Section XXI: The Three Branches: A Preview
If the existing architecture cannot recover in its current form, what does the supersession look like? The scale the branches sit on is the one the front set out, the arc entire. The object is the architecture rather than the state. And the mechanism that decides the question is Arrighi’s, stated in the pattern section, and the four cases behind it are the ones the front told.
The arc has faced its own ending twice before, and the record of the two survivals is what makes the present audit decisive. The seventeenth-century general crisis and the 1914-1945 conflagration were both moments when the sequence could have terminated, and both times it survived on the same four conditions. A larger container stood ready: each regime’s territorial base exceeded its predecessor’s, city-state, proto-national federation, national empire, continental state, so accumulation always had somewhere bigger to recompose. Mobile capital was free to migrate to it. A crucible war crowned it. And an outside remained: territory, populations, and nature not yet subsumed, into which the recomposed system could expand. Run the audit against the present. The container series is exhausted as a series of types, because the only type larger than a continental state is a world state. The successor’s container is the larger of the two continental states, as the pattern section requires, and no type stands behind it, and the world state is the one formation the arc has twice refused.
The world state’s failure under the Habsburgs is what opened the sequence, and its failure at Hormuz, if the tests of Section XXIX hold, is what closes it. The strait that was to be the successor’s jugular, the artery a carrier group could close in minutes, closed instead on the incumbent’s allies, with the successor’s imports rerouted through the pause and the refill and the allies rationing jet fuel. The bid for universal empire bookends the arc, failing at the birth and failing at the death, with opposite residues: the first failure left the plurality that hosted five centuries of accumulation, and the second leaves a single productive center that does not bid for the pool at all. The pool itself is converted, as the two-sided lock below states. The crucible is blocked, as Section XXIII establishes. And the outside is closed, extensively since 1916, and now intensively, with the frontier series Section XX traces terminating in the enclosure of cognition, the frontier that abolishes its own validation.
The internalization series states the same exhaustion from inside the regimes. Each cycle internalized a cost its predecessor externalized. The Dutch internalized protection, the British production, the Americans transaction, the vertically integrated corporation swallowing the market’s coordination work, and a fourth internalization was speculated for the regime that would follow: reproduction, the making of the workers themselves. The fourth has begun to arrive, and not in a capitalist form. The party-state holds housing, education, industrial formation, and the reproduction of the workforce as plan targets, denominated in use-values, at the commanding heights Section XXII examines, and Section XXIV measures how much of the household’s cost it has taken on so far. What the incumbent attempted instead was a fifth, the internalization of cognition, the labor process as such, and it is the one internalization the series cannot contain, because it consumes the wage relation every prior internalization still required. The series, like the container series, runs off its own end.
With the arc opening at Hormuz and the incumbent’s last reach made for the same strait, the circle is closing from both ends at once, and the convergence is the evidence that the form question is settling. The incumbent’s own late motion is toward the “state capitalism” it spent a century defining itself against: equity stakes in strategic firms and threats to seize defense contractors, export licenses converted into revenue shares, a 12-billion-dollar federal mineral stockpile, a development-finance cap more than tripled so the state can take positions in foreign state-owned firms, tariff walls run as industrial policy, and a Treasury administering the bond market as a defended price. In the week of the September visit its own trade establishment counselled adding to the stockpile by buying it from the adversary, and its transportation secretary wrote himself “deeply alarmed” by a Michigan battery plant the White House had praised nine days earlier, over cells built on the successor’s chemistry under license.259 The scholarship has the numbers and the name. Sovereign funds now manage 15 trillion dollars against less than one at the century’s start. State enterprises hold 54 trillion in assets, near half of world product. Industrial-policy adoption rose ninefold in six years, and the political economists Alami and Riofrancos, cataloguing what they call Trumpian state capitalism, find state intervention unmoored from its twentieth-century ideological associations and adopted by regimes of all stripes, with Washington a late adopter of a movement two decades old.260
The American president states the conversion himself: taking stakes in companies, he says, is very American. What is dying is the characteristic form of the long twentieth century’s capitalism, private allocation wearing a liberal state, and it is dying in its homeland by its own government’s hand. Every formation that survives the cascade arrives at the state commanding capital. What the convergence does not settle is the class content of the command, which is why the branches partition on class character rather than on the state-market axis the incumbent’s ideology still argues in, and why the envy with which Washington copies Beijing’s instruments is a tell rather than a paradox: the incumbent imitates the successor’s form in defense of the incumbent’s content.
The right image is therefore the Counter-Reformation rather than an emptied church. Rome after the Council of Trent, 1545 to 1563, had congregation in the hundreds of millions, wealth worth defending, and the discipline to adopt the reformers’ instruments, cleaning its own abuses and founding its own militant orders precisely to keep the hierarchy and the treasury intact, and on those terms it held half of Europe for more than a century. The militant orders have their modern counterparts in the economic-statecraft cadre of the councils and think tanks, the chips, minerals, and export-license specialists, founded like the Jesuits to fight the reformation on its own disciplinary terms. The state capitalism now running from Washington is Trent’s move with Trent’s purpose, the challenger’s discipline in service of the incumbent hierarchy’s position and wealth, and it commands a congregation that remains vast. What Trent needed and got was a crucible, 30 years of one, to fix the confessional map. This counter-reformation gets no crucible, so its map will be fixed by accretion, and the case in which Trent holds is the second of the paths below, which this work argues cannot hold for long.
The finding takes a name of its own alongside the fuze and the doom loop, because it is this work’s actual revision of the framework: the two-sided lock. Arrighi asked in 1994 whether the structures of US capitalism constituted the ultimate limit and left the answer open. In paraphrase, his finding was that capitalist power cannot expand indefinitely without undermining the interstate competition for mobile capital it rests on, the alliances of state and capital grow formidable enough to eliminate the competition and with it the possibility of a new capitalist power of a higher order, and the question is whether the structures of American capitalism are the ultimate limit of the six-century process.261 The answer this work argues is affirmative, on a mechanism with two sides. One is on the tape, and the other is the second claim, which Sections XXII to XXV test: the incumbent converted the mobile-capital pool into an enforcement instrument, and the one formation holding the productive mandate does not compete for the pool. No prior near-ending of the arc approached this configuration.
In the seventeenth century and in 1914-1945, some of the four conditions failed and the surviving ones rebuilt the rest, a larger container and an open outside available each time. A recurrence machine locked from one end strains, and the first half of this work records the strain. Locked from both ends it stops, and whether the second side holds is what the branches below partition. Three cautions bound the claim. The schema itself is an interpretive scheme, and the argument rests on the mechanism rather than on the numerology of cycle lengths. Robert Brenner’s standing charge, that cycle theory is Smithian, locating the dynamic in trade, finance, and hegemony while the crisis lives in production, is met by this work’s division of labor: the reproduction schemes and realization mechanics of Sections XII and XV supply the motor the schema lacks, the arc gives the crisis its 500-year address, and neither is asked to do the other’s work. And the one regularity of the cycle lengths worth keeping, their compression across the sequence from roughly 220 years toward one hundred, is claimed only as lineage: the inter-cycle ancestor of the compression thesis this work runs at intra-crisis scale.
The audit above has a military column, and it is datable at both ends with the same precision as the birth certificate. The plurality that made states biddable made them armed, and the competition that mobile capital financed produced, within two generations of the arc’s birth, a war-making capability no other civilization could match. The account opens in 1492. At Cajamarca 168 men took an empire’s sovereign in an afternoon, and for five centuries the pattern held wherever the arc’s core met the world it priced: the asymmetry seized the ghost acres the historian Kenneth Pomeranz counts, the Maxim gun held them, and the gunboat collected. The China bracket states the signature.
The fleet scrapped after 1433 was the largest in the world. Four centuries later gunboats on the Pearl River opened the market of the formation that had scrapped it. And that formation’s yards now hold more than 200 times the building capacity of the arc’s last hegemon, on that hegemon’s own estimate.262 The asymmetry is the audit’s fourth condition stated in military form. The outside stayed open because the core could always kick the door in, and every accumulation regime in the sequence rested on that capability the way the settlements rested on the plurality. An arc that opened with the conquistador closes when the doors stop giving, which is the fact Section XXIII’s crucible and Section XXVI’s launch rates record from two directions.
The prior transition clarifies the structure, on condition that the right formations are matched to the right roles. The Soviet Union was the encircled bubble on the capitalist sea, the premature and weaker formation the stronger system contained and outlasted, and its fate is the trapped-transition failure mode the rest of this argument works to distinguish China from. The USSR is the analogue of revolutionary France, the forcing carrier that was crushed and whose content diffused into the sea that beat it, the welfare settlement the imperial core adopted as the price of competing with the alternative the USSR represented. China is not that formation. China is the analogue of Britain in the prior transition, the mixed formation that was the rising center rather than the encircled periphery, carrying the more advanced base under the old order’s retained forms. Britain carried industrial capitalism while keeping a monarchy, a Lords, and a landed aristocracy it never fully cleared and has not cleared yet. China carries a directed and socialized base while keeping markets, private firms, and a bourgeoisie.
The mixed formation that wears the prior order’s birthmarks is what the durable carrier of a transition looks like, and the rising carrier turning on the premature one is part of the pattern: Britain organized and financed the coalitions that destroyed revolutionary France, and China defected from the Soviet bloc, took the American hand in 1972, and helped complete the encirclement that exhausted the USSR. The analogy has three limits, and the later sections carry them. Britain was financed by the incumbent it displaced, and so was China for 30 years, by the offshoring Section IX records. The difference is the return leg: China recycled its surplus into the incumbent’s debt, slowed the recycling after 2008 and stopped it after 2013, and has propped the incumbent since by other means, the consumption pause that held the world’s energy price through the closure being the latest, which is the anomaly of Section IX in a new form. Britain fought its crucible and wrote the peace, and China’s crucible is blocked. And Britain took an empire, which Section VI gives the successor’s reason for refusing.
This is where the three branches acquire their full stake. Branch 1 is the bourgeois succession ending for the first time in five centuries, the arc terminating in a non-bourgeois successor rather than recomposing. The recomposition case, in which China’s directed forms turn out to be the cosmetic and capital the essence, is the bourgeois arc continuing under a new center, the Arrighi sequence extended to a fifth seat, and it is the same outcome the imperialism test in Section XXV reaches when it asks whether Beijing converges on Lenin’s criteria. The frame does not decide between them. The frame is what the property record of Section VII and the override test of Section XXII adjudicate, raised to the scale at which the adjudication matters. The arc ending is named here as the thing the falsification structure is for, not asserted from the pattern.
Three resolution paths are available to the cascade.
The first path is a dominant Beijing plus Marxist articulation ascendant in the imperial core. The global productive economy reorganizes around planned political-economic principles. Party-state political control of capital becomes the dominant ordering form. Explicit productive planning frameworks replace the present regime of fictitious-capital accumulation. The eventual transition past the primary stage of socialism into socialism proper, which the CCP names as its long-term goal, becomes the operative trajectory. This branch represents the actual end of capitalism as a global ordering principle and its supersession by the form Marx specified as its dialectical successor when the world market stage had been reached and the contradictions had become globally synchronized.
The second path is contested. The Chinese productive substrate organizes through the Beijing-anchored architecture. The G7 reorganizes through Orbanization-pattern personalist-authoritarian consolidation, with variations in aesthetic register and policy from authoritarian-conservative to more-explicit-fascist depending on the case. The world system operates as asymmetric multipolar competition between the structurally durable Beijing project and the structurally declining American-aligned framework that persists in transformed form at sub-hegemonic regional scales. This branch terminates the same way as the first branch on a long arc because the substrate gap between the productive center and the contracting imperial core widens structurally and the contested configuration cannot hold against asymmetric productive dominance indefinitely.
The third path is the failure case for both other branches. Beijing fails, is destroyed, is captured by its security apparatus at a level its institutional capacity cannot contain, or is drawn into the Wilhelmine bid it has so far refused. Multiple regional reorganizations proceed without coherent successor architecture, with each region’s productive base too thin to anchor a new system.
Climate substrate compounding looked like the slower layer of this branch, instability arriving behind the financial-political rupture but ahead of any successor’s recovery capacity, and the summer showed otherwise. The continent-wide European drought and heatwave sequence ran inside the war window rather than after it, a season’s yields and a war’s inputs sampling the same throughput variable in the same quarter, which is the compression thesis operating on the climate layer itself. The slower timescale holds only for the deep-substrate terms of AMOC weakening, monsoon disruption, and agricultural belt migration, and even those can arrive in parallel.
Engels’s formulation, later compressed by Rosa Luxemburg into “socialism or barbarism,” names the alternative as regression rather than continuation. The barbarism path is not a science-fictional outlier. It is the path that comes if the structural durability of the alternative architecture turns out to have been overestimated and the political articulation in the imperial core turns out to have absorbed the recognition into forms unable to organize productive supersession.
The three paths are stated on the imperial-core axis, and a second question runs under all of them, the class content of the successor. Branch 1 and Branch 2 both assume that the directed forms are the essence and capital the instrument. If capital turns out to be the essence, the outcome is the recomposition. Sections XXIV and XXV test for it, and it can arrive beside a Marxist core or an Orbanized one. This work keeps the branch numbers for the core and calls that case the recomposition.
The branches are more coupled than parallel, and the coupling is material: the core’s collapse closes the recomposition’s exits, the export of overcapacity and the export of capital to a solvent periphery, which is what forces the choice. The structural durability of the Beijing project is conditional on the transition. A controlled American retreat leaves Beijing inheriting a functional system it can integrate with. A catastrophic American collapse leaves Beijing inheriting a wreck it may not have the institutional capacity to stabilize. Survival under those conditions requires the kind of mobilization that destroys the gradualist developmental path that Reform and Opening took up in 1978 and the primary-stage doctrine has protected since 1987. The structural difference between Beijing and the USSR holds under controlled transition conditions and compresses under catastrophic ones. The danger inside the controlled branch is the interwar one: Kindleberger’s sentence on the 1930s was that Britain could not and the United States would not take responsibility for the system, and a successor whose doctrine requires the transition to be slow can become a successor that will not, beside an incumbent that cannot.263
On agency
The branches are stated as if the formation whose trajectory decides them were choosing. Marx’s rule for that question is the one from the Eighteenth Brumaire: men make their own history, and not under circumstances of their own choosing. Beijing has a plan. The test is where in the sequence the plan is the cause and where it is the name the formation gives to what the base required of it.
The evidence for a forced hand is most of the record. The 2008 stimulus, 4 trillion yuan into infrastructure and productive capacity, was the only response a formation with that fiscal room and that base could make, as financial intermediation was the only response the incumbent could make. Each side chose the one response its base allowed. The property deflation from 2020 was a choice of timing over a fact: the sector had grown to a share of output no economy has sustained, and the three red lines decided when it would stop rather than whether. Involution, the deflator falling for nine quarters, the export surplus above a trillion dollars, the anti-involution campaign, are the base absorbing what a core that cannot buy no longer takes. The yuan bridge was built by the freezing of a sovereign’s reserves in February 2022, a necessity for every state that could imagine being next. The phosphate and rare-earth controls were built by the tariff war. The pattern states the rule: the challenger’s spring is the season of least freedom, because the challenger must produce to exist and cannot yet finance, and every successor in the arc was carried into its position by what the incumbent’s autumn did before it chose anything.264
The evidence for agency is the refusals, because a refusal is the only act a constrained formation performs that the constraint did not require. Beijing refused to rescue the developers or their creditors when the incumbent’s precedent was to rescue both. It refused to recycle its surplus into the incumbent’s debt after 2013, when the recycling was the price of the export market. It refused the platform-capital path at the Ant listing and the Didi filing, when that path was the one the incumbent’s own base had taken. It refused the Wilhelmine confrontation the incumbent’s planners have been inviting since 2017. It refused the rare-earth embargo the war invited and priced the licenses instead. It has refused the reserve role on the incumbent’s terms. Xi’s program for a financial power names a currency with global reserve status among its six aims and sets beside it a capital account opened only under control, an infrastructure under the state’s own control and finance held to the service of production. Its central bank’s governor proposed in 2009 that the world’s reserve be a basket no nation issues, and what it lays abroad is a payment rail. The role as the incumbent held it came with the deficits and the financial expansion that hollowed it. What Beijing has refused is that path, and whether the role can be held without it is untested. A seventh is often claimed and holds only in part. It ran no consumer-credit boom of the kind the incumbent’s households ran before 2008. It did let the household borrow for the apartment, and household debt went from under a fifth of output in 2008 to over three-fifths by 2024, which is the property expansion of the paragraph above seen from the buyer’s side.265
The incumbent’s counterpart to these refusals lies a decade before 2008. In the 1990s, the decade the next paragraph takes up, Washington could have kept the base and chose to ship it, and by 2008 finance was the one response the hollowed base allowed. Agency is measured by what a formation refuses when refusal costs it, and by that measure Beijing has it in the sequence, which crisis to take first and which to defer, the property sector before the platforms, the platforms before the war, and in the patience that could, on Cory Doctorow’s proposal, buy the chips at 10 cents on the dollar when the companies that bought them are bankrupt.266 What it does not have is the direction. The direction is the base’s, and the successor’s test of agency arrives only when it becomes the incumbent: whether a formation that was carried into production can refuse its own autumn when the autumn is offered. This work calls that the value-form test, and no successor in the arc has passed it. It waits on an autumn the successor does not yet have, which puts it outside this work’s window. The closing section carries its near form, the override test of Section XXII. The migration is underway and unfinished: the successor still sells for money, pays wages, and borrows on the world’s markets, and the claim is that the commanding heights run on the other law while the base stays entangled in the world market, which is why the closing section’s tests are this work’s argument rather than its conclusion.
The American 1990s are the control case, because they were the incumbent’s last season of freedom and the record shows what it did with it. In the decade after 1991 the United States believed it had left the old capitalism behind through computing and the network: the new economy, the productivity revival that ran from 1995 to 2004 and was real, the business cycle declared repealed, a magazine promising 25 years of prosperity. The mechanism was real, the one decade of measured upsurge the closing section’s test cites. The distribution was the choice: the gains went to capital and the top decile, the base was offshored while the productivity was counted at home, the crash of 2000 was answered with the housing bubble and the Iraq reach, and the top percentile’s share of income returned to its 1928 level by 2007 on Piketty and Saez’s series.267 The Gilded Age was restored by the technology that was supposed to end it, and the founder fortunes replaced the railroad fortunes as its emblem. AI is the same belief administered again, with the 1990s as the moment reached for: the network’s promise carried to its end, the incumbent reaching back for its last productive spring.
The industrial-revolution comparison the sellers make cuts both ways. The steam loom’s cloth was worse than the handloom’s for decades before it was better, and the workers who broke the looms were the most skilled technicians of their trade, angry about the quality as much as the wage. The machine did in the end exceed the artisan, and the transition destroyed a class and the knowledge it carried in the meantime. The present transition has the form, a knowledge-working class displaced by untrained labor and a machine that produces objects shaped like work, and lacks the ending on the incumbent’s books, because the frontier’s cost is not falling, the challenger’s is, and the phase in which the machine becomes better is being funded where the incumbent cannot collect on it. What returns is the social question of the 1830s, the poverty of those who live by selling their labor to a market that has found a substitute, and the articulation contest is where that question is being fought.
Every branch runs under a constraint the branches themselves cannot alter. The infinite growth that debt-based capitalism requires, the finite resources that planetary boundaries impose, and the stable climate that civilization requires form an impossible trinity: any two can be held for a time and all three cannot be held permanently. The clocks are synchronized, and two of them are load-bearing: the emissions deadline the climate assessments place at 2030, and the Taiwan window the American assessments place between 2027 and 2030. The AGI window the labs sell for the same years is a sales calendar, and the resource peaks sometimes set beside them, phosphorus, helium, the rare earths, are contested, the phosphorus peak most of all since the reserves were revised upward in 2010, and they are carried here as a flag, where the emissions deadline and the Taiwan window are clocks.268 The attempt to square the trinity is what drives the extreme behavior this work records at the strait and in the data centers, and it is why the compression is a property of the period rather than of any actor in it.
The question of which branch is realized turns on three live experiments: whether China is structurally different from the Soviet Union in ways that prevent the trapped-transition trajectory the Soviet experience exhausted. Whether the topology of the world system actually inverts in the way Beijing claims. Whether the political articulation contest in the imperial core resolves into the form that absorbs the rupture into Marxist content rather than into fascist content.
Notes
- 259Adam S. Posen, "China Still Has What America Needs," Foreign Affairs, September 21, 2026; the Department of Transportation’s letter of September 3, 2026, public September 8 (BlueOval Battery Park, Marshall, Michigan, on CATL-licensed cells; Ford’s reply); the White House release of August 31, 2026, praising the plant. Back
- 260Lenin had the incumbent’s form in 1917: Germany and America “regulate economic life” by reactionary-bureaucratic methods, war-time penal servitude for the workers and a paradise for the bankers and capitalists (Lenin, “The Impending Catastrophe and How to Combat It” (1917), Collected Works, vol. 25, chapters 4 and 11). Back
- 261Arrighi, The Long Twentieth Century, 19. Back
- 262The Office of Naval Intelligence slide cited in note 1. Back
- 263Charles P. Kindleberger, The World in Depression, 1929–1939 (1973); Joseph S. Nye, “The Kindleberger Trap”, Project Syndicate, January 9, 2017. Back
- 264State Council of the People’s Republic of China, November 2008. Back
- 265CEIC, on People’s Bank of China and National Bureau of Statistics data: 17.6% of GDP in December 2008 and 62.3% in March 2024; BIS total credit statistics, 62.1% for 2023. Zhou Xiaochuan, “Reform the International Monetary System,” People’s Bank of China, March 23, 2009, on the Triffin dilemma and a reserve built on the SDR. Xi Jinping, “走好中国特色金融发展之路,建设金融强国,” Qiushi, no. 3 (2026), published January 31, 2026, from remarks of January 2024: the first of six elements of a financial power is a strong currency, widely used in international trade, investment and the foreign-exchange market and holding the status of a global reserve currency, beside a financial infrastructure that is independent and controllable. Bordo and McCauley, “Triffin: Dilemma or Myth?” BIS Working Paper 684 (2017), against the reading that a reserve role requires the issuer to run current-account deficits. Back
- 266Doctorow, interview, September 2026: a national AI strategy should begin by doing nothing until the chips can be bought at ten cents on the dollar in bankruptcy. Back
- 267Piketty and Saez, Quarterly Journal of Economics (2003) and the updated series: the top percentile’s share of pre-tax income near 24% in 1928 and again in 2007. Back
- 268Cordell, Drangert, and White, “The Story of Phosphorus” (2009): peak between 2030 and 2040 on the USGS reserves of the time; Mohr and Craswell (2033); Van Kauwenbergh, World Phosphate Rock Reserves and Resources (IFDC, 2010) and USGS (2012) revised reserves to 60 and 71 billion tons, a reserve-to-production ratio near 370 years. Back