. The articulation contest in the imperial core is another. The two are coupled. The form the imperial-core configuration takes during the rupture phase determines part of what Beijing inherits, and what Beijing inherits determines what the rupture phase actually resolves into.

The transition this work treats as the clean upward supersession was a barbarism of the first order from inside it. The years from 1770 to 1848 were the enclosures and the expropriation of the English peasantry, the Atlantic slave trade at its British height with abolition arriving after the wealth was built, the clearances, the factory regime and the workhouse, the Napoleonic wars as continental mass death, and the Irish famine at the window’s close, a million dead while food was exported. Supersession and barbarism were one process seen from two sides, the productive leap and the human catastrophe one motion. Every mode transition on record ran through mass death, and the slogan Engels and Luxemburg left is about the terminus rather than the transition: the real distinction is barbarism that births a mode against barbarism that births nothing, and which one a transition was becomes legible only afterward. Branch 1 in reality looks like Branch 3 from inside it. Reading the catastrophe as proof of Branch 3 is the same error, in the other direction, as reading the recognition in the Tsinghua libraries as proof of Branch 1. The catastrophe cannot tell the branches apart, which is why the tests that can separate the branches are the ones on the successor’s tape. Compression concentrates the barbarism into a shorter window, and the climate layer is the medium it runs through. The strong analogy does not buy a clean resolution. It buys a supersession through catastrophe at best, and that is the ceiling.

The monarchical sea took 130 years from the crushing of its bubble to the universalization of the bubble’s content, telegraph, rail, and print doing the diffusing inside national containers. The capitalist sea drains on network speed, integrated supply chains, real-time capital, and a climate forcing with its own deadline, and the cracks of 1830, 1848, and 1870 have arrived early and stacked, 2008, the 2020s, Hormuz. Compression sets the speed and not the form. Compression is the other reason this work does not expect the century Plekhanov’s generation had to wait.

The coupling between the branches, held apart in Section XXI as a preview, became operational in the third week of July, and it has a date range. The strategic reserve asymmetry is now a live transfer visible in two weekly data series moving in opposite directions: the American reserve draining toward its authorization floor, at more than a million barrels a day through June and at about half that since, to hold a domestic price through an election season. China, having demonstrated during the closure that it can run for months at half its import rate on stockpiles and demand management, brings its arrivals back toward 9 million barrels a day while its stocks, on the tanker count, still fall, against a stockpile, state and commercial together, estimated at four times the American reserve. One great power is spending its buffer to maintain the appearance that no buffer is needed. The other spent the crisis proving how little it needs and is absorbing at the bid. This is the material substrate under the diplomatic surface: when Iranian state media confirms proposals from a mediator, the only mediator holding actual leverage is the one holding actual reserves, and Beijing’s interest in mediating is structural: the controlled transition of Section XXI is the branch in which Beijing inherits a functional system, and a disorderly autumn is the branch in which it inherits a wreck. The strongest argument that the rupture will be managed rather than catastrophic is that the actor with the deepest buffers is the actor whose entire doctrine requires the transition to be slow. Goodwill has nothing to do with it.³⁶⁹

The pace of the cascade is itself administered, and not by the incumbent. The 5-million-barrel consumption pause was the deepest buffer on the board, spent to slow the transition, which makes the tempo a policy variable held in Beijing. The pause ran from February into July, and on this work’s gauge the cascade ran slow while it held. Since the third week of July the successor’s arrivals have recovered while its stocks still draw on the September counts, and the long end has made its highs in the weeks since.³⁷⁰ A tempo that tracks the pause is the claim, and it can fail in both directions: a cascade that slows while Beijing buys, or speeds while Beijing holds off, is evidence against an administered pace.

One register of resistance to that adjudication has to be named, because it lives inside the tradition rather than outside it. Every prior succession looked, from inside the incumbent, like the loss of the human. The rising Dutch discipline read as Calvin’s Geneva against the humane corruption of the Catholic Mediterranean, the successor’s rigor ugly from inside the incumbent’s decadence, the incumbent’s humanism corrupt from inside the successor’s rigor, and each hegemon has moralized its accumulation in exactly this mirror. Marx caught the double edge in 1843, writing of Luther that he freed the body from chains by enchaining the heart. The fear that the present successor is Geneva again, discipline internalized where accommodation used to live, is the oldest internal criticism the socialist tradition carries, and it should be granted standing rather than argued away.

The loss should also be located precisely, because it is epochal rather than confessional. The mourning is for the characteristic form of the nineteenth and twentieth centuries’ capitalism, and that form is dying everywhere at once, in its homeland by its own government’s hand, as the convergence in Section XXI records. The era offers no choice between the old form and the new, only a contest over the new form’s class content. And the fear leads a double life that has to be kept separate. Inside the tradition it is criticism, and it has standing. In the incumbent’s apparatus it arrives fused with older material, the Oriental-despotism lineage running from the yellow-peril press through Wittfogel’s hydraulic state to the present civilizational genre, in which the loss of the individual is dreaded as the East rather than as Geneva.

The fusion is the evidence. Coding a class transition as a racial-civilizational threat is how the wealth question is kept from being asked. The orientalism supplies the affect and the affect launders the orientalism. The tests below owe nothing to that register. They adjudicate class content and nothing else. What can be said past that is this. The fear is what the falsification structure is for. The tests on the exit, the override, the articulation, and the class character of provisioning exist so that the Geneva question stays a question with observables attached, adjudicated by evidence rather than settled by the incumbent’s nostalgia or the successor’s catechism.

The claims above remain falsifiable, and the tests are near. A refunding operation that calms the long end through the autumn would demonstrate that the monetary toolkit is strained rather than exhausted, and the terminal reading would owe a revision.

The first test began adjudicating before this section closed. The 30-year yield reached a 19-year high in the third week of August, and the Treasury answered with the pre-registered signature itself: liquidity-support buybacks in the 10-to-30-year sectors at least doubled, to 4 billion dollars per operation, announced on the nineteenth, effective in September, described by the department as reflecting strong sponsorship and read by the desks as a circuit breaker at best. The first operation under the doubled program came on September 9 and disappointed the market it was built for: $6 billion against Street estimates of $8 to $10 billion. The 10-year rose to 4.86% on the announcement, its highest since late 2023, with the 20- and 30-year moving with it, and the department’s own auction that afternoon cleared the 10-year at 4.834% against 4.683 a month earlier, primary dealers taking their smallest share since September 2025 because investors bought the higher coupon.³⁷¹ The day before, on the yen, the secretary had told the market that he had asymmetric information and was the house now.

A house that bets small the day after announcing itself has told the table what it holds, and Société Générale’s head of US rates research said the rest: the buybacks show an understanding of the long end, and everything other than the direction of the debt and the deficit is cosmetic. The relief from the August announcement had lasted one session. The long bond gave back the 9 basis points the announcement bought within a day, and Deutsche Bank named the pattern, reading the buyback surprise and the same month’s yen support together as signs of increasing administration unease about the long end. Sustained strain at the Treasury’s own refunding operations was named above as the first signature of the handoff, and its first mark has appeared: the coupon it took to clear. The rest of the signature printed once on September 23: the five-year cleared at 5.033% against 4.393% a month earlier, 3.1 basis points above the when-issued, with bids at 2.21 times the offer against 2.37, the indirect share down from 61 to 54% and the dealers taking 16 against 10, on a day the flash PMI printed its highest since 2021, so the repricing carried a growth reading as well as a fiscal one.³⁷² The day after, on the Treasury’s second enlarged buyback, the 10-year touched 5.18% and the 30-year 5.47%, its highest in 22 years on the wire’s count, with oil at $108.³⁷³ One auction is a mark. The strain is repeated weakness against the when-issued with intermediation deteriorating, which is the criterion as it stood. The second signature has appeared as well, in the gold count above. The third, non-dollar settlement announced by producers, has not.

The month closed higher still, and the selling changed character. The 10-year ended September near 5.3%, more than half a point above where it began and at its highest since 2007, and the investors and traders the Financial Times spoke to described a loop: worry over the debt and inflation had driven yields to levels at which certain funds were obliged to sell, and their selling drove yields to the next such level. The forced seller of the earlier sections was a central bank defending its currency. This one is a fund obeying its own limits, and it needs no decision to sell.³⁷⁴

Three borrowers now draw on one pool of savings. The sovereign rolls and adds to its debt at the long end, the build-out borrows on the scale Section XVIII counts and will borrow more next year, and the listings, the largest in history among them, need buyers who must sell something else to make room. The price that clears the pool is the yield, which is why the 10-year’s climb was also the gated credit funds’ problem: a fund holding loans it cannot sell competes with a Treasury above 5% that can be sold any day, and the money its gates hold back would rather be in the sovereign’s paper, until the sovereign’s paper is the thing being sold.

A hyperscaler earnings cycle that validates the buildout through demonstrated, margin-visible productivity, not narrative, would rescue the circuit’s arithmetic. That test now has its arithmetic specified from outside the trade: net productivity is the automation saving minus the verification cost, against a comparison class, the computer revolution, that automated downtime rather than production, carried a verification cost of a different kind, code tested against a specification where generative output must be read for meaning, and still produced only a single decade of measured upsurge largely confined to the United States.³⁷⁵ Margin-visible validation means clearing that bar, not narrating it. The lines that would show it are filed ones: AI revenue from customers in whom the seller holds no stake, as a share of AI revenue, and the balance of assets not yet in service against AI capacity online, in the October and January filings.³⁷⁶ The one outside reading so far is Bloomberg’s September count of the largest cloud, given above.³⁷⁷ The seller’s own side has now priced the bar: Goldman Sachs’ strategists estimate that the cloud builders need about $300 billion a year of AI revenue to break even on their outlays, against cloud revenue running some $70 billion a year above its pre-AI trend, and AI users spending about $1 trillion a year on applications for the builders to earn solid returns, with 2027’s spending on track to take a larger share of output than any technology cycle since the railroads of the late 1800s.³⁷⁸ The demand is Lenin’s of September 1917, that the accounts of large-scale capital, already known to the hundreds of people inside it, be opened to the people it runs on, since commercial secrecy at that scale protects nothing but the concealment of “financial swindles and the fantastically high profits of big capital”.³⁷⁹

The restoration case those tests would confirm, stated as one sentence, is a conjunction: a technology real enough to validate the capex, not so real that it consumes the wage relation, financed to maturity without the circuit breaking. Each leg is possible. The product of the three is the width of the fourth autumn, and the tests on this list measure the legs separately, so the conjunction cannot be granted by mood or denied by it.

A Hormuz settlement that restores flows on the prior terms, American enforcement and no sovereign toll, would revive the enforcement function this work has buried. Five cycles have run toward that test and none has reached it. A conditional ceasefire on April 8 took Brent down nearly 16% in a session. A deal described in late May as largely negotiated took it under 98. Qatari reports of progress took it below 71 on July 2, beneath the pre-war price. The fourth cycle, opened when Washington halted strikes on July 24 and Iranian and Omani negotiators met over the following weekend, priced a joint regional arrangement under which Iran collects voluntary transit fees, and broke when the 60-day memorandum expired on August 17 without agreement, the blockade held, and Iranian strikes on shipping resumed within a day. The fifth, a phased framework announced by Tehran and Muscat for a jointly administered temporary corridor with joint mine-clearing, the demining itself contingent on Iranian cooperation, took crude down three consecutive sessions and broke at the end of August on the question recorded two paragraphs on.

Each time the price returned above a hundred. A market that has been wrong five times is not evidence. And the shape every cycle proposed is not the falsification: a settlement in which Iran collects transit fees is sovereign price administration at the chokepoint, conceded by the power whose enforcement function the chokepoint used to express, and it goes in the confirming column of the enforcement test, the relief that follows it scored on the prices. Each cycle arrives at the same result, a strait that reopens, if it reopens, under sovereign administration rather than restored enforcement. September showed what the escort can and cannot do. Under escort, with the main corridor cleared of mines, the Gulf’s exports rose to 12.8 million barrels a day, the most since the war began and well short of the 20 before it, and the week’s heaviest day, September 24, fell in a lull while Iran’s seven-day proposal was before the General Assembly. When Washington dismissed the proposal, a Kuwaiti supertanker was struck on September 28, and Iran said it had turned back nineteen tankers in two days. The escort can move barrels while Iran withholds fire. It has not been able to make Iran withhold it, which is enforcement performed and not restored.³⁸⁰ Iran’s team negotiating for a pause named itself Minab 168, after the count of the dead its government gives, and a society that had been in the streets against that government in January closed ranks after the strike, on the account of those who reported from it.³⁸¹

The dispute is no longer whether the strait is open. It is which sovereign’s lane is lawful, and the fifth cycle broke on that question on the same schedule as the fourth. Three tankers were struck inside a week at the end of August. On Sunday American forces struck two Iranian launchers on Larak Island, overlooking the shipping lanes, which Washington said Guard personnel were loading with rockets carrying sea mines, the first known US strike inside Iran since late July. Tehran answered with missiles and drones at the King Hussein and Al Azraq bases in Jordan, and Brent jumped back above 90 as soon as the shooting resumed. On Tuesday Central Command struck Guard targets around the strait after a pair of tankers came under attack, its published target list running to air defense sites, radar, maritime assets, mine-laying capability, and communications, and the president wrote that the strait currently has no mines one day after the launcher strike. The strikes came a day after reports that the president was weighing a plan for periodic attacks aimed at preventing Iran from rebuilding radar and missile capability around Hormuz.

The content of the exchange is the band made explicit. Washington insists the southern lanes near the Omani shore are open and safe. Iran insists all vessels coordinate with its military and use a northern passage close to Larak Island. Guard small craft identify ships at close range and order selected vessels to abort, Tehran has circulated a list of 46 ships it says breached its navigation regime, and by September 10 the discrimination was tightening against the accommodators too, Abu Dhabi blaming Tehran for strikes on two ADNOC vessels in the strait, and the maritime organization counts up to 400 ships and 6,000 seafarers still unable to leave the Gulf safely. The sovereign sell-off Section XVIII documents ran through the same sessions. A periodic-strike doctrine is the enforcement function reduced to a maintenance schedule, the sawtooth of Section XIX adopted as military policy, and each launcher destroyed on a rebuildable timeline is a round spent from the magazine Section X counts. The test keeps resolving the same way: the strait reopens, when it reopens, under an administered regime, and the administrator is the party being struck.

Two reservation prices were stated inside 24 hours. Iran’s president Pezeshkian on the successor’s stage at Bishkek, with Xi, Putin, and Modi in the room, offered immediate reciprocity if the United States returns to the memorandum it signed, called the February attack a violation of the Charter, and dated Washington’s compliance with its own signature at less than two weeks. At the Assembly on September 23 he added the strait’s rule, that no one enjoys free passage through a waterway it uses to attack Iran, the reopening on offer at seven days and the right to close it not on offer.³⁸² The president on his own platform announced he could not care less whether Iran signs a worthless agreement, that he prefers almost total control of the strait with their economy collapsing, and asked when the Iranian people will rise up. The Guard’s answer was that the strikes would tighten the lock, and the standing Iranian line is that if Tehran cannot export through the Gulf no one will. One party is speaking the language of the form. The other has stopped, and its Treasury secretary said as much at the G20 the same week: within two years the strait would be a worthless piece of water, 70% of its oil in pipelines.³⁸³ The owner of the enforcement function declared the chokepoint obsolete on the day it could not clear it. Both statements are floors, and the floors have moved apart. The incumbent’s own foreign-policy establishment had conceded the price the same week: the man who ran the Council on Foreign Relations for twenty years wrote in its journal that there was no way to justify the war, that the country and the region were worse off and Iran better off than before it began, that a fee structure for the strait had become unavoidable since the June text accepted Tehran’s right to collect payments, and that Washington should ask Beijing to bring Tehran to the table.³⁸⁴

The same 48 hours produced the split screen. At Asheville the Treasury secretary met the Russian finance minister, the first Russian at a G20 table on American soil since 2022, told him nothing was possible until the Ukraine war ends, and told the rest of the table that the economic-outcast sanctions announced the prior week on anyone facilitating Iranian trade expected their compliance, while the Europeans refused to stand in the photograph. At Bishkek Xi, Putin, Modi, and Pezeshkian sat in one room, and Modi’s readout with the Iranian president spoke of diversifying the trade basket. The incumbent’s core fractured in one venue while the successor’s assembled in the other.

A domestic labor test joins the test list above, named with the same discipline. Two consecutive quarters of positive real wage growth with rising participation would move the evidence toward the benign-transition branch, the passthrough arriving, the ethos revealed as an adoption lag rather than a substitute for capability. The current series runs the other direction on the wage leg, and August’s participation is one month against a two-quarter test. The test is stated now so that its outcome, in either direction, counts. A longer-dated credit test stands beside them: whether aggregate spreads remain tight two quarters after the equity complex has finished repricing.

The China-side tests join the same list, named with the same discipline, because the successor claim is falsifiable on the tape it prints. The household consumption share of GDP over the next four quarters, rising or anchored. The social wage as a share of the surplus, pensions, health, and housing. The destination of the margin the anti-involution campaign preserves, wages or retained earnings, readable in the industrial-profit series against the wage series. The terms of the next sovereign restructurings, haircut or duration. Local content in the periphery audited rather than announced, the Indonesian value-capture share published or still absent, and the African deficit narrowing or widening under the tariff. And the share of the plan routed to arms against the share routed to the household. A capitalist China prints as deflation with the banks holding the claims and the households holding the bill, or as the war economy. The other thing prints as the consumption share and the social wage rising while the export route is closed and while the private class loses command over investment, not one without the other. Either outcome counts, and the tests are stated before it arrives.

Four tests set in earlier sections join the list here, because those sections point to this one. The exit: the errors-and-omissions line of the balance of payments through the next season of pressure on the yuan, shrinking if the door closed in May holds and widening if it does not. The override: the day a market verdict at the commanding heights cannot be interrupted without systemic collapse is the day Section XXII said the state-capitalism reading wins. The class: if by the end of 2029 the strategic-industry fortunes have stayed intact through whatever successions have come, their owners sit in the organs that set policy beyond the consultative bodies they already sit in, and no act has come against them comparable to the three red lines, the capture reading has won, which is the property record of Section VII given a date.³⁸⁵ A fifth test, set at this revision on Cliff’s criterion and Lo’s: if by the end of 2029 labor power is still priced against accumulation’s needs, the labor share flat, alienated working time not shortened on the hours regime, the arrears record and the migrant’s status where they were, workers’ control over the labor process no wider, while the household’s share has not moved, the state-capitalism reading has won the ground the household test contests.³⁸⁶ And the articulation, which Section XXVIII dates to the contract expirations of May 2028 and scores on whether the siting coalitions federate.

A seat test joins them. This work claims the successor refuses the Wilhelmine path, and the claim fails on a stated observable: a quarantine or blockade of Taiwan initiated by Beijing before the end of 2030, absent a declaration of independence or a change in the island’s status initiated from Taipei or Washington, would show the successor taking the bid this work says it declines. A second observable measures the pressure toward it, the one Section XXVI names: the arms share already on this list. A third is the hull count: a sixth carrier laid down, or a second declared base abroad, before the end of 2030. A fourth observable names the case between the two tapes: a sustained campaign short of quarantine, ships stopped at sea, exclusion zones declared around the island, arms deliveries interfered with, would score against the refusal reading, short of the bid, without moving the end-2030 date.³⁸⁷ The confirming tape is the opposite pair, military pressure on the island falling while the incumbent is committed elsewhere, and the island’s terms moving by accretion. The peaking-power reading predicts the first tape and this work predicts the second, and the window closes inside five years.

One more test arrived after this list was first drafted and joins it under the same discipline, because the number attached to it does analytical work by itself. The laboratory at the front door of Section I has moved. A confidential prospectus is on file, half a dozen of its investors have confirmed a two-trillion-dollar target for an October listing to the Financial Times against projected year-end revenue of $100 to $120 billion, and one of them offered 3 trillion as, in his words, “on the incredibly low side.” The rival’s finance chief told employees its own listing comes in 2027 or sooner and called the IPO “not a finish line, it is a milestone, another fundraise,” the circuit describing itself in a sentence.

The target does not price on any history of earnings. Every entry on the structural side of the ledger runs against it. Price it from both directions. The valuation discipline has already run the reverse arithmetic: Damodaran’s generous case, a 30% mature margin and a 10% cost of capital, requires roughly $1.2 trillion in year-10 revenue, about 1.7 times Amazon’s entire 2025 sales, to justify the number, and Breakingviews prices the target at 40-plus times the spring run rate and 325 times annualized operating profit at the current margin. Every existing member of the two-trillion club earns $110 to $135 billion in net income. The candidate has barely crossed into operating profit at a 5% margin. The mark’s own velocity is the belief phase speaking: $183 billion in September, $965 billion in May, two trillion targeted in August, the doublings now running on projections of projections.

The funding side decides it, and the one real entry on the other side is growth. The two laboratories together must make or raise on the order of $1.25 trillion inside four years to meet their compute obligations, so the listing is the wall’s next tranche whatever else it is, which is what the rival’s finance chief said in plainer words. And the precedent is three months old: the June record listing at $1.77 trillion has since broken to lows that place it below 90% of billion-dollar debuts since 2009. The price signal arrived from inside the shop in the same month: Financial Times reporting of Ramp payment data shows the flagship model at roughly 11% of enterprise spending on the company’s models two months after launch, undersold by its own cheaper sibling released in late July, the first break in the pattern of enterprise customers migrating to each frontier release.

The bulls’ margin-expansion path requires pricing power at the frontier, and the frontier tier is the one going unsold, which is the value floor of Section XII arriving at the price. Against all of that stands one real entry on the other side. The revenue is real and unprecedented, a run rate from 9 billion to above 65 in seven months, and above 100 on a September report from one chain of reporting, enterprise-weighted, against recognized second-quarter revenue of 11.6 billion,³⁸⁸ and at the projected year-end figure the target prices at 17 to 20 times forward sales, rich against any history yet inside the range hypergrowth has commanded before. The balance still tilts to the bears, because the growth entry is the only one on that side and every structural entry sits on the other.

The test therefore takes its final form with a buyer condition attached. A print at or above the mark refutes nothing by itself, because Section XIII shows what a print does: it revalues the patrons’ stakes, books the revaluation as their earnings, and feeds the multiple the buy case rests on, the circularity’s largest single operation performed at the exchange. The falsifier is a completed listing at or above the private mark, held for two quarters, with the anchor demand visibly financed outside the patrons’ circuit, a sovereign buyer included if it passes the same test. A listing pulled, delayed past the window, or priced and then broken below the mark stays in the confirming column. So does one held up by captive books, where captive means money supplied by the patrons, or demand made to depend on their financing, their guarantees or their purchases in return, and the dependence has to show in the terms, since a buyer’s name or its wish for a higher price proves nothing. If the filings do not show who bought, the test scores nothing either way and the result is recorded as unresolved. The tape around it leans confirming: four senior executives out of the rival in a single summer, the data-center chief among them the day before the chip supplier reported, an operating loss there widening from 9 billion to 12 while quarterly revenue grew 18%, and the supplier’s own gross margin guided lower on memory costs even as it projected 70% growth two years out. The test is stated before the outcome so that either outcome counts.

And a Chinese stumble would collapse the second claim while leaving the first standing, which is the barbarism branch by another door. The stumble has six candidate doors: the property overhang; the deflationary undertow; the demographic arithmetic, which the robot count of Section XXVI answers in output and only the social-wage test can answer in distribution; the military-industrial capture Section XXVI refuses to rule out; the hinge the same section names, a Taiwan war that moves the successor from America’s seat to Germany’s; and the realization coupling Section XXIV now prices, in which the export circuit stalls on the very demand collapse this work documents, so that the stumble arrives because the first claim is right. Several of these tests print before the leaves turn, and one has already been run and postponed. At Oman’s initiative the foreign ministers of the six Gulf states, Iraq, and Iran were to meet in Salalah on September 14, the first meeting of the Gulf bloc with Tehran since the war began, to settle a transit regime for the strait, vessels entering through Iranian waters and leaving through Omani ones, with Iran asking for fees Oman rejects and conditioning the strait’s security on the lifting of the American blockade of its ports. Washington had opposed any arrangement that excluded it and had warned the Gulf states against negotiating directly.

On the eve, Oman postponed the meeting without a date, in its foreign minister’s phrase in the interest of consensus, after Bahrain declined to attend and Riyadh, on the Iranian account and on the American reporting, submitted amendments because the draft would have written a new status quo it could not accept, in the week its pipeline was shut and its client lost the coast, as Section XI records. Brent opened on Monday, September 14, above $106 and passed $108. The test stands as stated, and Salalah is a choice between two confirming outcomes. The falsifier is the one given above, American enforcement and no sovereign toll. If the region settles the strait without the incumbent, the enforcement function has changed hands in the one place the front said it could not. While the region cannot agree, the toll regime stays Tehran’s alone, and the Times’ report that the July attacks in the strait that broke the June memorandum were carried out by hardliners without the president’s or the Guard commander’s authorization is the measure of how many hands the regime is in.³⁸⁹ Two more tests carry dates just beyond them: the suspension of Beijing’s extraterritorial rare-earth controls expires on November 10 and the American-specific suspension on November 27, and the Pentagon’s rule reaching the Chinese mining and processing behind the magnets, tantalum and tungsten in defense supply chains takes effect on January 1, 2027, so that the valve turned by license since April 2025 meets a buyer barred by its own rule, exceptions aside, from the magnets that come through it. The claim is Section XXVI’s: the license desk persists past November, renewed or replaced, and no embargo comes. This work was written into a window. The window is now closing around it.

One phrase from the successor’s own vocabulary should be read once more, at the arc’s depth, before this work closes. The rejuvenation of the Chinese nation, the formulation Xi fixed at the center of the project, is translated by the incumbent’s commentary as nationalist boilerplate, a slogan of arrival. Against the branches section it is a periodization claim, and a precise one. Rejuvenation names a return, and the thing returned to is the condition the arc interrupted: the Sinocentric material economy that held the hemisphere’s center for centuries before 1433 and stands ready to hold it after. The fleet scrapped then and the fleet launched now bracket the interregnum, and the five centuries between them are the anomaly the word files away, the long twentieth century and its three predecessors reduced to a parenthesis in someone else’s sentence. The incumbent has no equivalent word, because the incumbent’s history is the parenthesis. What the phrase does not settle is this work’s question. A restored centrality says nothing by itself about the class character of the center, and the official teleology would read the same whether the returning formation is socialist construction or recomposed capital wearing the mandate. The branches remain the adjudication. The word tells the reader where on the calendar the adjudication sits.

The structural conditions for the end of the existing architecture are present. The form of the supersession depends on what is built in the rupture phase rushing to meet us.

Postscript: a conversation this work carries

Arrighi quoted Marx on the sequence of lending states and then corrected him in the next sentence. Marx cannot answer, but his texts do, and the argument between them, read with current events in hand, is an argument this work has been trying to fuse.

They concur on the history. In the chapter on the genesis of the industrial capitalist Marx wrote that the villainies of the Venetian system formed one of the secret bases of the capital-wealth of Holland, to which Venice in her decadence lent large sums of money. So it was with Holland and England: Holland, having ceased by the eighteenth century to be the nation preponderant in commerce and industry, turned to lending enormous amounts of capital to its great rival. And the same thing, he added, was going on in his own day between England and the United States. .

Arrighi’s correction is that the units in Marx’s sequence are of increasing size, resources, and world power, which Marx had not noted, so that the flow from the declining center to the rising one is also the passage to a larger container. Marx likely would not have contested it. The concentration and centralization of capital is the general law he stated for firms, and Arrighi applied it to states. The correction extends the law rather than reversing it. What Marx did not say and Arrighi did is that the sequence has a shape with an end, and the end is what the two of them argue about.³⁹⁰

Marx’s M-C-M’ is the movement of value through production to expanded value. Arrighi read it as a “preference for liquidity,” capital investing in a particular input-output combination only as a means to a greater freedom of choice later, and reverting to its money form when the expectation of that freedom is unfulfilled, which Braudel had called the “sign of autumn.” Marx’s answer is in the third volume of Das Kapital: The reversion to money is the form the falling rate of profit takes when the route through production stops paying, and the credit system that carries it is the lever of over-production and over-speculation, the purest and most colossal system of gambling and swindling, and at the same time the abolition of capital as private property within the framework of capitalist production itself.

The disagreement is real and it is open. Braudel and Arrighi read the financial expansion as a symptom of maturity that recurs. Marx read it as a fever that intensifies the contradictions of the system itself each time it recurs. The news has both potentials.

On the state they argue past each other, and this work needs both. Marx wrote that the alienation of the state, its public debt, marked the capitalist era with its stamp, and the Manifesto’s executive committee is the same claim from the political side: the state is capital’s instrument. Arrighi, through Braudel and Weber, holds that the important transition is from scattered to concentrated capitalist power, that capitalism triumphs when it is the state, and that the competition among states for mobile capital is the motor of the whole expansion. Marx would reply that the competition among states is the competition among capital displaced onto the state system, and Arrighi that Marx’s focus on the domestic side of accumulation kept him from seeing the interstate system as anything but a by-product, and from seeing what national debts continued to do in a system of states competing for capital.

Marx was more accurate about the one thing Arrighi’s Smithian frame cannot see, the class question. When a successor’s state stands over capital rather than under it, Arrighi can only call it a market economy that is not capitalism. Marx’s question is which class the state serves and whether the transfer runs toward the plan or away from it, and that question is the Preobrazhensky.

On China they were both predictive and both partly wrong, in opposite directions. Marx’s 1853 prediction, given in the third transition’s ledger, had the periphery’s crisis detonating the core economies, and the crisis he expected did not follow. The walls came down as the Manifesto predicted, and what returns now is the direction he gave the spark under a condition he could not have had, the core’s lines running on industrial inputs China makes, a dependence the core did not have on it in 1853, and it is that dependence and not the 170 years that carries the comparison. Arrighi in 1994 saw the East Asian archipelago, islands of capitalism rising above a sea of horizontal exchanges, thrusting their roots into the region’s laboring masses without giving them the means to rise to sea level, and in 2007 named China a non-capitalist market economy in Smith’s sense, accumulation without dispossession, the state controlling the capitalist class. He was wrong, on Marx’s terms, to call the formation Smith’s, because a market economy that makes its own proletariat by plan and holds capital under a party-state is the weak link’s transitional form and not Smith’s natural progress, and the difference decides whether the successor is a fifth capitalist container or the first non-capitalist one. Marx’s China is the more accurate on the class content. Arrighi’s on the geography and the timing.³⁹¹

On the limit they concur, and the concurrence is this work’s title. Marx’s sentence is the one the circuit section reads, that the real barrier of capitalist production is capital itself. Arrighi’s is the built-in-limit sentence the branches section gives, the same wall reached from the interstate side. The two sentences describe one wall from two sides, the production side and the interstate side, and this work reads both as denying that a fifth capitalist container can form, a reading Arrighi himself left open in 2009. Marx’s barrier is met by a buildout that would abolish variable capital. Arrighi’s by an incumbent so formidable that no capitalist power of a higher order can form against it, which is why this work asks whether the formation that did form is one. The thesis of this work is the conjunction of the two sentences, and neither man could have written the conjunction without the other.³⁹²

On the anomaly Arrighi was more predictive, and this work extends him. He noted in 1994 that the flows Marx had described as the instrument by which declining centers claimed a share of the rising center’s surplus had run the wrong way in the 1980s, from the rising center to the declining one. The bifurcation of military from economic power that the air-naval section records is the same finding from the other side. Marx had no category for a flow that ran the wrong way, because in his sequence the money followed the production. The anomaly is Arrighi’s finding. The tape’s finding is the sequel: the flow that ran the wrong way in the 1980s now runs nowhere, New York financing New York, and the military power that bifurcated from the economic one in the 1980s has since lost the strait. Arrighi saw the anomaly. This work records its completion.³⁹³

On the ending Marx was the more predictive and Arrighi the more exact. Marx’s ending is the abolition of the wages system, and the anatomy has the sellers’ own aspiration for it. Marx’s credit system abolishing capital as private property inside capitalism is the two-trillion listing whose revenue is its buyers’ investment. Arrighi’s endings are three, given in the question-mark section, world empire on the enduring superiority of Western force, a world market society in which that force withers, or the escalating violence of the transition consuming humanity before either, and this work’s branches are his partition. Where Arrighi is exact is the dating and the form. Where Marx is predictive is the content, because the market society Arrighi’s second ending imagines is a society in which value is still the measure of wealth, and the tape’s productive center has for a decade overruled that measure at the commanding heights, which Sections XXII and XXIV test as the beginning of a different law. The argument between them closes on the question mark in this work’s title: whether the fourth autumn ends capitalist history through Arrighi’s second ending or through the thing Marx described as the ultimate abolition, and whether the difference is one the successor’s formation will be allowed to decide.³⁹⁴