The Chinese Revolution is the name given to the process by which the Chinese Communist Party, driven from the cities in 1927 and reduced there to a hunted underground, rebuilt itself among the peasantry of the interior, expanded enormously during the war with Japan, and defeated the Nationalist government of Chiang Kai-shek in the civil war that concluded in 1949. A party that had come close to extinction as an urban organisation ended by founding a state, the People’s Republic, and remaking the society it governed. On the sequence of events the scholarship is broadly agreed. On the question of why the communists won, it is not agreed at all, and the disagreement is the most instructive thing a reader can take from the subject. One celebrated account explains the victory by nationalism: the Japanese invasion created a mass patriotic response, and the communists were the force positioned to organise it. A second, equally celebrated account explains it by social and economic reform: in the base areas the party governed, it reduced rents and interest, organised cooperatives, and administered honestly, and peasants supported the power that had improved their lives. A third body of scholarship, working later from local archives, added a factor neither of the first two had foregrounded: organisation itself, with its discipline, its surveillance, and its coercion, as a means by which support was manufactured and held. This article maps that three-way dispute, weighs it, and then turns to the theoretical question the victory poses for Marxism itself, since a revolution made by a peasant army in a country with a negligible industrial proletariat either extends the theory or breaks its central category. The reader who finishes here should be able to state the dispute precisely, assign each explanation its evidence and its limits, and say on what criteria the argument about the revolution’s character actually turns.

Two cautions govern everything that follows. The first is that the three explanations are not three opinions of equal looseness; each was built by a named scholar or school from a distinct body of evidence, and each will be given here at full strength with its attribution attached, because an explanation weakened for convenience teaches nothing. The second is that the article treats the revolution historically throughout. It names no serving official, evaluates no later policy, and offers no commentary on the politics of any successor state; its subject is a dispute among historians about events that closed in 1949 and the years immediately following, together with the theoretical problem those events left behind. Where the record involves violence, and it does at every stage, the violence is reported as a fact of the record, without detail and without the casualty figures that the sources themselves cannot agree upon. Where the record is genuinely open, the article says so and shows the reader the shape of the opening, rather than closing it by editorial preference. The dispute about why the communists won has lasted more than half a century because the evidence genuinely supports more than one reading, and the honest scholarly position, which this article adopts and defends under the name of the three-factor settlement, is that all three factors operated, in proportions that shifted from one district and one phase of the war to another, and that the long quarrel between the first two explanations was conducted before the archival evidence for the third had become available to either side.
The Chinese Revolution in outline: a party destroyed and remade
The Communist Party of China was founded in 1921, in Shanghai, by a small circle of intellectuals and students, with encouragement and organisational assistance from the Communist International, the Comintern, the Moscow-based body that directed the world communist movement. Its first strategy was set for it by that parentage. On Comintern advice, and against the instincts of some of its own leaders, the young party entered an alliance with the Nationalist Party, the Kuomintang, whose leader Sun Yat-sen had accepted Soviet assistance for his project of national reunification. Communist members joined the Kuomintang as individuals while retaining their own organisation, a arrangement the Chinese called the first united front, and within it the communists grew rapidly by organising where the Nationalists did not: among the industrial workers of the treaty-port cities, in the new trade unions, and in the peasant associations of the southern provinces. When Sun died in 1925, the alliance passed to his military heir, Chiang Kai-shek, who led the Northern Expedition of 1926 to unify the country against the warlords, with communist organisers mobilising workers and peasants ahead of his armies. The partnership carried the seeds of its own destruction. The communists were building an independent mass movement inside the Nationalist revolution, and Chiang, whose officer corps and financial backers feared that movement more than they feared the warlords, resolved the tension by force. In April 1927 his forces and their allies in the Shanghai underworld turned on the city’s organised workers, and the communist-led unions of Shanghai were destroyed in days. The Wuhan government, the Nationalist left’s rival seat of power, expelled the communists months afterward. The party’s urban organisation, its unions, its press, and a large part of its cadre were annihilated or driven into an underground existence from which the cities were never recovered as the party’s home.
What happened to the communist organisation in the year of the rupture?
In April 1927 the Nationalist commander Chiang Kai-shek turned on his communist allies, and the party’s urban organisation was destroyed: its unions smashed, its press closed, its cadres killed, imprisoned, or driven underground. The Wuhan Nationalist government expelled the communists soon afterward, and the party’s first strategy, built on the cities and the alliance, ended in catastrophe.
The catastrophe of 1927 is the hinge on which the whole history turns, and its scale needs to be felt before the later achievement can be measured. The party had entered the year as a growing movement with real strongholds in the most modern sectors of Chinese society: the mills and docks of Shanghai, the railways, the mines, the universities. It ended the year as a remnant. Attempts to recover by insurrection, at Nanchang in August, in the Autumn Harvest risings in the countryside of Hunan and Jiangxi, and in the desperate seizure of Guangzhou in December, which held the city briefly as the Canton Commune, were each defeated. The Comintern’s line, which had first required the communists to remain inside the Kuomintang and then, after the rupture, demanded immediate uprisings in the cities, had produced disaster at both of its stages, a fact the article returns to when it weighs the Comintern’s role in full. What survived the year survived in two forms: an underground apparatus in the cities, perpetually raided and never again the movement’s centre of gravity, and a scattering of armed bands in the hill country of the provincial borders, where defeated soldiers and organisers took refuge among a peasantry the modern state had always governed loosely. The party’s future, though no one could yet have described it, lay with the second of these. The movement that would conquer China in 1949 was not the continuation of the urban party of 1926 by other means; it was a different organisation, built in the countryside out of the wreckage, by men who had learned from the wreckage that survival required an army, a territory, and a rural population organised as the base of both.
The inheritance: a country of fragments
The revolution’s first task, before any question of doctrine, was set by the condition of the country it proposed to remake, and a reader who does not hold that condition in view will misjudge every explanation of the outcome. The empire that had governed China for two millennia collapsed in 1911, and its successor republic fragmented within five years into the satrapies of competing militarists, the warlords, whose armies taxed, conscripted, and fought across the provinces for a decade. The warlord era’s legacy was not merely disorder. It was a population schooled by experience to regard every government as an exaction and every army as a calamity, a countryside in which the state’s agents appeared as tax farmers and press gangs, and a political culture in which the reunification of the country had become the one programme that commanded assent across every class and region. The Nationalists’ Northern Expedition of 1926 to 1928 achieved a partial unification, and the Nanjing government of the following decade built real institutions, a central bank, a modern bureaucracy’s beginnings, treaty revision, in the territory it firmly held, which was never more than a fraction of the whole. Its writ ran weakly in the provinces, where militarists retained their armies under new titles, and hardly at all in the villages, where the county government remained what it had been under the empire, a yamen at the county town taxing a countryside it did not administer. This is the vacuum into which every actor in this article moved. The communists’ eventual achievement, the extension of an administered state to the village level across the entire country, was revolutionary not only in its social content but in its mere administrative fact, against a background in which no Chinese government of the modern era had accomplished it. The point disciplines the first explanation in particular. The nationalism that the invasion aroused was powerful precisely because the nation it invoked had never yet been made into a state the villager could touch, and the party that promised to embody the nation arrived, uniquely, with the organisation to make the embodiment local.
The founding and the first mass movements
The party that perished in the cities in 1927 had been, in its brief life, a more various movement than its end suggests, and its early repertoire explains much of what the rural party later revived. Founded in 1921 by a dozen delegates meeting in Shanghai and aboard a boat on South Lake, under the guidance of the Comintern’s emissary, the Dutch communist Maring, the party began as a circle of intellectuals armed with a translated doctrine and a conviction that China’s salvation ran through its small modern working class. Within five years it had built, or helped to build, three mass movements of genuine scale. In the treaty ports its organisers led the labour movement through the great strikes of the middle 1920s, of which the Guangzhou-Hong Kong strike and boycott of 1925 and 1926, sustained for sixteen months with its own pickets, courts, and dormitories, was the most formidable labour action of the decade anywhere in Asia. In Hunan, Guangdong, and the Yangzi provinces its peasant organisers, of whom Peng Pai in the Haifeng district of Guangdong was the pioneer, built peasant associations that enrolled villagers by the hundred thousand, reduced rents, and humbled the local bullies whom the associations’ reports catalogued with relish. And in the universities its student members formed the cadre of the May Thirtieth movement of 1925, the patriotic explosion that followed the shooting of demonstrators in Shanghai and that carried the party’s name into every school in the country. The repertoire on display in these years, the association, the boycott, the mass meeting, the patient organisation of a constituency around its grievances, was the party’s invention in China as much as the Comintern’s, and it survived 1927 because the men who had practised it survived, in the hills, and taught it to the peasant armies. The rural party’s methods were not a rustic improvisation. They were the urban movement’s methods, transplanted to the only soil in which the Nationalist terror had left them room to grow, and the continuity of personnel and practice across the catastrophe is one of the underappreciated facts of the history.
Who was Peng Pai, and why does he matter to this history?
Peng Pai was the party’s pioneer of peasant organisation, who from 1922 built the Haifeng peasant association in Guangdong into a movement that reduced rents, defied the landlords, and briefly governed its district in 1927. He demonstrated, before the rural turn was forced on the party, that the Chinese peasantry could be organised on a mass scale.
The early movement also fixed a liability the party carried into its rural decades: its dependence on translation, in every sense. The doctrine arrived in texts rendered from Russian and Japanese, its categories, proletariat, bourgeoisie, feudal, fitted to a European history that China’s social order resembled only in patches, and its strategic instructions arrived from a headquarters whose knowledge of China was filtered through the same translations in reverse. The party’s best organisers were precisely those who violated the translations in practice: Peng Pai, who learned to speak of the association in the idiom of the village and to burn the debt contracts in the temple courtyard; Mao, whose Hunan report insisted that the peasantry’s own categories of grievance, the landlord, the usurer, the corrupt clerk, were the revolution’s real syllabus. The sinification proclaimed at Yan’an was thus the formal recognition of a practice as old as the party’s successes, and its opposite, the mechanical application of imported schema, was the common thread of the party’s disasters from 1927 to the loss of Jiangxi. The reader who grasps this has grasped something the three explanations all assume: the party that won in 1949 was not the party that had been founded in 1921 with a name added, but an organisation repeatedly destroyed and repeatedly rebuilt by the friction between a universal doctrine and a particular society, and its victory, when it came, belonged to the version of it that the friction had produced.
The rural turn: base areas and the Jiangxi Soviet
The most consequential of the survivors was Mao Zedong, a Hunanese organiser whose report on the peasant movement in his home province, written early in 1927, had already argued that the peasantry was not the passive object of revolution but its storm force, rising, as he put it, like a tempest. After the failure of the Autumn Harvest Uprising he led his remnant to the Jinggang Mountains on the Hunan-Jiangxi border, where in 1928 he was joined by the mutinied soldiers of Zhu De. There, in poor hill country where state power was thin, the elements of the later system were assembled: a Red Army that lived off and protected a defined territory, a soviet government that administered villages, and a programme of land revolution that confiscated the holdings of landlords and rich peasants and redistributed them to the poor. Other base areas formed elsewhere, in the hills and lake districts of several provinces, each a small armed republic in embryo. In 1931 the Jiangxi base, the largest, proclaimed itself the Chinese Soviet Republic, with its capital at Ruijin, a state within a state, issuing currency, running courts and schools, and conducting a land reform that in its thoroughness went far beyond anything the party would attempt during the later war years. The Jiangxi period matters to the historiography for two reasons. It proved that a communist movement could govern Chinese territory and survive for years against a hostile state, which no urban underground could do. And it fixed the party’s social character: whatever its doctrine said about the proletariat, the organisation that survived the early 1930s was a peasant army officered by intellectuals, and every later debate about the revolution’s nature starts from that fact.
Why did the party rebuild in the countryside rather than the cities?
Because the cities had become unsurvivable. Nationalist police power, informer networks, and the destruction of the unions after 1927 made urban organisation a matter of months between arrest and execution, while the hill districts of the provincial borders offered what the cities could not: distance from state power, defensible ground, and a peasantry with grievances the party could organise.
The rural turn should not be mistaken for a doctrinal preference arrived at in a library. It was imposed by defeat, and for years it was contested inside the party itself, whose Comintern-instructed leadership repeatedly ordered the base areas to attack cities they could not hold and regarded the rural soviets as a temporary expedient pending the recovery of the urban movement. Mao’s contrary argument, set out in his writings of the period, was that the countryside was not a waiting room: in a country where state power was unevenly distributed, strong in the cities and along the railways and weak in the interior, a revolutionary force could build durable power in the gaps, accumulate strength over years, and in the end encircle the cities from the countryside. The argument fitted the facts of Chinese geography and administration as the urban line never had, but it carried a theoretical price that the party paid knowingly. A strategy of protracted rural war required an army recruited from peasants, supplied by peasants, and fighting for objectives peasants could be brought to support, and it subordinated every other element of the classical Marxist programme, the factory, the strike, the insurrection of the industrial working class, to the needs of that army and its bases. The party that resulted was Marxist in its doctrine, its organisation, and its ultimate aims, and peasant in its composition and its daily practice. Holding those two halves of the description together, without letting either erase the other, is the discipline this article maintains, because the scholarly dispute about the revolution is, at bottom, a dispute about which half of the description explains the victory.
The soviets in detail: land revolution and its costs
The Jiangxi Soviet’s internal order repays a closer look, because the classification machinery and the radicalism of those years became the template, and the warning, for everything the party later did in the countryside. Land revolution in the soviet districts proceeded by investigation and classification: every household was assigned a class status, landlord, rich peasant, middle peasant, poor peasant, or hired labourer, according to its landholding, its hiring of labour, and its income from rent and usury, and the land of the first categories was confiscated and redistributed by the village soviets in mass meetings. The system created, for the first time in the Chinese countryside, a politics in which a family’s fate turned on an administrative classification decided by its neighbours under party direction, and the line between the categories was neither self-evident nor stable: the definition of a rich peasant shifted with directives, households manoeuvred to escape the fatal labels, and the party’s own corrections of leftist excess, as it called the overreach of its zealots, alternated with campaigns against the rightist error of sparing the enemy. The soviet also reproduced, inside the movement, the violence of the world it was fighting. In 1930 a mutiny and purge in the Jiangxi forces, the episode remembered as the Futian incident, saw the party turn its security apparatus on its own ranks in a hunt for alleged enemy agents, and the pattern, accusation, confession, execution, recurred in the soviet period’s internal campaigns. These passages belong in the article because they establish a continuity the later sections need: the organisational explanation of the wartime victory, with its classification, surveillance, and coercion, was not invented in the war against Japan. It was inherited from the soviet period, moderated during the united front years when the alliance required moderation, and available in full whenever the party’s situation again permitted or demanded it. The wartime base area, with its reduced rents and its three-thirds governments, and the Jiangxi soviet, with its confiscations and its purges, are two settings of a single instrument, and the historian who understands the instrument’s range understands the party better than the historian who knows only one setting.
How did classification work in the land revolution?
Investigators and village meetings fixed each household’s class label from the land it held, the labour it employed, and the dues it drew from tenants and debtors, and the label decided confiscation, redistribution, and political rights. The boundaries were contestable and shifted with directives, so classification became the village’s political act and the party’s deepest instrument of knowledge and control.
The Jiangxi experiment also carried a strategic lesson that the party absorbed at great cost and never forgot. The soviet had attempted to hold fixed territory against a modernising state army, defending a capital, a frontier, and a conventional front, and it had been strangled by blockhouses in consequence. The doctrine that emerged from the defeat, elaborated in Mao’s military writings of the later 1930s, inverted the priority: territory was expendable, the army and the population were not; the enemy was to be drawn in deep, worn down by guerrilla action across a countryside the party organised, and destroyed in mobile battles at moments of chosen concentration. The Long March itself was the doctrine’s first terrible application, the abandonment of the base to save the force. When the war with Japan came, the party therefore expanded without repeating Jiangxi’s error: its base areas behind Japanese lines were instruments, not idols, to be contracted when the enemy’s offensives required and re-expanded when they passed, and the leadership’s willingness to trade ground for survival, learned in the loss of its first republic, distinguishes the wartime movement from every Chinese rebellion that had died defending its holy city. The point matters for the weighing of the explanations, because it identifies a component of the victory that none of the three factors quite names: accumulated strategic learning, purchased in defeat and stored in doctrine, a form of organisational memory that the party’s rivals, whose defeats taught them nothing because their organisations could not retain the lesson, conspicuously lacked.
The Long March and the remaking of the leadership
The Jiangxi Soviet did not fall to internal failure; it was crushed by methodical pressure. Chiang Kai-shek, having subdued or bought off rival militarists, directed a series of encirclement campaigns against the base areas, and the fifth of them, conducted with blockhouse lines that advanced slowly and strangled the soviet’s economy, succeeded where the earlier drives had failed. In October 1934 the main Red Army force broke out of Jiangxi and began the retreat that entered history as the Long March: a fighting withdrawal across the length of western China, over snow mountains and through marshlands, harassed by Nationalist armies and provincial forces, which ended a year later when the survivors reached the loess country of northern Shaanxi and made their headquarters at Yan’an. Other Red Army groupings converged on the same region by 1936. The army that arrived was a fraction of the one that had set out, and the losses were not only numerical. The march functioned as a selection mechanism of terrible efficiency. Cadres who could not endure it fell away; commanders who had failed were displaced; and at the Zunyi Conference, held in January 1935 in a captured town in Guizhou, the leadership was remade, with Mao Zedong rising to effective direction of the army and the party after years in which Comintern-backed rivals had overruled him. The biography of that ascent, and the intellectual formation behind it, belongs to the study of Mao Zedong’s intellectual biography, and is linked rather than retold here. What belongs to this article is the march’s structural consequence: the party that settled at Yan’an was smaller, harder, more cohesive, and more completely identified with a single strategic doctrine, the protracted rural war, than the party that had left Jiangxi, and it carried the march ever afterward as its founding ordeal, the proof in its own mythology that it could survive anything.
What did the Long March decide besides survival?
It decided the leadership and the doctrine. At the Zunyi Conference of January 1935 the Comintern-backed leadership that had lost Jiangxi was displaced, Mao Zedong rose to effective command, and the strategy of protracted rural war, building base areas and encircling the cities from the countryside, became the settled line of the movement that reached Yan’an.
Yan’an itself requires a word of description, because the name does double duty in the scholarship, standing both for a place and for a model of rule. The place was a small town in the poor loess hills of northern Shaanxi, capital of the Shen-Gan-Ning Border Region, the one base area that survived from the soviet period into the war years. The region was remote, agriculturally marginal, and lightly held by the Nationalist state, which blockaded it but could not easily invade it. Into this backwater the party brought its central apparatus, its schools, its printing presses, and, as the war developed, tens of thousands of young patriotic students and intellectuals who made their way there from the occupied and Nationalist-held cities, drawn by the party’s reputation as the most determined force of resistance to Japan. Yan’an in the war years was thus two things at once: the headquarters of a rural insurgency, and a magnet for the urban educated youth whom the party had lost in 1927 and now recovered in a new form, as administrators, teachers, propagandists, and officers. That double character, a peasant army governed from a town full of students, is essential background for the historiographical dispute, because the second of the three explanations, the account of the Yan’an Way, is precisely an account of what the party did in and around this capital: how it governed, how it taxed, how it reduced rents without confiscating land, and how it converted a poor border region into the showpiece by which millions of Chinese who never saw it judged the movement. Whether the showpiece explains the victory, or merely advertised it, is one of the questions the dispute exists to answer.
The other Long March: Zhang Guotao and the road not taken
The march that reached Yan’an was not the only one, and the party’s survival required the defeat of an alternative leadership as well as of the Nationalists. The Fourth Front Army, commanded by Zhang Guotao, a founder of the party and the most senior of Mao’s rivals, was the strongest of the communist forces when the two columns met in western Sichuan in the summer of 1935, and Zhang’s price for unity was a share of command that the Zunyi settlement had not contemplated. The dispute over strategy, whether to strike north toward the Soviet border or to consolidate in the west, became a dispute over the party’s crown: Zhang established a rival central committee, led his army south in a campaign that cost it half its strength against the Sichuan militarists, and only rejoined the northern forces in 1936 under the pressure of defeat and of the Comintern’s insistence on unity. A detachment sent westward into the Gansu corridor, the Western Route Army, was annihilated there by the cavalry of the Ma family militarists, a loss the party’s later historiography laid at the door of Zhang’s line. Zhang himself, assigned honourable and empty office at Yan’an, defected to the Nationalists in 1938, the most senior apostate in the party’s history. The episode matters beyond its drama. It demonstrates that the unity the rectification later codified was not the movement’s natural condition but an achievement, won against a rival who commanded more soldiers and an older pedigree, and won by the methods this article has tracked throughout: the patient accumulation of legitimacy, the appeal to the international centre’s arbitration, and the retrospective writing of the history, in which Zhang’s substantial record as an organiser of the early movement was reduced to a cautionary tale. The party that took China was a party that had first taken itself, and the taking was neither gentle nor foreordained.
The Long March as memory: the ordeal and its uses
No sooner had the march ended than its narration began, and the narration is part of the history, because the Long March as the world knows it was composed, in the first instance, at Yan’an, for audiences the party needed to persuade. In 1936 the American journalist Edgar Snow reached the border region and recorded long interviews with Mao and other leaders, published the following year as Red Star Over China, which carried the march’s epic, the river crossings, the snow mountains, the daily battles of a retreating army that refused to dissolve, to readers across the world and across China itself, where the book circulated in translation among exactly the students the party hoped to draw north. The march in Snow’s telling was a feat of endurance without parallel, and the telling was not false; but it was a selection, and the party’s own historiography completed the selection over the following decade, fixing the march as the movement’s passion and resurrection, the ordeal that proved its election. The historian’s task with such a source is double and should be performed openly. The march’s facts, the route, the losses, the Zunyi Conference, the arrival, are recoverable from records independent of the epic and are stated in this article from those records. The march’s meaning, its function as the party’s charter of legitimacy, is a separate fact of equal reality: the generation that ruled China after 1949 was the generation the march had selected and bonded, its survivors staffed the state to its summit for decades, and the claim to rule of the men of the Long March rested, in their own presentation, on the ordeal as a title deed. To note the function is not to debunk the ordeal, which needed no invention to be terrible. It is to observe that the party understood, earlier than any of its rivals, that a revolution must narrate itself to survive itself, and that the narration, in this case, was among the most successful political texts of the century, recruiting the students of the 1930s, instructing the cadres of the 1940s, and framing the victory of 1949, for foreign and Chinese readers alike, as the conclusion of a story they already knew.
The march through the minority regions: an early lesson in coalition
The Long March’s route carried the Red Army through territories inhabited by peoples whom the Chinese state had governed, when it governed them at all, through their own chiefs and by intermittent coercion, and the army’s conduct there became one of the foundation stories of the party’s later nationalities policy. The most famous episode belongs to the passage through the Yi country of Sichuan in 1935, where the advance column’s commander, Liu Bocheng, concluded a sworn brotherhood with a Yi chieftain, Xiaoyedan, whose clan then guided the army through territory where other forces had been stripped and destroyed by the peoples of the mountains. The episode was exceptional and was meant to be remembered as exceptional: an army that paid for its supplies, punished its own stragglers for theft from minority villages, and negotiated passage as a supplicant rather than a conqueror, in a region where the Nationalist and warlord forces had ruled by punitive expedition. The party drew the general lesson into its programme, promising the minority peoples autonomy within a united China and enshrining the promise in the constitutions of its successive states. The historian should register both the episode and its limits. The alliance of the march was an instrument of survival adopted under extreme necessity, like much of the party’s coalition practice, and the autonomy later practised was autonomy within a centralised party-state whose leading organs the minorities did not control. But the contrast the episode established, between a revolution that negotiated with the peoples on its path and states that had merely raided them, formed part of the party’s claim to embody the nation more completely than the government that claimed the nation’s name, and the claim was heard. In the competition for China, the communists presented themselves, with increasing plausibility as the war proceeded, as the only force that proposed to include the whole of the country’s population, Han and minority, peasant and worker, in a single political order, and the inclusiveness of the proposal, whatever qualifications the sequel attached to it, was itself a factor in the victory that the three explanations, each focused on the Han village, tend to leave uncounted.
The Xi’an Incident: the hinge on which the war turned
Between the arrival at Yan’an and the outbreak of full-scale war stands an episode so improbable that narrative histories linger on it, and it earns its place here because it determined the political conditions under which the expansion occurred. Through 1936 Chiang Kai-shek had pressed his campaign to exterminate the Shaanxi communists before turning to Japan, assigning the task in part to the Northeastern Army of Zhang Xueliang, the former warlord of Manchuria whose troops, exiled from their homeland by the Japanese seizure of 1931, had little appetite for killing Chinese communists while their own provinces remained occupied. The communists, for their part, had adopted under Comintern influence the line of a broad anti-Japanese front and directed their propaganda at the Northeastern Army’s homesickness, with a success that fraternisation across the lines made visible. In December 1936 Chiang flew to Xi’an to compel a final offensive, and Zhang, with the commander Yang Hucheng, detained him, demanding a united resistance to Japan and the end of the civil war. The communists, invited to the negotiations that followed, faced a choice their doctrine had not prepared: Chiang in captivity could be killed, as many of their cadres demanded in memory of 1927, or released as the indispensable leader of a national front. Zhou Enlai, negotiating for the party, chose the front, a decision in which Soviet preference for a Chinese government capable of tying down Japan weighed heavily, and Chiang was released on Christmas Day with an unwritten undertaking to end the civil war and resist. He honoured it sufficiently. The incident’s significance for this article is structural rather than dramatic. It converted the communists from hunted rebels into prospective partners of the national government at the exact moment when invasion made partnership valuable, it demonstrated to every wavering force in China that the party could act with a statesmanship its enemies denied it possessed, and it set the terms, an armed party inside a patriotic alliance, under which the eight years of expansion were conducted. Without Xi’an there is no second united front in the form it took, and without that form the three explanations of the victory lose their common stage.
Why did the communists help to release the general who had tried to destroy them?
Because a dead or captive Chiang promised them nothing they could use: his death would have fragmented the Nationalist camp, invited the Japanese advance it was their strategy to resist, and branded the party as the obstacle to national unity. Released, he headed a government committed, however reluctantly, to the alliance under which the party could legally expand.
The war with Japan and the second united front
The full-scale war began in July 1937, after the clash at the Marco Polo Bridge outside Beijing escalated into a Japanese invasion of China proper. The ground for cooperation had been prepared the previous December, when Chiang Kai-shek, detained at Xi’an by his own subordinate commanders who demanded resistance to Japan over civil war, was released after negotiations in which the communists, following the Comintern’s new line of broad anti-fascist alliance, played a mediating part. The second united front followed in 1937. Its terms transformed the communists’ position. The Red Army was redesignated as Nationalist forces, the Eighth Route Army in the north and the New Fourth Army in the central river valleys, while remaining under communist command in fact. The soviets were renamed as border region governments within the republic. The land revolution was suspended: confiscation gave way to the milder policy of reducing rents and interest, which rewarded tenants without dispossessing landlords, many of whom the party now wished to keep inside the patriotic alliance. In exchange, the party gained a recognised place in the national resistance and the freedom to operate, behind Japanese lines, in the vast countryside of north and central China where the Nationalist state had been swept away and the Japanese army, holding cities and railway lines, could not administer the villages. It was into that vacuum that the party expanded. Its organisers followed its guerrilla units into village after village, building governments, associations, militias, and tax systems where none existed, and over eight years of war the party’s membership grew by more than an order of magnitude, from a remnant headquartered in one border region to a movement governing base areas across the north and centre of the country. That expansion, its scale, its speed, and above all its causes, are the central fact the three explanations compete to explain.
How did the second united front differ from the first?
In the first united front the communists had entered the Nationalist Party as individuals and possessed no army or territory of their own, so the alliance’s collapse in 1927 destroyed them. In the second, from 1937, they kept their own army, their own governments, and their own territory inside the alliance, and expanded behind Japanese lines while it held.
The contrast between the two united fronts is the clearest evidence that the party of 1937 had learned the lesson of 1927 in organisational form. The earlier alliance had been a merger in which the weaker party dissolved its independence and was destroyed when the stronger partner turned. The later alliance was a coalition between two armed states in embryo, each with its own territory and forces, cooperating against a common invader while competing, sometimes violently, for the countryside between them. Friction was constant: the Nationalist blockade of the border region, clashes between the two armies in the contested zones, and in January 1941 a Nationalist attack on a column of the New Fourth Army in the south, an episode the communists commemorated as proof of their ally’s bad faith. Yet the front was never formally abandoned by either side before the Japanese surrender, because each needed it: the Nationalists needed the appearance of national unity and the communist divisions tying down Japanese troops, and the communists needed the legal and political shelter under which their base areas grew. Readers should hold the structure firmly in mind, because it disciplines the historiography. When the first explanation says the party grew by organising nationalism, it means nationalism within a front that legitimised the party’s presence. When the second says it grew by governing well, it means governing inside enclaves the front protected. When the third says it grew by organisation and coercion, it means organisation carried into villages under cover of a war the party did not have to fight alone. None of the three explanations makes sense without the front as their common condition.
The bean ballot: elections in the base areas
The base area elections deserve a closer description than the word election usually receives in accounts of one-party movements, because their mechanics show the system’s character in miniature. The franchise in the border region was, on paper, the broadest in China: every resident above the qualifying age, without distinction of sex, property, or literacy, held the vote, at a time when the Nationalist republic’s own electoral practice was indirect, restricted, and largely notional below the national level. The mechanics were adapted to an illiterate electorate with an ingenuity that became famous: in the method remembered as bean voting, the voter dropped a bean into the bowl of the candidate of his choice, the bowls screened from observation, so that the count was public and the choice was secret. Assemblies at village and district level debated candidates, and the record contains genuine instances of officials rejected by their electors and of the party’s nominees withdrawn after hostile meetings. The limits were equally structural. The party determined the framework within which choice operated: the three-thirds rule allocated the outcome’s proportions in advance, the associations through which candidacies were canvassed were the party’s own, and no organised alternative to the movement’s rule could present itself for election, so that the voter chose among persons within a settlement he had no means of voting against. The honest description is neither the democracy the border region’s publicity claimed nor the fiction its enemies alleged. It was a real practice of consultation and selection inside a monopoly, which gave the villager a voice in the choice of the men who would tax and judge him, and denied him a voice in the choice of the power that appointed the taxers’ masters. Both halves of that description belong to the second explanation’s evidence and to the third’s, and the bean in the bowl, at once a genuine ballot and a bounded one, is as good an emblem as the period affords of the system that cast it.
The students’ road to Yan’an: the patriotic generation
The party’s recovery of the educated youth it had lost in 1927 was one of the war’s quieter transformations, and it began before the war did. In December 1935, students in Beijing, communist-influenced and nationalist in temper, marched to demand resistance to Japanese encroachment in the north, and the December Ninth movement carried the demand through the universities of the country, creating a generation for whom patriotism and the demand for a government worthy of it were the same emotion. When the war came and the universities of the coast evacuated inland, a stream of these students made a different journey, slipping through the lines to Yan’an, where the party received them with an apparatus built for the purpose: the schools of the border region, Kangda first among them, which gave the new arrivals short courses in the doctrine, the discipline, and the practical work of the movement, and sent them out to the base areas as the cadres described in a later section. The migration mattered out of proportion to its numbers, which were in the tens of thousands across the war years. It reversed the social geometry of 1927. The party that had been expelled from the cities returned to them in the persons of their own children, educated, patriotic, and persuaded that the movement in the loess hills was the nation’s most serious enterprise; and the Nationalist government, which conscripted the same generation’s peasants and policed its students, watched its natural constituency drain away to a rival it had blockaded. For the historiography, the students’ road is evidence that cuts across the three explanations. These were not peasants moved by rent or invasion in their villages; they were the urban educated, moved by a nationalism the party had organised and by the reputation of a government they had read about, a reputation the second explanation’s showpiece had created and Snow’s book had carried to them. The factors of the settlement operated on China’s students as they operated on its villages, through different mixtures, and a dispute that confines itself to the peasantry misses the constituency that staffed the revolution.
Yan’an and the base areas: governing behind enemy lines
What the party built in the base areas was a state in miniature, and its design is recoverable in detail from the administrative record. The border region governments were constituted, from 1941, on the three-thirds system, under which communists were to hold one third of government posts, with the remaining two thirds divided between non-party progressives and the so-called middle elements, landlords and merchants included, a deliberate display of coalition in a period when the party sought the widest possible patriotic alliance. Village and county assemblies were elected, on franchises broader than most of China had known, though the party’s influence over candidacies and agendas was pervasive. The economic programme was calibrated to the alliance: rents were reduced by a standard fraction and interest on loans capped, but titles to land were left untouched, so that a landlord lost income without losing his property or his place in the village, while his tenants gained measurably in every harvest. Peasants were organised into associations, women into associations of their own, and producers into cooperatives for spinning, transport, and mutual aid in the fields. Taxation, levied in grain, was regularised and, by the standards of the warlord and Nationalist practice the peasantry had known, assessed with unusual publicity and restraint. Around these arrangements stood the less visible machinery: the party cell in the village, the militia, the security organs, and the mass campaigns through which policy was driven downward and information gathered upward. The whole apparatus amounted to a claim about how China should be governed, made in administrative form. Where the Nationalist state had reached the village mainly as a tax demand and a press gang, the communist base area reached it as a government that organised production, settled disputes, ran schools, and asked for support in the currency of services rendered, while never ceasing to be an armed organisation preparing for the civil war its leaders expected to follow the national one.
What was the three-thirds system in the base areas?
It was the rule, applied in the border region governments from 1941, that communists should hold only one third of official posts, the rest going to non-party progressives and to middle elements such as landlords and merchants. The system advertised coalition, reassured the propertied allies the united front required, and left the party’s real influence intact.
Two features of this system deserve emphasis before the dispute is entered, because each of the rival explanations seizes on one of them and passes over the other. The first is its moderation. The party that had confiscated land in Jiangxi now protected titles, courted landlords, and presented itself as the administrator of a patriotic coalition rather than the agent of class war. Critics of the movement have treated the moderation as camouflage, and defenders as proof of a genuine new democratic programme; the historiographical question is narrower and more useful, namely what the moderation did. It plainly widened the party’s appeal beyond the poor peasantry to the middle peasants, the merchants, and even the gentry who feared Japanese exactions and Nationalist corruption more than they feared communist discipline, and it made the base areas governable without a permanent terror against the propertied. The second feature is its penetration. The base area state did not merely tax and judge; it organised. The peasant who joined the association, the woman who joined the spinning cooperative, the youth who joined the militia, each acquired a relationship to the party that was also a channel through which the party knew them, counted them, and could mobilise them. Sympathy and organisation were not alternative fuels for the same engine; the system was designed so that each produced the other, and it is precisely this fusion that makes the later dispute about causes so difficult to settle, since the nationalism, the reform, and the organisation arrived in the village in the same persons, wearing the same clothes, carrying the same programme.
The New Fourth Army incident: the front’s substance ends
The united front’s pretence of partnership did not survive its third year, and the episode that ended it deserves record because it fixed the terms on which the two parties thereafter coexisted, armed neutrality within a nominal alliance. The New Fourth Army, the communist force operating in the Yangzi valley under the front’s arrangements, had grown into a power in central China that the Nationalist command resolved to confine south of the river, and in late 1940 it was ordered to move north across the Yangzi by a prescribed route and date. A column of its headquarters troops, moving in January 1941 through southern Anhui in compliance or in delay, the record of intention is itself contested, was ambushed and destroyed by Nationalist forces many times its strength; its commander was captured and its political commissar killed. The communists named the episode the Southern Anhui Incident and made of it the centrepiece of a propaganda campaign that ran through the Chinese and foreign press: the party that was fighting Japan had been attacked, at the height of the war, by the government that claimed to lead the resistance, and the moral capital of the national cause passed, in the eyes of much of the educated public, measurably toward Yan’an. Chiang, under domestic and foreign pressure, disavowed the ambush as a disciplinary matter and let the matter rest without the civil war the incident might have licensed. The party rebuilt the New Fourth Army under new command, larger than before, and treated the front thereafter as what it had become, a diplomatic form covering a competition for territory that both sides expected to settle after Japan’s defeat. For the historiography the incident is a fixed point of unusual clarity. It demonstrates that the expansion the three explanations compete to explain proceeded, from 1941, in the teeth of the senior partner’s hostility as well as the invader’s, and it supplies the nationalism thesis with its sharpest complication: the party’s patriotic credit rose highest not when it fought the Japanese but when it was seen to suffer, resisting moderately and protesting loudly, at the hands of the Chinese government itself.
The storm years: the great offensive and the contraction that followed
The expansion was not a smooth ascent, and its interruption is as informative as its course. In the second half of 1940 the communists launched their largest operation of the war, the Hundred Regiments Offensive, a coordinated assault by regular forces on the railways and strongpoints of the north China occupation. The offensive demonstrated a strength the party had previously concealed, from its enemies and to a degree from itself, and the Japanese response reordered the war in the north. Through 1941 and 1942 the occupation conducted its counterinsurgency at maximum intensity, the campaigns of devastation remembered in the party’s vocabulary as the Three Alls, and combined them with blockade, fortified lines, and the systematic destruction of the villages that sustained the guerrillas. The base areas contracted sharply; populations under communist government fell away as districts were lost; the party’s forces, short of food, ammunition, and medicine, dispersed into smaller units and dug into the villages for survival; and Chongqing’s blockade of the border region, tightened after the destruction of the New Fourth Army column in January 1941 ended what trust the front retained, cut the Yan’an government off from its subsidies and much of its trade. These were the years in which the movement came nearest to being reduced a second time, and they perform a service in the historiography out of proportion to their place in popular memory, because they discriminate among the explanations. Nationalism, at its maximum provocation, did not prevent the contraction: patriotic feeling could not hold districts against methodical devastation. Reform, in the border region, was curtailed by blockade economics to the production campaigns described in a later section. What carried the movement through was the third factor in its purest form, organisation sunk so deeply into the village, in cells, militias, tunnels, and intelligence networks, that the occupation could burn the surface and not reach the root, together with a strategic patience that refused decisive battle and waited for the correlation of forces to change. When expansion resumed from 1943, as Japanese strength was drawn off to the Pacific war and the puppet administrations decayed, it resumed on the organisational foundations the storm years had tested. Any weighing of the three factors that ignores 1941 and 1942 is weighing only the fair weather.
What did the contraction of the storm years prove about the base areas?
That their survival did not rest on popularity or patriotism alone. Under maximum counterinsurgency the districts held where organisation had been sunk deepest into the villages, in cells, militias, and intelligence networks, and were lost where it had not, a discrimination among the explanations that the years of easy expansion could not provide.
Discipline and its shadow: the rectification movement
No account of the base areas is complete without the machinery by which the party disciplined itself, and that machinery matters here because the third of the three explanations rests on it. Between 1942 and 1944 the party conducted the rectification movement at Yan’an, a campaign of study and criticism directed first at the party’s intellectuals and cadres. Its stated objects were habits of work: subjectivism, the mechanical copying of texts and foreign models; sectarianism, the defence of one’s own bureau or faction; and stereotyped party writing, the substitution of formulae for investigation. Cadres studied a canon of documents, wrote self-examinations, and criticised one another in small groups, and the movement standardised the party’s language and its loyalty around the leadership that had emerged from the Long March and around Mao’s claim, increasingly explicit, that Marxism had to be sinified, made Chinese in form and content, rather than imported as a finished doctrine from Moscow. The movement also had a coercive underside, which the record does not permit a honest account to omit. In its later phase, a campaign to rescue those who had allegedly erred subjected cadres, including young volunteers who had come to Yan’an out of patriotism, to interrogation and pressure to confess to espionage that in most cases had not occurred; the excesses were halted and apologies issued, but the episode demonstrated, inside the party itself, the readiness of its organisational methods to pass from persuasion into compulsion. Historians weigh the rectification differently according to their explanation of the victory. For the second explanation it is an embarrassment to be acknowledged and passed over; for the third it is a central exhibit, the moment when the party forged the disciplined instrument that could hold a hundred base areas to a single line. Both readings can cite the same documents, which is a pattern the reader will meet again.
What was the rectification movement in Yan’an?
A campaign of 1942 to 1944 in which the party subjected its cadres to intensive study, self-examination, and mutual criticism, standardising doctrine around the sinification of Marxism and around the leadership formed on the Long March. In its later phase it passed into interrogation and forced confession before the excesses were halted, displaying the coercive underside of party discipline.
The technique the party called the mass line belongs in the same discussion, since it was the bridge between the discipline of the cadre and the government of the village. In its formal statement, the mass line required the organiser to gather the scattered views of the population, concentrate them into a policy tested against doctrine, and return the policy to the population as its own, repeating the cycle as implementation revealed new conditions. As a theory of leadership it was genuinely distinctive: it cast the peasant neither as the passive material of revolution, the role the classical texts had assigned, nor as its spontaneous hero, but as a partner whose experience the party was obliged to consult and entitled to overrule. As a practice it cut both ways, and the scholarship is right to insist on both. It produced administrators who knew their villages intimately and policies fitted to local conditions in a way no distant ministry could match; and it supplied a vocabulary in which the party’s decisions, however arrived at, could be presented as the people’s own conclusions, a legitimation that cost the leadership nothing it valued. When the third explanation speaks of organisation as the decisive factor, the mass line is part of what it means: not merely cells and files and security organs, but a method that converted consultation into mobilisation and mobilisation into knowledge of the governed, a governing technology as much as an ideology.
Rent and interest reduction in practice: the moderate programme at village level
The wartime programme’s moderation can be made precise, and precision here serves the dispute, since the first explanation rests part of its case on how little the programme offered. Rent reduction, as standardised in the border region and applied with variations across the base areas, required the reduction of existing rents by a quarter from their pre-war level and capped the rent thereafter at a fixed share of the harvest, while interest on loans was capped at a rate that ended the compounding by which village debt had traditionally become hereditary. Titles were untouched: the landlord remained the owner, received his reduced rent through the association’s enforcement rather than his own bailiffs, and was himself protected against the tenant’s default by the same government that had cut his income. The programme’s genius, and the source of the scholarly argument about it, lay in its double address. To the tenant it delivered a measurable improvement in every harvest, delivered not as charity but as law, by an organisation in which he held membership and could complain. To the landlord it delivered a smaller income and a larger security, in districts where the alternative authorities offered him neither, and it preserved his place in the village order at the price of his dominance within it. The party’s cadres enforced the settlement against both directions of violation, against landlords who evaded the reduction and against poor peasants who pressed, in the first enthusiasm, toward the confiscation the united front forbade, a discipline that cost the movement some of its poorest supporters’ ardour and bought it the acquiescence of the village’s upper half. Johnson read the moderation as evidence that social reform could not explain the growth, since so moderate a programme could not purchase a revolution. Selden read the same facts as the reform itself, a redistribution of income and dignity within the village that peasants experienced as the substance of a new order. The archival scholarship added the third reading: the reduction was administered, and administration is the operative word, by an apparatus that registered every tenancy and every loan in the village, so that the programme’s moderation and the organisation’s penetration advanced together, each the other’s instrument. All three readings are of the same ledger book, and the ledger book survives, which is why the dispute about it will not die.
The cultural army: propaganda teams, theatre, and the wall newspaper
The party’s presence in the village had a cultural arm whose work is easily underestimated by readers who meet propaganda as a word for lies. The base areas were served by troupes and teams of every scale, professional companies at Yan’an performing opera and drama in renovated forms, and at the district level the small propaganda team, a dozen young people with drums, gongs, a portable stage, and a repertoire of skits, songs, and living newspapers that enacted the village’s own affairs back to it: the landlord who had evaded the rent settlement, the recruit whose family the association had helped, the spy unmasked. The yangge, the rice-sprout song dance of the northern countryside, was taken up and refitted with new words until the festival procession itself carried the movement’s messages into the lanes. Wall newspapers appeared at the village threshing ground, handwritten sheets in large characters that reported production figures, praised model households, and named laggards, legible to the newly literate and read aloud to the rest. The cultural work performed functions the explanations divide among themselves. It was the nationalist factor in performed form, staging the invasion and the resistance for audiences who had seen the war only as rumour. It was the reform factor’s advertisement and its school, teaching the new rules by dramatising their application. And it was the organisation’s softest instrument of surveillance, since the team that performed in a village also learned it, and the wall newspaper that praised also recorded. The point for the weighing is the same one the school and the cooperative taught: in the base areas the party’s appeal was never a message delivered to a passive audience but a set of activities in which the audience was enrolled as performer, pupil, and subject at once, and the enrolment, repeated at every festival and market day, built the familiarity on which the harder demands, the tax, the recruit, the classification, could later draw. Movements are believed, in the end, by the people who have sung in them.
The Yan’an Forum: literature conscripted
The rectification’s cultural front deserves a section of its own, because it shows the organisational method applied to the realm where the second explanation’s admirers most expected freedom, the life of the mind. Yan’an in the early war years had enjoyed a relative cultural spring: writers, artists, and musicians drawn from the cities produced work in the border region’s relative safety, and a critical literature circulated that examined the new society’s own shadows, its hierarchies, its privileges, its treatment of women. In May 1942 the party convened the Forum on Literature and Art at Yan’an, and Mao’s talks there, published afterward as a canonical text, settled the terms on which culture would be permitted: literature and art were to serve the workers, peasants, and soldiers who were the revolution’s constituency, to be judged by their political effect before their artistic merit, and to take their standpoint from the party’s. The writers who had practised criticism of the movement itself were corrected in the rectification’s machinery; the essayist Ding Ling, who had written with unusual frankness about the position of women in Yan’an, was criticised and withdrew her line; the writer Wang Shiwei, whose essays had attacked privilege and demanded equality within the revolutionary camp itself, was denounced as an enemy agent in a campaign of mass meetings, imprisoned, and in 1947, during the evacuation of Yan’an, executed, a sequence the party itself later acknowledged as a wrongful case when his name was formally cleared decades afterward. The episode belongs in this article for the same reason as the rectification’s coercive phase: it is the third explanation’s evidence at the level of culture, the demonstration that the party’s organisation of society extended to the organisation of imagination, and that the showpiece government’s tolerance had a boundary at the point where criticism touched the legitimacy of the organisers. The foreign observers who praised Yan’an’s cultural vitality in 1944 were not inventing what they saw; the spring had been real. They arrived after the conscription that ended it, and the distinction between the Yan’an of the spring and the Yan’an of the forum is one more instance of the article’s recurring finding, that the base areas must be periodised as carefully as they are weighed, since the same name covers a society that changed its character under the pressure of war and discipline.
The party as its own historian: the official Resolution
Among the rectification movement’s products, one deserves separate notice because it shaped every later telling of the history this article recounts, including the tellings the scholarship had to unlearn. In April 1945, on the eve of its Seventh Congress, the party adopted the Resolution on Certain Questions in the History of Our Party, a document drafted under Mao’s direction that rendered the party’s official verdict on its own first quarter century. The Resolution arranged the past as a struggle between the correct line, identified with Mao’s, and a succession of erroneous lines, the right opportunism of 1927, the putschism that followed, and above all the dogmatism of the Comintern’s returned students in the early 1930s, to whose account the loss of Jiangxi and the disasters of the Long March’s outset were assigned. As history the document is a victor’s brief, and historians use it as such, a primary source for the leadership’s self-understanding rather than a reliable account of the disputes it settles. As politics it was a masterpiece of the organisational method: it gave the unified party a unified past, it disposed of rival claims to authority by disposing of their historical records, and it installed the interpretation, that the party wins when it follows the sinified line and loses when it copies foreign models, as the lens through which members were thereafter required to view their own experience. The significance for the reader of this article is a caution that operates in both directions. Much of what the world first learned of the revolution came through the frame the Resolution built, transmitted by the party’s historians and its foreign sympathisers; and much of the later scholarship, Johnson’s and Selden’s included, defined itself against that frame, which is one reason the dispute took the polar form it did, each new explanation announcing itself as the correction of the official story. The settlement defended in this article is written against that frame’s method as much as within its shadow: the past of this revolution was not a struggle between a correct line and its deviations, but a collision of a doctrine, a nation, and a peasantry, in which the men who won were as often improvising as applying, and in which the errors and the insights were distributed on all sides.
What did the party’s official Resolution do to its own history?
It rewrote it as the struggle between a correct line, identified with Mao, and a sequence of erroneous lines, assigning the catastrophes of 1927 and of Jiangxi to deviation and the survivals to fidelity. Adopted as the party unified for its final campaigns, it installed the official interpretation through which members, and many outsiders, long viewed the revolution.
The Nationalists’ war: the other half of the explanation
Every explanation of why the communists won contains, implicitly, an explanation of why the Nationalists lost, and the second half of the question deserves its own space, because the Nationalist government was not a hollow shell waiting to be pushed over. It was the recognised government of China, it commanded the country’s largest armies and its modern sector, and it bore the main weight of conventional resistance to Japan in the war’s first years, fighting the great battles for Shanghai, Wuhan, and Changsha while the communists waged guerrilla war in the north. The costs of that conventional war fell on a state whose fiscal foundations had always been shallow. The Nationalist government had never effectively taxed the countryside it claimed to rule; its revenues came disproportionately from the maritime customs and the cities of the coast, precisely the assets the Japanese invasion seized. Driven inland to Chongqing, it financed the war increasingly by printing money, and the inflation that began as an expedient became, across eight years, a solvent of the social order: officials and soldiers paid in depreciating currency turned to speculation and exaction, the salaried classes were ruined, and the state’s promises lost credit with the population that was asked to redeem them. Conscription, conducted through quotas imposed on villages and enforced by press gangs and the chaining of recruits, made the central government hated in the countryside it seldom otherwise visited, while the grain tax in kind, levied from 1941, fell on peasants already stripped by requisition. The historian Lloyd Eastman, in studies that remain standard, described the result as a state rotting from within during the very years of its supreme exertion: factionalised at the top, corrupt in the middle, and predatory at the base. None of this made communist victory inevitable. It made the Nationalist state a competitor for the peasantry’s allegiance that arrived, in most villages, carrying demands and offering little, against a rival that arrived carrying organisation and offering order, and the three explanations of the communist victory all presuppose that asymmetry even when they do not state it.
Why did the Nationalist state weaken during the war it was leading?
Its finances rested on the coastal cities and customs revenues that the invasion seized, so the inland government at Chongqing financed resistance increasingly by printing money, and inflation corroded salaries, discipline, and public credit. Conscription by quota and the grain tax made the state hated in villages it otherwise barely reached, while factionalism paralysed its summit.
Fairness to the Nationalist record requires the complementary point, which the better scholarship insists on and which any honest weighing must carry. The Nationalists’ conventional armies absorbed the main force of the Japanese offensives for most of the war, and the strategy of trading space for time, retreating into the vast interior and refusing surrender, denied Japan the political settlement it needed and kept China in the war as a recognised belligerent. The communists’ guerrilla expansion, for all its scale, took place in a theatre the Japanese treated as secondary, against garrison forces and puppet troops rather than the field armies engaged in the south and centre. When the first explanation attributes communist growth to the nationalist response to occupation, it quietly assumes a division of labour in which the Nationalists held the front of the war and the communists organised its rear; and when the civil war came, the Nationalist armies that faced the communists were the survivors of eight years of that attrition, commanded by generals whose forces had been husbanded for the internal struggle in some cases and shattered in others. The point does not rehabilitate the Nationalist state’s domestic record, which its own officials lamented in terms as severe as any communist pamphlet. It locates the Nationalists’ failure where the evidence puts it: not in an absence of resistance to Japan, but in the inability of their state to convert sacrifice into legitimacy, an inability whose roots, fiscal shallowness, factional command, and a party that had ceased to organise anything below the level of the state, long predated the invasion and were merely exposed by it.
The Nanjing decade reconsidered: the alternative that the war destroyed
Fairness to the losing side requires one more section, because the Nationalist government of the decade before the invasion has undergone a scholarly reconsideration that complicates the morality of the whole narrative. The Nanjing decade, 1927 to 1937, was long written as a prelude of failure: a government of landlords and compradores, in the communist phrase, or of faction and futility, in the liberal one. Later scholarship recovered its substance: a state that unified the currency, built the beginnings of a modern fiscal and banking system, extended railways and roads, codified law, recovered tariff autonomy, and presided over measurable growth in the modern sector, while its writ in the countryside remained as shallow as ever and its party decayed into an apparatus of office. The reconsideration matters to the dispute because it reframes the counterfactual that haunts the first explanation. Had the invasion not come, the argument runs, the Nationalist state might have consolidated by the methods of the decade into the government of a unified China, and the communists, confined to their northwestern redoubt and shrinking under blockade, might have dwindled into a regional sect: the war with Japan, on this reading, did not merely assist the revolution but rescued it, and the nationalism thesis is entitled to claim the rescue as its strongest form. The counter-argument is equally documented. The Nanjing state’s consolidation was precisely what its structure forbade: it could not tax the countryside, discipline its own militarists, or reform the land relations on which its local allies rested, and the communists it had failed to exterminate in five campaigns survived because the state’s reach ended where the revolution’s constituency began. The article does not adjudicate between the decade’s promise and its limits, since the invasion rendered the question permanently hypothetical. It insists only that the Nationalist government be treated, here as in the section on its wartime decay, as a serious state that lost, and not as a stage property whose fall explains itself. Revolutions are explained by the strength of what they overcame as much as by the strength of what overcame it, and a Nationalist China of genuine achievement makes the communist victory more remarkable, and more in need of the three-factor explanation, not less.
The civil war and the founding of the People’s Republic
The Japanese surrender in August 1945 opened a competition for the succession that both parties had spent the war preparing for. Communist forces moved to accept the surrender of Japanese and puppet units across the north and to enter Manchuria, the industrial north east, where Soviet occupation forces, in a transfer whose scale and significance historians still debate, allowed captured Japanese arms to pass into communist hands. Negotiations between Chiang and Mao, conducted in Chongqing under American pressure and continued through the Marshall Mission of 1946, produced agreements neither side expected to keep, and by the middle of 1946 the country was in full civil war. The Nationalists opened with apparent advantages in numbers, equipment, and international recognition, and took Yan’an itself in March 1947, an evacuation the communists converted into a mobile defence that bled the occupying force across the loess hills. The deeper contest was fought in the countryside, where the party now reversed its wartime moderation: a directive of May 1946 authorised the transfer of landlord land to the tillers in the liberated areas, and the Outline Land Law of 1947 generalised a radical reform that confiscated landlord holdings and equalised village landholding. Land reform in the civil war years was at once a social revolution and a mobilisation machine: the peasant who received land in the struggle meetings acquired a stake in the communist victory and an obligation to defend it, and the armies that fought the great campaigns of 1948 and 1949 were fed, carried, and reinforced by the populations the reform had organised. In the cities, the Nationalist position disintegrated on the schedule Eastman’s analysis predicts: the attempt to stabilise the currency with the gold yuan reform of 1948 collapsed within months, prices escaped all control, and the urban classes that had been the regime’s constituency concluded that it could no longer govern. The three great campaigns of the war’s final phase, in Manchuria, in the Huai River plain, and around Beijing and Tianjin, destroyed the Nationalist field armies; Nanjing fell in April 1949; and on the first of October 1949 the People’s Republic of China was proclaimed in Beijing. The Nationalist government withdrew to Taiwan, where its state survived, a fact of geography and naval power that lies outside this article’s subject, which closes with the mainland victory and the order the victors proceeded to build.
Why did the Nationalists lose the civil war after an apparent start of strength?
Their advantages were wasting assets: an inflation that destroyed urban confidence and military pay, armies corroded by faction and poor treatment of conscripts, and a government that could not tax or administer the countryside. Against them the communists combined land reform as mobilisation, disciplined forces living off organised base areas, and a strategy that traded territory for the destruction of armies.
The land reform of the civil war years deserves a further paragraph of its own, because it is the point at which the historiographical dispute and the theoretical question intersect most sharply. The reform was violent. The transfer of land was accomplished through mass meetings at which landlords were struggled against, their property inventoried and divided, and in many places their persons attacked; the party’s own directives oscillated between encouraging the movement and restraining its excesses, and the record includes both the emancipation of tenants who received land for the first time in their families’ memory and the persecution of households classified as landlord or rich peasant, a classification that could be arbitrary, hereditary in effect, and ruinous. This article records that violence as it records the violence of the Japanese counterinsurgency and of Nationalist conscription: plainly, on the testimony of all parties, and with neither narrative elaboration nor the numerical totals that the sources give in irreconcilable forms. What matters for the argument is the function the reform performed in each of the three explanations. For the second, the reform was the Yan’an programme radicalised, the social content that bound the peasantry to the party by interest and gratitude. For the third, it was an organisational instrument of classic form: the classification of every household, the staging of accusation, the distribution of spoils, and the recruitment of activists created a village order in which the party knew everyone and everyone was implicated. For the first, the reform mattered chiefly after 1945, when nationalism had done its work and a new bond was required. The same events carry all three readings without strain, which is why the dispute survives.
Manchuria and the race for the north east
One theatre of the civil war deserves separate treatment because its outcome was disproportionate to its place in the popular narrative, and because it exhibits the party’s method in a region where none of its wartime advantages applied. Manchuria, the three north eastern provinces, held China’s heaviest industry, its best railways, and a population among whom the communists had almost no organisation in 1945, their wartime bases having lain south of the Great Wall. The Japanese surrender delivered the region to a race. Soviet forces, entering in the war’s last days, occupied the cities and the railways; Nationalist forces, airlifted and sealifted by their American ally, moved to take possession of the prize; and the communists, marching overland from their northern bases, infiltrated the countryside and the smaller towns, absorbed surrendered puppet troops and local recruits, and acquired, in circumstances historians continue to argue over, substantial stocks of the surrendered Japanese army’s weapons from Soviet hands. The communists’ first strategy in the region repeated the error of every Chinese army that had coveted its cities: they attempted to hold the great urban centres and the railway corridors, and in the spring of 1946 the Nationalists’ best American-equipped armies drove them from the cities with heavy loss. The recovery that followed was the Jiangxi lesson applied on a continental scale. The communist forces under Lin Biao abandoned the cities, dispersed into the Manchurian countryside, built base areas in a region where land reform could be conducted against the collaborators’ and absentee holdings of a frontier society, and returned in 1948 with field armies raised, supplied, and politically rooted in the north east itself, to win the Liaoshen campaign that destroyed the Nationalist position in the region entire. Manchuria then became the anvil of the civil war: its industry, its arsenals, and its veteran armies fed the campaigns that followed in north and central China. The episode matters to the dispute because it is the nearest thing the history offers to a controlled experiment. In Manchuria there was no Japanese occupation to generate the first explanation’s nationalism in the party’s favour, the occupier having been the party’s own sometime patron’s army; there was no Yan’an with a decade of reform to exhibit; there was only the method, land reform, organisation, strategic patience, applied to fresh ground, and it sufficed within three years. The third explanation has no stronger exhibit, and the honest weighing must register it.
The working class as audience: the cities that did not rise
The classical theory expected the revolution to be made in the cities by their workers, and the Chinese Revolution’s urban history is the record of that expectation being inverted with deliberation. After 1927 the party’s urban underground survived as an intelligence and liaison apparatus, never again as an insurrectionary force, and the great industrial centres, Shanghai above all, passed the war years under occupation and the postwar years under a Nationalist administration whose labour policy combined suppression of independent unions with the corruption of the official ones. The party’s instructions to its urban organisations in the civil war years were a studied reversal of 1927: no uprisings, no premature exposure, the preservation of cadres and the protection of plant, on the reasoning that the cities would be taken by the armies from outside and that the working class’s contribution was to deliver its factories intact to the new order rather than to seize them in its name. The instruction was obeyed to a degree that surprised the victors themselves: in city after city in 1949, workers guarded machinery against removal and sabotage, maintained production through the transfer of power, and received the entering armies as a government rather than joining them as an insurrection. The episode completes the article’s evidence on the classification question from the urban side. The proletariat of the classical texts did participate in the Chinese Revolution, at its conclusion, in a role the texts had not written for it: not as the revolution’s maker but as its custodian, holding the modern sector in trust for a state made by others. Defenders of the party’s Marxism reply that the party acted throughout as the proletariat’s organised vanguard and that the class’s physical absence from the armies is a detail of Chinese geography; the objection replies that a vanguard which substitutes for its class at every stage, in the war, in the countryside, and finally in the factory itself, has made the substitution the system. The article has given its verdict on the criteria and does not repeat it here; but no reader should leave the subject believing the urban working class was simply absent from the story. It was present as the audience whose judgement both parties courted, whose allegiance the Nationalists forfeited and the communists husbanded, and whose disciplined reception of the new order in 1949 was among the victory’s most consequential facts.
Taking the cities: the administrative problem of victory
The entry into the cities in 1949 was planned with a care that reveals the victors’ understanding of their own history, since the party’s one previous experience of urban power, in the soviets’ brief holdings and in the disastrous attempts of 1927 and 1930, had taught it how quickly a city could ruin an army that treated it as plunder or as a puzzle. Cadres for the takeover were trained in advance and held behind the advancing armies; military control commissions assumed authority in each city as it fell; and the instructions issued to the entering forces addressed the revolution’s urban problem at the level of the individual soldier’s conduct: industry and commerce were to be protected intact, foreign residents and their property unmolested, supplies paid for, troops quartered without requisitioning houses, the famous image of the victorious army sleeping in the streets of Shanghai rather than disturb its households being the deliberate emblem of the policy. The policy had a hard administrative core. The party that had governed villages for a decade now had to govern ports, railways, banks, and factories whose operation it barely understood, staffed by the personnel of the fallen regime, whom it retained wholesale in the first instance for want of replacements, while it learned. The cities were also where the new state’s promises met their first test in the currency: the runaway inflation inherited from the Nationalists was mastered within a year by the unification of fiscal control and the balancing of the budget, an achievement of administration that did more for the new government’s credit with the urban classes than any proclamation. The episode closes the civil war narrative with the theme the whole article has pursued. What the communists brought to the cities, as they had brought to the villages, was not primarily an ideology, which the urban population received variously, but a government: predictable, disciplined, and capable of making its writ run. In a country that had known four decades of fragmentation, the provision of government was itself the revolution’s most widely felt content, and the historians who ask why the communists won are sometimes answered, by the record of 1949 and 1950, with the observation that by the end they were the only organisation in China still capable of supplying it.
The second front: the cities, the students, and the collapse of confidence
The civil war was also lost and won in the cities the communists did not yet hold, on what the party’s own vocabulary called the second front. The Nationalist government’s urban crisis after 1945 compounded every wartime failing. The takeover of the formerly occupied territories was conducted with a corruption so open that the returning officials earned a popular name as requisitionists, and the assets of the enemy and of collaborators flowed into private hands while the currency resumed its fall. Students, the constituency whose patriotism had supplied the December Ninth movement of 1935 and the volunteers of Yan’an, turned against the government in waves of protest: against the civil war itself, against hunger and the price of rice, against the brutality of the police who answered demonstration with beatings and shootings that the press then carried to the reading public. The intellectuals and the professional classes, whose allegiance a modernising government might have claimed, concluded from the inflation and the repression together that the regime had forfeited the capacity to govern, a conclusion the communists’ urban underground, rebuilt patiently after 1945 under strict instructions to preserve itself rather than to rise, did much to organise and nothing to discourage. Suzanne Pepper’s study of the civil war years remains the standard account of this dimension, and its finding disciplines the military narrative: by the time the field armies met in 1948, the Nationalist state had already lost the allegiance of the urban society that financed, staffed, and informed it, so that its collapse, when it came, had the speed of a structure whose supports had been withdrawn in advance. The communists did not create the urban crisis; the government’s own conduct created it. What the party supplied, here as in the countryside, was the organised alternative into which disaffection could flow, and the discipline, unusual in victorious movements, to enter the cities in 1949 with instructions to protect industry, pay for supplies, and sleep in the streets rather than requisition houses, converting the second front’s sympathy into the new order’s initial credit.
Why did the urban classes abandon the Nationalist government before its armies were beaten?
The takeover of the recovered territories was conducted corruptly, inflation destroyed salaries and savings a second time, and police violence against student protest completed the alienation of the educated public. The regime lost the allegiance of the society that financed and staffed it while its armies were still in the field.
The war of the railways: sabotage, protection, and the guerrilla economy
The railways were the skeleton of the occupation and therefore the guerrilla war’s most contested ground, and the fighting along them exhibits the party’s method in its economic form. Japan’s hold on north China was a hold on lines: the garrisons sat in the towns the railways served, the blockhouses guarded the embankments, and the countryside between the lines was the sea in which the guerrillas swam. The communist forces and the village militias waged against the lines a war of interruption, lifting rails, cutting telegraph wires, ambushing repair parties and the trains that carried the occupation’s coal and grain, while the population along the corridors was organised to deny the enemy the labour and intelligence his patrols required. The war had a second face that the sabotage romance conceals. The base areas themselves needed the lines’ traffic, smuggled, taxed, and at times protected: salt, cloth, medicine, and ammunition moved across the corridors by night in both directions, and the party’s economic organs traded with the occupied cities through intermediaries, so that the railway was simultaneously the enemy’s artery and the base areas’ supply route, to be cut at the points where cutting hurt the occupier and preserved where its commerce fed the resistance. The famous railway guerrillas of the Shandong corridor, whose exploits entered popular legend, were the romantic surface of this double economy. Beneath it lay the calculation that governed the whole base area system: the enemy’s infrastructure was to be used, taxed, and denied in proportions set by the movement’s needs, never merely destroyed for the satisfactions of destruction. The Hundred Regiments Offensive of 1940, which tore up the lines on a grand scale, was the exception that revealed the rule, and the retaliation it provoked taught the party the price of advertising its strength against the skeleton all at once. Thereafter the war of the railways returned to its proper tempo, a thousand small interruptions administered by the people who lived along the lines, invisible in the campaign histories and decisive in the arithmetic of occupation, since an empire that must garrison every mile of its communications has already conceded the countryside between them.
On Protracted War: the strategy text of the war’s opening phase
The party’s military method had a canonical statement, and the statement’s argument is worth recovering because it shows the leadership thinking about the war in categories the historiography later borrowed. In the spring and summer of 1938, in lectures at Yan’an published as On Protracted War, Mao addressed the two errors he opposed: the expectation of quick victory, and the theory of inevitable subjugation, which held that China could not defeat a modern industrial power. His answer was a theory of the war’s shape. China was vast, populous, and progressively mobilised; Japan was small, overextended, and fighting a war its economy and its international position could not indefinitely sustain; the war would therefore be protracted and would pass through three stages, a strategic defensive in which Japan took the cities and the railways, a strategic stalemate in which the guerrilla war behind the lines became the war’s centre of gravity and the invader’s conquests digested him, and a strategic counteroffensive when the correlation of forces had turned. The analysis did the political work the party required: it made a virtue of the strategy the party was actually pursuing, guerrilla war from base areas, by installing that strategy as the war’s decisive theatre in its middle stage, and it promised the party’s followers a victory whose timetable was long enough to organise for. As prediction it was substantially vindicated in form, if not in the agency of the final stage, which arrived in 1945 through Japan’s defeat in the wider Pacific war rather than through the communist counteroffensive the text projected. As evidence in the dispute it cuts in an interesting direction. The text is a nationalist document in its premises, a theory of how China, not how the proletariat, would win its war of resistance; and it is simultaneously a party document in its machinery, assigning the decisive role in the decisive stage to the forces the party commanded. The first and third explanations can each quote it as a charter, and the fact that the movement’s own canonical strategy sustains both readings is a small, perfect emblem of the settlement this article defends.
The military contest in its own terms
The civil war’s military history is sometimes treated as the mere ratification of political facts decided elsewhere, and the treatment is an error the article should correct, because armies that are politically favoured still have to win battles, and the communist armies of 1946 to 1949 won them against an enemy superior, at the outset, in numbers, artillery, and air power. The method was the one the Jiangxi defeat had written into doctrine, scaled up. Communist strategy refused the defence of places, including, in 1947, Yan’an itself, and traded territory for time and for the destruction of the enemy’s forces in mobile battles fought at heavy local superiority, assembled by marching farther and faster than the Nationalists believed infantry could march. As the war proceeded the guerrilla formations of the base areas were consolidated into field armies with artillery of their own, much of it captured, and the war’s character changed from insurgency to conventional campaigning without the strategy changing at all: the great campaigns of 1948 and 1949, Liaoshen in the north east, Huaihai in the central plain, Pingjin around Beijing, were annihilating battles of encirclement conducted by armies that had begun the war as guerrillas. The logistics of those armies were the land reform’s military dividend made visible: the Huaihai campaign was supplied by a host of civilian porters and handcarts mobilised through the village organisations the party had built, a supply system no requisition could have matched, which the communist command itself commemorated as the people’s contribution to the victory. On the other side, the Nationalist armies were dissipated in garrisons, their best formations committed to holding cities that became traps, their command divided between the generalissimo’s interventions and his generals’ rivalries, and their conscripts, ill-fed and ill-treated, surrendered or defected in formations once the issue of a battle was decided, swelling the communist ranks with trained men faster than any recruiting could have done. The military verdict and the political verdict are the same verdict stated in two vocabularies: the side that had organised the society behind it could concentrate force at the point of decision, and the side that had not, could not.
The dispute in outline: three explanations of the victory
Why did the Chinese Communists win?
historians identify three factors: the nationalist response to Japanese occupation, which the party organised; social and economic reform in its base areas, including rent and interest reduction; and organisational discipline and coercion. Their relative weight varied by region and period, and the dispute remains open.
With the sequence established, the article turns to its centre: the dispute about causes. The dispute has a history of its own, and its stages correspond to the stages in which evidence became available. In the first stage, the question was debated from wartime reporting and captured documents, and the party’s own account, that the masses had risen under its leadership because its programme expressed their interests, competed with the dismissals of its enemies, for whom the victory was an accident of Japanese invasion and Soviet assistance. In the second stage, beginning in the early 1960s, professional scholarship took the question up with the tools of the archive then open: a study of 1962, built largely on Japanese military and intelligence records, argued for peasant nationalism; a study of 1971, built on the published record of the border region and the testimony of the wartime observers, argued for social and economic reform. The two positions defined a polarity, nationalism against social revolution, within which most subsequent writing arranged itself. In the third stage, from the 1980s, scholars gained access to local Chinese archives, county records, and the party’s internal documents for particular base areas, and the picture that emerged complicated both poles: organisation, discipline, surveillance, and coercion appeared not as background conditions but as mechanisms in their own right, through which the other two factors were delivered to the village and their effects secured. The argument of this article, stated here so that the reader can test it against everything that follows, is the three-factor settlement: all three causes were at work in the expansion and in the victory that followed it, in mixtures that shifted from district to district and from year to year, and the celebrated quarrel between the first two accounts was fought out, on both sides, before the local archives that document the third had been opened to anyone, which is why each of the original positions could be so well documented and yet so incomplete. The settlement is not a compromise designed to offend no one. It is a claim about the evidence, and the sections that follow give each explanation its full strength before the weighing begins, since a settlement reached without hearing the cases is worth nothing.
The three explanations grid
The dispute can be set out in a single view. The three explanations grid below gives one row to each explanation and sets against it the mechanism it claims, the testimony it rests on, the ground and the years where it holds strongest, and the residue it leaves unexplained. The grid is an instrument, not a verdict: its purpose is to let the reader see, at any point in the sections that follow, which cell of the argument is being filled.
| Explanation | Mechanism claimed | Evidence relied on | Regions and periods it fits best | What it cannot explain |
|---|---|---|---|---|
| Peasant nationalism | Japanese occupation generated mass patriotic resistance, and the party, present and armed behind the lines, organised that resistance into support | Japanese military and intelligence records; the correlation between occupation and party growth; the party’s wartime moderation of its social programme | North China under direct occupation, 1937 to 1941 | Why the Nationalists, also resisting Japan, failed to harvest the same nationalism; why victory followed in 1949, four years after the occupier left |
| Social and economic reform | Rent and interest reduction, cooperatives, honest administration, and participation in the base areas gave peasants material reasons to support the party | Border region documents and publications; wartime observers’ reports; the administrative record of the Shen-Gan-Ning region | The Yan’an border region and the more secure base areas, 1941 to 1945 | Growth in contested areas where reform was thin or absent; the atypicality of the border region, which the enemy largely left alone |
| Organisation and coercion | Party penetration of the village, classification, surveillance, and disciplined coercion manufactured and maintained support that sympathy alone would not have secured | County archives and internal party documents opened to scholars from the 1980s; the records of particular base areas in central and north China | Central China under the New Fourth Army; contested zones where control was fought village by village | Why coercion succeeded for this party and failed for the Japanese and Nationalist forces that also practised it; the durability of support once the fighting moved on |
Read the grid by its last column. Each explanation is strongest exactly where its evidence was gathered and weakest where it is asked to travel: the nationalism thesis was built from the records of an occupying army watching resistance grow under its hand; the reform thesis from the publications of the one base area secure enough to publish; the organisation thesis from the internal files of a party that recorded its own methods with bureaucratic candour. The history of the dispute is therefore in part a history of archives, of which records survived, which were opened, and which a scholar could read in which decade, and the three-factor settlement is in part the observation that a dispute among archives cannot be settled by loyalty to any one of them. The tradition to which this article belongs, the study of Marxism in China as a whole, supplies the doctrinal background against which the dispute’s terms were set; what follows here is the dispute itself, explanation by explanation, each stated as its author would wish it stated.
The first explanation: peasant nationalism and Chalmers Johnson
The nationalism thesis was given its classic form by Chalmers A. Johnson, a scholar of Japan at the University of California, in Peasant Nationalism and Communist Power, published in 1962. Johnson’s question was precise: what explains the transformation of the Chinese Communist Party between 1937 and 1945 from a battered remnant into a mass movement? His answer displaced the assumption, common to the party’s friends and enemies alike, that the communists had won the peasantry by social revolution. The party’s wartime programme, he observed, was conspicuously moderate: land confiscation had been abandoned, rents merely reduced, landlords courted. A movement offering so little social transformation could not have grown so fast on the strength of its programme. What had changed in 1937 was not the programme but the country: Japan’s invasion, and in particular the savagery of Japanese counterinsurgency in the north, the punitive expeditions and the scorched-earth campaigns by which the occupation tried to drain the countryside of support for guerrillas, had created a mass peasant nationalism, a determination to resist the invader, which required only organisation to become a political force. The communists supplied the organisation. Their rivals could not: the Nationalist armies had withdrawn to the interior, the local elites had collaborated or fled, and the secret societies and village defence corps that sprang up spontaneously lacked the discipline and the political form to sustain a war. On Johnson’s account, the party’s wartime growth was therefore a national phenomenon wearing a communist uniform: the appeal was patriotic, the content of the party’s social policy largely incidental, and the mass support it enjoyed was, in his phrase for the phenomenon, a response to the breakdown of the old order under foreign attack rather than an endorsement of the new order the party intended eventually to build.
What evidence did Johnson use for the peasant nationalism thesis?
Johnson worked chiefly from Japanese sources: the military and intelligence records of an occupying army documenting, district by district, the growth of the resistance it was fighting. The correlation those records showed, between the intensity of Japanese presence and brutality and the growth of communist organisation, carried his argument that occupation itself had manufactured the party’s mass base.
The strengths of Johnson’s account are the strengths of its evidence and its logic. The Japanese records are hostile witnesses, and testimony from an enemy’s files carries a weight that a movement’s own publications cannot: the occupation’s intelligence officers had every incentive to understand the resistance correctly and no incentive to flatter its causes. The correlation at the argument’s centre is also real. Communist organisation did grow fastest where the occupation pressed hardest, in the north China plain, and grew least in the one region, the Yan’an border region, that the Japanese never seriously attempted to take, a distribution the reform thesis, built on that same border region, is embarrassed to explain. And the argument performs a permanent service in dissolving the assumption that support for a communist party must be support for communism: Johnson taught a generation of scholars to ask what the peasantry thought it was joining, and to allow that the answer might be the nation rather than the doctrine. Its limits are the mirror of its strengths. The Japanese records could register the growth of resistance but not its interior composition: they could not show what the party did inside the village to convert patriotic anger into durable allegiance, and Johnson, whose evidence ran out at the village gate, treated that interior as organisation in the thinnest sense, a flag and a rifle squad. The thesis also struggles with the sequel. Nationalism against Japan should have spent itself in 1945, when Japan surrendered; instead the party emerged from the war stronger than ever and won a civil war against a Chinese government within four years, a continuation that requires either a second mechanism, the land reform and the organisation, or an implausible extension of the first. Johnson himself, in later reflections, conceded ground on the social content of the movement’s appeal; the thesis survives as one factor of three, which is exactly the status the settlement assigns it.
The second explanation: the Yan’an Way and Mark Selden
The reform thesis received its classic statement a decade later, from the opposite side of the Pacific war’s memory. Mark Selden, an American historian, published The Yenan Way in Revolutionary China in 1971, working from the published record of the border region governed from Yan’an, the reports of foreign observers who had visited the wartime capital, and the party’s own administrative literature, read with a sympathy his method made explicit: he took the border region’s experiment seriously as a model of development, in an era when the failures of imposed modernisation elsewhere had made such experiments interesting far beyond China. Selden’s Yan’an was a place where a poor and neglected region was governed, for the first time in its inhabitants’ experience, in their interest: rents and interest reduced by enforceable rule, taxes assessed progressively and collected without the traditional squeeze, cooperatives organising production and credit, elections giving villagers a voice in the choice of the officials who governed them, campaigns against illiteracy and opium, and a party and army held to standards of conduct toward civilians that the population could compare, from experience, with the behaviour of every other armed force in China. On this account the party’s support was neither a reflex of invasion nor a product of manipulation; it was earned, by a programme of social and economic reform that answered the peasantry’s material grievances within the limits the united front allowed, and the Yan’an Way, the combination of mass participation, cooperative economics, and honest administration, was both the explanation of the wartime growth and a standing alternative model of how a poor society might develop. Where Johnson’s peasant was a patriot, Selden’s was a citizen in the making, and the dispute between them is, among other things, a dispute about what moved the Chinese villager of the 1940s: the nation insulted, or the rent reduced.
What did Selden mean by the Yan’an Way?
Selden’s term named the border region’s governing model as a coherent whole: rent and interest reduction enforced by rule, progressive taxation, cooperatives, mass participation through associations and elections, and a party and army disciplined in their conduct toward civilians. For Selden this model, and not patriotic reflex, was what earned the party the peasantry’s support.
Selden’s account has the great merit of taking the base areas seriously as governments, which Johnson’s, written from the occupier’s archive, never had to do. It explains features of the record the nationalism thesis passes over: the party’s extraordinary administrative energy, the care lavished on tax schedules and cooperative charters in the middle of a war, and the testimony, from observers with no communist sympathies, that the border region was palpably better governed than the China around it. It also explains the movement’s self-understanding, since the party’s own cadres believed they were winning support by serving the people’s interests, and a history that cannot accommodate what the actors believed is missing a mechanism by which belief became effort. The limits of the account are the limits of its showpiece. The border region around Yan’an was the most secure, the most lightly attacked, and in some ways the least typical of the base areas: a region the Japanese did not occupy and the Nationalists merely blockaded, where the party had governed since before the war and could conduct its experiment under laboratory conditions that the contested guerrilla zones of the north China plain, where growth was fastest, never enjoyed. Generalising from Yan’an to China is the move the thesis requires and the move its evidence cannot quite fund. Selden himself, in the revised edition of his study published in 1995, conceded much of this: he gave greater weight to nationalism, to coercion, and to the darker passages of the Yan’an record, including the rectification campaign’s coercive phase, in a revision that stands as a model of scholarly honesty and that moved his position, by his own hand, a considerable distance toward the settlement this article defends.
The third explanation: organisation, surveillance, and coercion
The third explanation arrived with a new archive. From the 1980s, as Chinese local archives opened selectively to researchers and the party’s internal documentation for particular regions became obtainable, a body of scholarship reconstructed the wartime movement from below and from inside, and its findings disturbed both of the established positions. Its landmark is Chen Yung-fa’s Making Revolution, published in 1986, a study of the communist movement in eastern and central China, the territory of the New Fourth Army, built from county archives and party records. Chen’s communists did not simply harvest nationalism or dispense reform. They built power methodically: entering villages behind their armies, identifying and cultivating activists, classifying households, constructing party branches and mass organisations that doubled as instruments of surveillance, controlling the flow of information, and applying coercion, against collaborators, against recalcitrant landlords, against villages that resisted taxation or recruitment, with a deliberateness that the language of mass support tends to conceal. Kindred studies filled in the regional picture: Kathleen Hartford’s work on the contested zones, Odoric Wou’s on organisation in Henan, David Goodman’s on the Taihang base area in Shanxi, each showing a party whose decisive technology was organisational, the capacity to be present in every village, every night, with a structure that remembered everything, in a countryside where its rivals were present intermittently and its enemies only in punitive columns. On this account, the question why the communists won receives a less inspiring and more mechanical answer than either predecessor offered: they won the competition for the village because they out-organised every competitor, and organisation, in a setting of war and state collapse, included the credible threat and frequent use of force against those who stood outside the order being built. Support, in this literature, is not denied; it is redescribed, as something manufactured by structures that also made opposition dangerous and neutrality difficult, so that the historian can seldom say, of any given village, where enthusiasm ended and prudence began.
What did the county archives add to the dispute?
They added the interior of the village. Japanese records had shown where resistance grew and border region publications how the model government presented itself; the county archives and internal party documents showed how power was actually constructed household by household, through classification, surveillance, controlled distribution, and coercion, a dimension neither of the earlier explanations had foregrounded.
The third explanation’s strength is that it explains the mechanism the other two presuppose. Nationalism does not organise itself, and reform does not administer itself; between a grievance and a government stands the party’s apparatus, and the archival scholarship is the first to describe that apparatus at the level at which it operated. It also explains the dark passages of the record that the reform thesis tended to hurry past: the rectification’s coercive phase, the security campaigns in the base areas, the violence of classification and struggle in the land reform, all of which fall into place as applications of a general method rather than aberrations from a benign one. Its limits are equally definite, and the settlement requires them to be stated. Coercion was not a communist monopoly: the Japanese practised it on an industrial scale and the Nationalists practised it through the press gang and the tax farmer, and both failed to convert it into durable allegiance, so coercion alone cannot be the differentia of the communist success. Nor can organisation explain its own reception: the party’s structures worked in villages where its programme answered real grievances and its nationalism was credible, and failed or stalled where those conditions were absent, as the uneven map of the base areas shows. What the archival turn established, properly weighed, is not that the third factor replaces the other two but that it is the medium through which they acted: nationalism and reform reached the peasant through an organisation that also watched him, and any account that omits the medium mistakes the message for the event.
Weighing the three: the three-factor settlement
The three-factor settlement can now be stated in its full form. Nationalism, social reform, and organisational coercion all contributed to the party’s wartime expansion, and none of the three, taken alone, explains the record. Their relative weight varied by region and by period, and the variation is not a fudge but a finding, recoverable from the geography of the war. In the occupied north China plain between 1937 and 1941, where Japanese garrisons and punitive expeditions made resistance the village’s first question, the nationalist factor predominated, and the party grew by being the resistance’s most durable local form. In the border region around Yan’an, unoccupied and governed continuously from before the war, the reform factor predominated, since no occupying army threatened it and the party’s claim rested on the quality of its government. In the contested zones of central China, where communist, Nationalist, puppet, and Japanese forces disputed the same villages in succession, the organisational factor predominated, because allegiance there was a function of which power could protect, punish, and remember. Period tells the same story in time rather than space: the expansion of 1937 to 1940 rode the nationalist surge; the contraction of 1941 and 1942, under the weight of Japanese offensives and blockade, tested the organisation and thinned the reform; the recovery of 1943 to 1945, and the civil war mobilisation that followed, fused all three, with land reform adding a social radicalism the war years had deliberately restrained. A dispute that began as a choice between nationalism and reform thus ends, on the fuller evidence, as a problem in weights and measures, and the scholarly task it leaves is not to elect a winner but to specify, for each base area and each phase, the mixture that the local record supports. That is a less dramatic conclusion than any of the three positions advertised, and it is the one the evidence, all of it, sustains.
Why did the dispute between the first two explanations last so long?
Because each was built on a different archive that genuinely supported it, and the archive that complicated both, the local Chinese record of how power was built inside the village, was closed to scholars until the 1980s. Johnson read the occupier’s files and Selden the border region’s publications; neither could see what the county records later showed.
One further observation completes the weighing, and it concerns the village’s-eye view, which all three explanations claim and none fully occupies. To the household that lived through the war in a north China village, the factors did not arrive as alternatives. The same party team that organised the resistance association also enforced the rent reduction and also compiled the household register; the same young activist who taught the literacy class also reported the village’s mood to the district committee. The villager who supported the movement might do so from patriotism, from gratitude, from ambition, from fear, and from the absence of any safer course, in proportions he could not himself have stated, and the historian’s three factors are analytic separations of a single lived experience of a new power arriving in the village and staying. This is not an argument against analysis; the separations are necessary, and this article has maintained them. It is an argument for the settlement’s form. An explanation of the victory that cannot be reassembled into the unity the participants lived is a diagram of the event rather than an account of it, and the test of the three-factor settlement is that it reassembles: a party that organised nationalism, delivered reform, and monopolised coercion, in mixtures set by place and time, against rivals who could do at most one of the three. On that test the settlement stands, and the article adopts it as its verdict on the dispute, while leaving the local weights, as the scholarship must, to the local records.
The visitor from Singapore: two governments compared by a guest
Among the witnesses whose testimony the historiography prizes, one occupies a category of his own, because he inspected both Chinas in a single journey and published his comparison. Chen Jiageng, known in English as Tan Kah Kee, was the leading figure of the overseas Chinese of Southeast Asia, a Singapore rubber magnate whose fundraising for the resistance had made him the indispensable intermediary between the diaspora and the homeland’s war. In 1940 he led a relief mission through the country, received with ceremony at Chongqing and entertained, as he recorded, at banquets whose cost and elaboration disgusted him amid a war of national survival. He then travelled, against his hosts’ wishes, to Yan’an, where he was lodged plainly, fed simply, and received by Mao in conversation he reported as frank and unceremonious, and he returned to Singapore to announce the conclusion that made his name a byword in the politics of the diaspora: that China’s hope lay with the communists, whose frugality and discipline he set against the extravagance and corruption of the capital he had just left. The episode’s value to the historian is not the verdict, which a single guest’s itinerary cannot carry, but the comparison’s structure. Chen Jiageng was no communist and no theorist; he was a patriotic capitalist measuring two governments by the standards of a businessman donor, economy, honesty, and the visible relation between what an administration said and what it did. That the border region passed his inspection and the wartime capital failed it, in the account of a man whose money both sides courted, is evidence for the second explanation from the quarter least suspected of manufacturing it. The diaspora’s allegiance, and its remittances, were among the war’s contested resources, and the contest for them was decided, in this emblematic case, by exactly the qualities the Yan’an model’s admirers described: a government whose leaders lived as they instructed others to live.
The Dixie Mission: the Americans who came to Yan’an
In July 1944 a United States Army observer group, remembered by its codename as the Dixie Mission, established itself at Yan’an, and its reports form a source for the base areas whose value the historiography has steadily raised. The mission’s political officers, John Stewart Service and John Paton Davies among them, were China-born diplomats of professional formation, and their dispatches described, with an admiration their government had not commissioned, a movement they found disciplined, popular in its districts, militarily effective against the Japanese, and led by men they judged to be communists of a distinctly Chinese temper who might be dealt with as a national force. The mission explored, inconclusively, the possibilities of American supply to the communist forces and of cooperation against Japan, possibilities the ambassador in Chongqing and the Nationalist government’s lobbying in Washington worked to close. The episode’s interest for this article is threefold. It supplies independent testimony, from officers of a power with no stake in the party’s mythology, that corroborates the second explanation’s portrait of the border region’s government and the army’s conduct at the moment the portrait was painted. It marks the high point of the alternative the counterfactual section considers, an American accommodation with the communists as a Chinese force, foreclosed by the politics of Washington and Chongqing together. And its aftermath, the destruction of the China Hands’ careers in the loyalty investigations of the following decade, when their accurate reporting was recast as the treason that had lost China, stands as the classic demonstration that the historiography of this revolution was never a merely academic pursuit: the victory’s causes became, for a generation of American politics, a question of who was to blame, and the scholarship this article follows was written, in its first decades, in the shadow of that inquisition, a shadow that pushed Johnson toward the occupier’s archive and a whole field toward the safety of documents over the danger of judgement.
The enemy converted: Japanese prisoners and the Emancipation League
The base areas conducted one experiment so contrary to the instincts of total war that its record deserves a place in any account of the movement’s methods. Japanese soldiers captured by the communist forces were not, as a rule, killed or held in the conditions that prevailed elsewhere in that war; they were fed at the standard of their captors, lectured, argued with, and recruited. Under the guidance of Sanzo Nosaka, the Japanese communist leader resident at Yan’an, the converted prisoners were organised in the Japanese People’s Emancipation League, which conducted propaganda against the imperial army in its own language: leaflets and loudspeaker appeals at the front, broadcasts, and letters to the garrisons, urging the conscripted soldiers of the empire to recognise the war’s authors and to turn their weapons, or at least their feet, homeward. The numbers involved were small, and no serious historian credits the League with a measurable effect on the occupation’s military position. Its significance lies in what it reveals about the movement’s understanding of its own struggle. The party’s doctrine distinguished, with a consistency its enemies found unnerving, between the Japanese militarists who were the war’s authors and the Japanese people who were among its victims, and it acted on the distinction at the cost of feeding and protecting prisoners whom its own soldiers, whose villages the empire had burned, had every reason to hate. Discipline of that order is not a natural product of a peasant war; it is the product of an organisation capable of imposing a doctrine’s abstractions on its members’ strongest emotions, and of calculating propaganda value across years. The episode is a small, bright exhibit for the argument this article has made at scale: that what distinguished this movement from the other armed forces of its time and place was the subordination of impulse to a programme, and that the subordination, visible alike in the treatment of prisoners, of landlords under the united front, and of the party’s own erring cadres, is the organisational fact beneath all three of the explanations that compete to account for its victory.
The Soviet Union and the revolution: patron, model, and interested party
The Comintern’s tutelage had a state behind it, and the state’s interests require separate statement, because they explain the pattern of help and harm better than any theory of error. The Soviet Union’s China policy across the whole period was governed by the security of its own eastern frontier: it required a China that would absorb Japanese expansion, and it preferred, at each juncture, the Chinese force most capable of that absorption, which was usually the Nationalist government it recognised, armed, and advised. Hence the sequence that baffled the Chinese party’s own members: Soviet advisers built the Nationalist army at Whampoa while Soviet doctrine built the communist one; Moscow sustained the united front into the teeth of 1927 because the alliance served the frontier; and in August 1945 the Soviet Union signed a treaty of alliance with the Nationalist government, recognising it as China’s legitimate ruler and counselling the communists, who were urged by Stalin to join a coalition under Chiang rather than fight for power, a counsel the Chinese leadership outwardly entertained and inwardly disregarded. Soviet forces entering Manchuria in 1945 behaved as an occupying power with its own account to settle, removing industrial plant as reparations on a large scale while permitting the communist infiltration of the countryside and the transfer of captured arms described earlier, a mixture of exaction and assistance that the Chinese party accepted as the price of a patron whose interests were not its own. Mao’s later recollection, that Stalin had distrusted the Chinese communists as peasants in Marxist dress and had expected them to lose, is self-serving as all such recollections are, but the documentary record of Soviet policy supports its substance. The finding matters to the theoretical question, because it dissolves the simplest version of the claim that the revolution was made in Moscow: the revolution’s patron doubted it, hedged against it, and armed its enemy, and the party won in the space between the patron’s interests and its own. It matters equally to the first explanation’s critics, who once attributed the victory to Soviet engineering: the engineering, such as it was, ran at least as strongly the other way.
The Comintern: an adviser whose map did not fit
Behind the narrative stands a relationship that the misreadings of the subject habitually soften: the Chinese party was, for the first half of its life, a section of an international organisation directed from Moscow, and the direction was frequently catastrophic. The Comintern’s contribution to the party’s existence was real: it encouraged the founding, supplied funds, training, and the Leninist organisational model itself, the disciplined vanguard party that proved, in Chinese conditions, to be the decisive import. Its strategic advice, by contrast, was consistently unsuited to the conditions it purported to describe. The first united front, in which the communists were instructed to subordinate themselves to the Nationalists and restrain the mass movement they were building, was a Comintern line, maintained from Moscow long after its danger was visible in Shanghai, and its collapse in 1927 was the line’s collapse. The insurrectionary orders that followed, demanding the seizure of cities by a party that no longer held them, repeated the error in the opposite direction. In the early 1930s the Comintern’s returned students, the group known as the Twenty-eight Bolsheviks, governed the party by the same imported certainties and presided over the loss of Jiangxi. Mao’s ascent, ratified at Zunyi, was in substance a rebellion against this tutelage, and the doctrine he built upon it, the sinification of Marxism, was a declaration of methodological independence: the universal theory was to be applied through Chinese conditions as investigated by Chinese communists, not administered as instructions. The Comintern dissolved itself in 1943, in part to ease the wartime alliance of its Soviet patron, and by then the Chinese party had long ceased to take its orders. The verdict the record supports is double and should be kept double: the organisational form came from Moscow and was indispensable, and the strategy came increasingly from Yan’an because Moscow’s strategy, whenever it was obeyed, had led the party toward destruction.
What was the Comintern’s most consequential error in China?
The maintenance of the first united front line into 1927: the instruction that the communists remain inside the Nationalist Party, restrain the mass movement they were organising, and trust an alliance whose military partner was preparing to destroy them. The Shanghai catastrophe that followed was the direct price of strategy dictated from Moscow against local evidence.
The relationship’s later chapters complicate any simple story of emancipation, and honesty requires them. The Chinese party’s independence was won in strategy, not in allegiance: it continued to regard the Soviet Union as the head of the world movement, sought its recognition and its material help, and received, at the civil war’s opening, the ambiguous assistance of the Manchurian transfer, while being refused the full commitment it sought. The tension between a national road to power and an international movement’s discipline runs through the whole history and beyond it, into the quarrel between the two communist states that the victory made possible, a quarrel treated in the comparison of Maoism and Soviet Marxism and only flagged here. For the purposes of the dispute about 1949, the Comintern’s significance is as a control case. Here was an organisation with the correct doctrine, by its own account, the correct international connections, and a Chinese party eager in its early years to obey, and its interventions damaged the cause they directed at almost every point where they were decisive. The Chinese party won when it stopped being directed. That finding, unwelcome to every mythology, is among the best attested in the whole literature, and it disciplines the theoretical discussion to which the article now turns: if the victory was won by a party that succeeded by adapting its doctrine to its conditions against the orders of the doctrine’s custodians, the question of whether the victory was a Marxist one cannot be settled by pointing to the party’s credentials. It must be settled by criteria, and the criteria must be stated.
The debate inside the tradition: the Chinese Trotskyists and the charge of substitution
The objection that the revolution was a peasant war under Marxist management was first pressed, with the greatest theoretical precision, from inside the Marxist movement itself, and its history belongs in the article because it shows the classification question being fought over before the victory, by men who shared the doctrine and disputed its application. The catastrophe of 1927 had been, among other things, a battle in the Soviet succession: the Trotskyist opposition charged that Stalin’s Comintern, by chaining the Chinese party to the Nationalists, had strangled a revolution that the Chinese proletariat, in the Shanghai risings of 1926 and 1927, had shown itself ready to lead, and Trotsky drew from the defeat the doctrine of permanent revolution in its modern form, that in a backward country the democratic tasks could be completed only by a working class that passed without pause to socialist ones. The Chinese party’s founder, Chen Duxiu, made the scapegoat for the line he had been ordered to execute, moved toward the Trotskyist position after his expulsion, and the small Chinese Trotskyist movement that formed around his legacy maintained, through the decades of the party’s rural exile, the classical indictment in its purest form: a party that had abandoned the cities and the working class for the countryside and the peasantry had, whatever its banners, abandoned the revolution’s agent, and would build, if it won, a state in the peasantry’s image and the bureaucracy’s interest. The indictment was answered by the party with excommunication and, where its power reached, with persecution, and the Trotskyists remained a sect without an army in a century that rewarded armies. But their question outlived their movement, because it is the theoretical question of this article stated from the left: not whether the party was sincere, but whether sincerity is a criterion. The scholarship that later asked whether 1949 was a Marxist revolution was, knowingly or not, re-asking the question the Chinese opposition had posed in 1929, and the criteria set out in the following section, agent, programme, leadership, outcome, are the terms into which that old quarrel is best translated for a reader who meets it now.
The theoretical question: what would make this a Marxist revolution?
The historiographical dispute asks why the communists won. The theoretical question asks what their winning was, and it is the harder of the two, because it cannot be answered from the record until the criteria of classification are made explicit. In the classical Marxism the party professed, the revolutionary agent is the industrial proletariat: the class created by capitalism itself, concentrated in its factories, disciplined by its production, and positioned by its lack of property to expropriate the expropriators and socialise the means of production. China in the 1940s possessed such a class only in embryo, a few million industrial workers in a population of hundreds of millions of peasants, concentrated in coastal cities the party did not hold until the revolution was already won. The force that made the Chinese Revolution was, by every sociological measure, the peasantry, organised by a party whose own membership was by then overwhelmingly of peasant origin, fighting a war whose programme, in its decisive years, promised national liberation and moderate agrarian reform. Whether the result is a Marxist revolution depends on which of four criteria is taken as decisive, and the article states them plainly because the literature too often argues past itself by switching among them. The first criterion is the agent: who made the revolution. The second is the programme: in whose name, and for what announced ends, it was made. The third is the leadership and its doctrine: whether the directing organisation was Marxist in its self-understanding, its theory, and its ultimate aims. The fourth is the outcome: what order the revolution built, and in particular whether it transformed the property relations of the society it conquered. On the first criterion the Chinese Revolution fails the classical test outright. On the second it passes only in part, since its wartime programme was national and democratic rather than socialist. On the third it passes completely. On the fourth, as the next sections argue, it passes in the end, though the passage was completed after the victory rather than in the act of winning it. A classification that does not say which criterion it is using is not a classification but a loyalty, and the article’s procedure is to hold all four in view while the strongest objection is stated and answered.
On what criteria does the classification of the revolution turn?
On four that the literature habitually confuses: the class character of the agent that made it, the programme it proclaimed while making it, the doctrine of the leadership that directed it, and the property relations it established afterward. The Chinese case divides across the four, which is why the argument about its nature has never closed.
The party’s own theoretical answer to the problem deserves statement before the objection, because it is more sophisticated than either its enemies or its admirers usually allow. In the writings of the Yan’an period, above all the essay On New Democracy of 1940, Mao argued that a country in China’s condition, semi-colonial and semi-feudal in his analysis, could not pass directly to socialism, and that its revolution necessarily had two stages: a democratic stage, directed against imperialism and feudal relations, which would create a new democratic republic of several revolutionary classes under communist leadership, and a socialist stage to follow at a distance the first stage would determine. The peasantry’s centrality was not an embarrassment to this theory but a theorem of it: in the democratic stage the peasantry was the main force, the proletariat, through its party, the leader, and the distinction between the class that supplies the force and the class in whose name leadership is exercised was elevated into doctrine. The theory of stages had cousins elsewhere in the Marxist tradition, most sharply in the Trotskyist doctrine treated in the study of permanent revolution, which had drawn the opposite moral from the same premise, holding that a democratic revolution in a backward country could succeed only by passing immediately into socialist measures under working-class leadership, and the Chinese party’s polemics against that doctrine were a live part of its self-definition. What the Yan’an theory achieved, at the level of the party’s own practice, was a licence: it permitted the movement to be, simultaneously and without bad faith by its own lights, a peasant war for land and national liberation and a Marxist revolution in progress. Whether the licence is valid is precisely what the objection contests.
The strongest objection: a peasant war with Marxist leadership
The objection must now be stated at full strength, in the form its most serious proponents have given it, without softening and without an answer interleaved. It runs as follows. A revolution is classified by what it is, not by what its leaders have read. The Chinese Revolution was made by a peasant army, recruited from villages, sustained by villages, and fighting for the objectives villages understood: resistance to a foreign invader, relief from rent and usury, land. Its social composition was peasant, its theatre was the countryside, its methods were those of the peasant wars that run through Chinese history, and its programme during the whole period of its expansion was, by the party’s own design, not socialism but national resistance and moderate reform, a programme a patriotic landlord could and often did endorse. The industrial proletariat, the agent to whom Marx assigned the revolution and in whose name the party claimed to act, was a spectator of the process, small in number, absent from the party’s ranks and armies, and admitted to the drama only at its conclusion, when the victorious peasantry’s party entered the cities it had encircled. Lucien Bianco, in his study of the revolution’s origins, gave the objection its scholarly spine, arguing that the revolution’s energy came from the accumulated grievances of the Chinese peasantry, which would have sought an outlet under any banner, and that the party’s achievement was to channel a peasant revolt it had not created and could not have created. The Trotskyist critics of the party drew the complementary conclusion from within the Marxist tradition itself: a party of peasants, however disciplined, would govern as peasants writ large, and its revolution would issue in a nationalist state, not a socialist one. On this reading, to call the events of 1927 to 1949 a Marxist revolution because the victorious organisation carried a Marxist name is to empty the category of content: if a peasant war under a moderate programme qualifies, then the term distinguishes nothing, and any insurgency whose officers have read Marx becomes a Marxist revolution by courtesy. The objection concludes that the honest description is the plainer one: a peasant war with Marxist leadership, a great national and agrarian revolution, directed by men formed by Marxism, which is a different and lesser claim than the one the party’s historiography advances.
Does calling the revolution Marxist empty the category?
On the objection’s terms, yes: if the agent, the programme, and the social content are all non-proletarian and non-socialist during the decisive years, then the label rests on the leadership’s doctrine alone, and a category that can be satisfied by doctrine alone will stretch to cover any movement whose officers profess it.
That is the objection at its strongest, and the article has deliberately left it standing unanswered through this section, because the checkpoint of honest treatment is that the reader meet it whole. Two of its premises are simply true, and the answer that follows does not contest them: the wartime programme was not a socialist programme, and the agent of the revolution was the peasantry. What the answer contests is the inference from those premises to the classification, and it contests it on the fourth of the criteria, the outcome, by asking the question the objection avoids: what did this peasant war do with its victory? Peasant wars are among the best documented phenomena in Chinese history, and their outcomes form a pattern against which the communist victory can be measured. The great rebellions either failed, as the Taiping failed in the nineteenth century, drowning their kingdom in its own contradictions, or succeeded in the only way the old order permitted, by founding a new dynasty, as the peasant rebel Zhu Yuanzhang founded the Ming, exchanging one imperial house for another while leaving the structure of landholding, taxation, and village hierarchy intact beneath the change of surname. A peasant movement’s horizon, in the classical pattern, is a better emperor and a remission of taxes; it does not aim at, and cannot conceive, the abolition of the landlord class as a class, the reorganisation of the entire society on new property relations, or the construction of a state that penetrates every village in the country permanently. The victors of 1949 did all of these things, and did them because of what their leadership believed. That is the ground on which the answer stands.
The answer: what the party did with power
Consider the record of the first years after the victory, stated as bare sequence. The new state unified a country that had been fragmented among warlords, occupiers, and rival governments for decades, and extended a single administration to the village level everywhere on the mainland, something no Chinese government of the century had achieved. It then carried out, under the Agrarian Reform Law of 1950 and in the years immediately following, the completion of the land reform: the confiscation of landlord holdings throughout the country and their distribution to the peasantry, which abolished the landlord class, the central social institution of the old countryside, not as a wartime expedient in contested zones but as a permanent act of state covering the whole society. The reform was enforced, as its civil war predecessor had been, through mass campaigns in which violence against those classified as landlords was a recurring feature, noted here as the record attests it and without elaboration. Within the same half decade the state suppressed the armed and organisational remnants of the old order, broke the power of the secret societies and the urban underworld, unified the currency and the fiscal system, and began the transfer of industry and commerce toward state and collective ownership that the following years completed, together with the collectivisation of the agriculture the land reform had just distributed. Whatever judgement is passed on these acts, their character is not in doubt: they are the acts of a state executing a programme of social transformation derived from a theory, and no peasant movement in the Chinese record, victorious or defeated, ever attempted their like. The party also did the quieter things a peasant war never does. It subordinated its army permanently to its political organisation, so that the gun never became an independent principality of the kind every Chinese militarist had built. It staffed its state with cadres trained in a doctrine it taught as a universal science of society. And it understood itself, and taught its members to understand themselves, as the Chinese section of a world transformation, a self-understanding that shaped decisions from the land law to foreign policy. The answer to the objection is therefore that the leadership’s doctrine was not a courtesy title worn over a peasant war; it was the efficient cause of everything that distinguished the victory’s sequel from every peasant victory before it.
What did the party do that a peasant movement would not have done?
It abolished the landlord class as a class across the whole country, built a party-state that reached permanently into every village, kept its army subordinate to its political organisation, and within a decade had moved industry and agriculture toward socialised ownership. Earlier victorious peasant movements had changed dynasties; none had attempted to remake the property relations of the society.
The concession must now be made as plainly as the brief of honest treatment requires, and made without the qualifications that usually smuggle it away. The party’s programme during its decisive expansion, the programme under which it recruited its armies and governed its base areas between 1937 and 1945, was not a socialist programme, and it was not presented as one. It promised resistance to Japan, rent and interest reduction, cleaner government, and a new democratic order in which private capital, including the national bourgeoisie, had a protected place; Mao’s own texts of the period are explicit that socialism belonged to a later stage and that to confuse the stages was an error of the left as much as of the right. Peasants who joined the movement in 1940 were not enlisting for collectivisation, and the collectivisation that arrived in the middle 1950s was, for many of them, an unwelcome second revolution imposed on the first. The article’s verdict on the theoretical question follows from holding the answer and the concession together. Judged by its agent and its wartime programme, the Chinese Revolution was a peasant war and a national revolution, and the objection is correct. Judged by its leadership’s doctrine and by what that leadership did with power, it was a Marxist revolution, and the objection’s inference fails. The classification is therefore not a fact to be discovered but a choice among criteria to be defended, and the scholarly obligation, which the party’s historiography and its detractors alike have shirked when it suited them, is to state the criterion before announcing the verdict. Stated so, the question that titles this debate dissolves into the more useful one the criteria expose: how a doctrine carried by a party can redirect the outcome of a war fought by another class, which is a question about parties, not about peasants, and it is the question the Chinese case bequeaths to the theory of revolution.
The agrarian argument inside the party: the rich peasant question
Beneath the party’s successive agrarian policies runs a single recurring argument, and following it gives the reader a thread through the whole history that the three explanations, focused on the party’s appeal to others, tend to miss: the argument was inside the party, about the peasantry itself. The question was the rich peasant, the household that owned its land, worked it with some hired help, and stood between the landlord above and the poor peasant below. To confiscate the rich peasant’s surplus land, as the most radical reading of land revolution required, was to strike at the village’s most productive farmers and to frighten the middle peasantry, the revolution’s necessary allies, with the spectacle of a prosperity punished. To protect the rich peasant was to leave the village’s internal exploitation half reformed and the poor peasantry, the party’s shock troops, half satisfied. The party oscillated along this seam for thirty years, and the oscillation was never merely tactical: it expressed the genuine duality of a movement that needed the peasantry as a whole against its enemies and needed the poor peasantry against the village’s own order, and could not finally have both allegiances at full strength. In Jiangxi the line swung to the left, toward the elimination of the rich peasant as a category, and the swing contributed to the soviet’s isolation. In the war years the line swung to protection, rich peasants enrolled among the middle elements of the three-thirds system, and the swing underwrote the united front’s breadth. In the civil war land reform it swung left again, and the party’s own directives of 1948, correcting the excesses of the 1947 radicalism, swung it partway back, in a public admission that the movement’s activists had misclassified and ruined households the policy required it to protect. After 1949 the argument was settled by abolition of the question: collectivisation dissolved the rich peasant into the cooperative along with everyone else, and the category that had tormented the party’s agrarian thought for a generation ceased to exist as a form of property. The thread is worth following because it shows the peasantry of the party’s theory to have been, in the party’s practice, not one class with one interest but a stratified village whose strata the party courted, sacrificed, and recombined as its situation changed, a record that the objection’s portrait of a simple peasant war, and the defence’s portrait of a simple emancipation, both flatten beyond recognition.
The paperwork of revolution: deeds burned, certificates issued
Land reform was conducted as much in documents as in meetings, and its documentary rituals repay attention because they show the revolution registering itself in the village’s memory. The old order’s countryside ran on paper of a particular kind: the landlord’s deed chest, the debt contract with its compounding clauses, the tenancy agreement witnessed by the village elders, instruments that embodied, in the peasant’s experience, the whole machinery of obligation. The reform’s ceremonies inverted the archive. Deeds and contracts were produced at the struggle meeting, read out, denounced, and in many districts burned, the destruction witnessed by the village as the public cancellation of the past. In their place the new government issued its own paper: the land certificate, made out to the household and, in the reform’s more careful phases, naming its members, which the peasant received as the first document in his family’s history that the state had issued in his favour rather than against him. The certificate’s authority rested on the same apparatus that had compiled the classification, and the peasant who framed it above his door understood both halves of its meaning: it was his title against the old landlord and his registration in the new order’s ledger, a protection and an enrolment in a single sheet. Historians of the reform have learned to read these rituals as the revolution’s self-inscription, the moment at which an abstract change of property relations became a household’s most treasured possession and a government’s most intimate presence. The episode generalises the article’s finding about the party’s method. Where other Chinese powers had taxed the village and conscripted it, this one documented it, and a documented population is a governed population in a sense the tax farmer never achieved: the state’s knowledge and the household’s security were manufactured together, in the same act, from the same paper.
Pacification: the new state and the armed countryside it inherited
The order the victors built had first to survive the countryside they had inherited, and the pacification campaigns of the new state’s opening years belong to the record of what the party did with power. The China of 1949 was saturated with arms and with men accustomed to living by them: defeated soldiers adrift from their units, the bandit bands of the hill country swollen by a decade of war into armies of predators, the local militias and secret society forces that every collapsing authority had armed and none had disarmed. In many districts the writ of the new government at its founding ran no farther than the county town, and the roads between towns were unsafe in the manner of the warlord years. The campaigns that followed, conducted by the field armies turned inward and by the public security organs the state was building, suppressed the banditry district by district over several years, in operations that combined military encirclement with the political method the party had perfected: the mobilisation of the villages against the bands, the offer of surrender terms to the rank and file, and exemplary severity toward the leaders. The violence of the pacification is reported here under the article’s standing rule, as a fact attested by the record, without elaboration and without the totals the sources state in incompatible forms. Its significance for the argument is administrative rather than military. Every previous claimant to govern China in the century had faced the same armed countryside and had compromised with it, taxing through it or renting it; the new state was the first to disarm it systematically and to replace its patchwork of private forces with a single public monopoly of violence reaching the village. Whatever else the revolution changed, it changed that, and the change was felt by every peasant household in the country as the elementary fact of a government that could, for the first time in living memory, guarantee the road to market. The monopoly so established was the foundation on which the transformations described in the following section, for good and ill alike, were erected.
The second act: from land reform to collectivisation
The sequel that the answer relies on has its own internal history, and stating it protects the answer from the charge of reading the victory backward. The order the party built after 1949 was not constructed to a single blueprint unveiled at the founding. The new democracy of the Yan’an texts was, in the first years, substantially practised: a coalition government in form, a mixed economy in fact, private industry and commerce retained and even encouraged under state direction, and a land reform that created, by design, a countryside of small proprietors, the classical outcome of a bourgeois agrarian revolution. The turn toward socialism came by stages whose timing the leadership itself debated: mutual aid teams gathered the new proprietors into seasonal cooperation; lower cooperatives pooled their land for shares; and in 1955, in a decision that overruled the gradualists in his own leadership, Mao accelerated the movement into higher cooperatives that abolished private landholding in the villages in all but name, a transformation substantially complete by the end of 1956, together with the transfer of urban industry and commerce to joint state-private and then state ownership. Industry followed the same rhythm, protection first, conversion after. The sequence is evidence, and both sides of the classification debate cite it correctly for opposite conclusions. For the answer, the sequence shows a leadership whose socialist destination governed its patience: the stages were tactics, the direction was doctrine, and a peasant movement does not collectivise the land it has just distributed. For the objection, the sequence shows exactly the substitution it warned against: a peasantry that had fought for land received collectivisation at the hands of a state it did not control, and the revolution’s second act was performed upon its makers rather than by them. The article’s verdict has already been given and is not altered by the sequel; but the sequel is the strongest single body of evidence each side possesses, and a reader who understands why it serves both has understood the classification problem itself, which is that the same facts, the party’s doctrine and the peasantry’s experience, diverge more widely the longer the story is followed.
The comparison with October: two revolutions, two roads
The inevitable comparison is with the revolution from which the Chinese party’s organisational model derived, and the comparison is instructive precisely because the two cases agree in form and differ in almost everything else. The October Revolution of 1917, treated in this series in the Marxist analysis of the October Revolution, was an urban insurrection: a party of the industrial working class and the garrison soldiers seized the capital of a state whose authority had collapsed, in a matter of days, and then fought a civil war to defend the seizure. Its agent matched the classical theory tolerably well; its road to power was the shortest imaginable; and its party spent the following years improvising the government of a country it had taken before it had a programme for holding. The Chinese Revolution inverted every term. Its agent was the peasantry; its road was the longest in the history of successful revolutions, twenty-two years from the catastrophe of 1927 to the proclamation of 1949; its party took power not by seizing a collapsed centre but by building a rival state, territory by territory, until the rival was the stronger, and entered the capital at the end as a government already formed, with its army, its administration, its currency experience, and its habits of rule a decade old. Where October was a coup that became a revolution, China was a revolution that concluded with the formality of a transfer of capitals. The doctrinal consequences were correspondingly opposite. October’s success had appeared to confirm the classical sequence, proletarian insurrection at the centre of a modern state; China’s success proposed a different sequence altogether, protracted people’s war from the countryside, and proposed it with the authority of the larger country and the longer struggle. After 1949 there were two models of Marxist revolution in the world, not one, and the second was the one addressed to the agrarian majority of mankind, which is why the Chinese case, and not the Russian, became the template studied by insurgents across Asia and beyond.
How did the Chinese road to power differ from the Russian one?
October was an urban insurrection that seized a collapsed state at its centre in days and improvised a government afterward. The Chinese road was a protracted rural war in which the party built a complete rival state in its base areas over more than a decade, and arrived in the capital already governing.
The comparison also disciplines the theoretical question from the other side. The Russian party of 1917, whose revolution fits the classical criteria far better, proceeded after its victory to measures, the expropriation of the landlords and the socialisation of industry, that the Chinese party matched and in agriculture eventually exceeded; while the Chinese party, whose revolution fits the criteria worst, produced in its base areas a decade of moderate, coalitionist government that the Russian party, in power, never attempted for a comparable span. The two cases cross-classify on every criterion the earlier section distinguished, agent against programme, programme against outcome, which is the empirical reason the article refuses a single-criterion verdict. It is also the reason the Chinese case mattered so much within the Marxist tradition itself: it demonstrated that the Leninist party, the one element the two revolutions shared exactly, could be detached from the class and the city that had defined it, and set to work in a countryside among a peasantry, with results the theory had not predicted and could not, afterward, disown. What survived the transplant was the party. What the transplant changed was everything the party was for, in the short run; and the long run, in which the party used its state to attempt the socialisation its war had postponed, is the ground on which the two revolutions converge again, as different roads surveyed by the same doctrine arriving, by opposite itineraries, at the party-state.
The peasantry in the theory: from a sack of potatoes to the main force
The Chinese outcome was an affront to the classical texts on their most sensitive point, the political capacity of the peasantry, and the history of that point is worth tracing because the Chinese party’s theorists traced it themselves. Marx’s judgement on the smallholding peasantry, delivered in The Eighteenth Brumaire of Louis Bonaparte, was the bleakest in his work: the French smallholders, he argued, formed a vast mass whose members lived in similar conditions but without entering into manifold relations with one another, a class in itself that could not become a class for itself, could not represent itself, and had to be represented, their millions aggregated, in his image, like potatoes in a sack, by simple addition rather than by organisation. The judgement was not contempt; it was a structural claim about dispersion, the isolated household on its plot, as a barrier to the collective action that factories imposed on workers by the architecture of production itself. Engels had written more warmly of peasant war in his history of the German rising of 1525, and Lenin, governing a peasant country, had made the worker-peasant alliance the practical centre of his politics while never conceding the alliance’s senior partnership. The Chinese party’s experience pressed the tradition past all of these positions. In the Hunan report of 1927, Mao had already reversed the valuation, presenting the poor peasantry not as the revolution’s drag but as its most revolutionary stratum, and declaring of their rising that “a revolution is not a dinner party”, not an essay or a painting, but an insurrection, an act of violence by which one class overthrows another. The Yan’an theory completed the reversal by constitutionalising it: the peasantry as the main force of the democratic revolution, led by a party that embodied the proletariat’s consciousness independently of the class origins of its members. The theoretical repair is elegant and its risk is obvious, and the tradition’s critics named the risk early: if proletarian leadership can be supplied by a party regardless of who composes it or whom it leads, then the class analysis that was supposed to discipline the party’s claims has been replaced by the party’s claim about itself, and substitutionism, the substitution of the organisation for the class, ceases to be a warning and becomes a constitution. The Chinese case does not refute that warning. It exhibits, on the largest scale history affords, both the power of the substitution, a party that made a revolution its class could not have made, and its price, a revolution whose direction was thereafter whatever the party decided it was.
Why did classical Marxism distrust the peasantry as a revolutionary agent?
Because the peasant household produces in isolation. Marx argued that smallholders, dispersed on their plots and bound to locality and custom, form a class by shared condition but cannot organise themselves as a class, and so are represented by others rather than representing themselves, a structural disability the factory was supposed to spare the proletariat.
It is worth pausing on the evidentiary subtlety here, because the Chinese party did not refute Marx’s structural claim so much as engineer around it. The party’s whole organisational art, the cells, the associations, the militia, the mass line, consisted in supplying from outside exactly the connective tissue whose absence Marx had diagnosed: the relations among households that the plot did not generate, the party generated, and maintained by discipline where the factory maintained them by production. The peasantry of the base areas did not cease to be dispersed smallholders; it was organised as if it were a concentrated class, by an institution that never slept. This is why the third explanation of the victory, the organisational one, and the theoretical question are the same question seen from two sides. If the organisation was the decisive factor in the victory, then the victory was the party’s work in a sense stronger than leadership: the party manufactured the agent that the theory required and the society lacked. And if that is so, then the objection that the revolution was not Marxist because its agent was the peasantry, and the claim that it was Marxist because its party was, are both half of a single finding, that in China the party preceded the class it claimed to represent and created the conditions of its own mandate. Historians of the revolution and theorists of Marxism have been circling that finding for decades, under different vocabularies; the article’s contribution is to state it in one place and to let the reader see that the historiographical dispute and the theoretical dispute are, at this depth, one dispute.
The army as a social institution: the third member of the triangle
The base area system is usually described as a triangle of party, government, and army, and the army’s side of the figure is the one foreign accounts most often flatten into a merely military instrument. The Red Army, renamed with its fronts and field armies across the period, was designed as a political organisation that fought. Its soldiers were recruited overwhelmingly from the peasantry it protected, and its discipline toward that peasantry was codified in rules whose fame in China exceeds that of any battle: the Three Main Rules of Discipline, obedience to orders, no taking of so much as a needle or thread from the masses, the turning in of all captured goods, and the Eight Points for Attention, which descended to the level of courtesy itself, speak politely, pay fairly for purchases, return what is borrowed, replace what is damaged, do not hit or curse, do not damage crops, take no liberties with women, do not ill-treat captives. The rules were enforced, and their enforcement was a strategic weapon of the first order in a country where the population’s experience of armies, warlord, Nationalist, and Japanese alike, was of licensed predation: a force that paid for its sweet potatoes and slept on the doorsteps it had swept was a walking argument for the order it served, and peasants who had hidden their grain from every uniform learned to feed this one. The army also worked: its units farmed wasteland in the production campaigns, dug the terraces and the tunnels, taught in the literacy classes, and supplied the cadres of new base areas when a district was opened, so that the boundary between the army and the administration behind it was deliberately porous. And the army was schooled, its political departments and its rectification sessions making it, in the party’s phrase, a fighting force, a working force, and a school at once. When the objection describes the revolution as the work of a peasant army, the description is accurate and incomplete: it was a peasant army of a kind no Chinese peasantry had ever produced by itself, an army whose conduct, cohesion, and purposes were the party’s manufacture, and the distinction between the two halves of that sentence is, once again, the distinction on which the whole classification debate turns.
What were the rules that governed the army’s conduct toward civilians?
The Three Main Rules of Discipline required obedience to orders, forbade taking anything from the population, and required captured goods to be turned in; the Eight Points for Attention prescribed courtesy in the small change of daily contact, fair payment, return of borrowed articles, replacement of damage, and protection of captives and of women.
The influence abroad: people’s war as an export
The Chinese Revolution’s largest effect on the world was as a method. What the victors exported, in the decades after 1949, was not primarily arms or money, though both flowed, but a doctrine of how revolution could be made in an agrarian country: a disciplined party, a people’s army built from the peasantry, base areas governed as states in embryo, a united front that gathered every force the principal enemy had alienated, and a protracted war that traded space and time for the enemy’s exhaustion, the whole summarised in the strategic writings that insurgents elsewhere studied as manuals. The doctrine’s first and most consequential application was in Vietnam, where the Viet Minh had already developed along parallel lines and where, after 1949, Chinese advisers, supplies, and training stood behind the campaign that ended at Dien Bien Phu in 1954; the Vietnamese victory over a colonial power was read everywhere as the Chinese method’s second proof. Korean communists with experience of the Yan’an years formed one of the currents in the northern state founded after the Second World War. In Malaya, by contrast, the same method failed, and the failure is as instructive as the successes: the Malayan Communist Party drew its strength overwhelmingly from the ethnic Chinese minority, could not root itself in the Malay peasantry whose country it proposed to liberate, and faced a colonial power that severed the guerrillas from their supporting population by resettlement, so that the conditions the Chinese case had enjoyed, a party of the majority peasantry in a countryside the state could not administer, were absent in every particular. Later movements in Asia and Latin America, down to the Maoist insurgencies of Nepal and Peru in the century’s last decades, took the Chinese writings as their charter, with results that repeated the original’s dependence on conditions: where a collapsed or illegitimate state faced a party rooted in a majority peasantry, the method retained its force; where it was applied as a formula to societies lacking those conditions, it produced catastrophe for its adherents. The comparative record across the anti-colonial world is surveyed in the study of anti-colonial Marxism’s case studies; the point for this article is narrower. The influence of 1949 was the influence of a demonstrated possibility, that the classical theory’s missing agent could be replaced by an organisation, and every movement that borrowed the method borrowed that premise with it, whether or not it knew the premise’s name.
Why did the people’s war model fail in Malaya?
The Malayan party was rooted in an ethnic minority and never won the Malay peasantry, while the colonial state physically separated the guerrillas from their suppliers by resettling the population that fed them. The Chinese method presupposed organisers drawn from the countryside’s own majority, in districts beyond the state’s administrative reach; Malaya offered neither condition.
A subtler influence ran in the opposite direction, from the Chinese example into the theory of the communist world itself. The existence of a second successful revolution, made by a different road in defiance of the first revolution’s custodians, broke the monopoly of interpretation that Moscow had exercised over what Marxism permitted. Parties and movements across the colonial world drew the conclusion that doctrine was to be investigated against local conditions rather than received as instruction, a conclusion the Chinese had drawn at Zunyi under the pressure of survival and now radiated as prestige. The quarrel that followed between the two communist powers, conducted in the language of doctrinal purity and fuelled by every material rivalry, lies beyond this article’s period and is treated in its own place in the series. What falls within the period is the demonstration effect itself, and it should be weighed as part of the revolution’s historical substance, not as an afterthought: the Chinese Revolution altered what revolutionaries elsewhere believed to be possible, and beliefs about possibility are causes in history as surely as rents and invasions are. An explanation of the victory that stopped at the victory would miss the victory’s largest consequence, which was to teach a generation of movements, rightly and wrongly, that the peasantry of the agrarian world was not the theory’s embarrassment but its reserve army, awaiting only the organisation that the Chinese had shown could be built.
Counterfactuals: the coalition that was not built, and the partition that was not imposed
Serious history permits one disciplined indulgence, the counterfactual that tests a causal claim, and the literature on this revolution contains two that the weighing should register. The first is the coalition of 1946. The Marshall Mission’s agreements, had they held, promised a coalition government and a national army into which the communist forces would be integrated at fixed ratios, a settlement that would have made the communists a great party within a republic rather than the founders of a rival state. The agreements failed on the questions such agreements always fail on: which army would demobilise first, whose police would govern the liberated areas, and whether either leadership believed the other would permit an election it might lose. The counterfactual’s lesson is about the depth of the civil war’s causes: the contest was not a misunderstanding between patriots but a competition for a state, between two organisations each of which had spent a decade building the apparatus of one, and the three explanations of the victory are explanations of why, when the competition was renewed in arms, the apparatus that prevailed was the communists’. The second counterfactual is partition. In 1945 and 1946 a division of China between a communist north and a Nationalist south, on the lines the century was drawing in Korea and Vietnam, was canvassed in more than one foreign ministry, and would have suited the Soviet preference for a buffer and the American preference for an end to the fighting. It was never seriously available, for a reason the counterfactual usefully isolates: neither Chinese party would accept it, both being committed to a unified China as the premise of their legitimacy, and the communists’ strength, unlike that of their Korean and Vietnamese counterparts, was distributed through the whole country rather than concentrated in a zone an occupier could assign them. The revolution was therefore always going to be decided whole, and the explanations of its outcome must explain a total victory, not a regional one, a requirement that tells, once more, against every single-factor account: nationalism alone does not take Nanjing, reform alone does not take Manchuria, and coercion alone does not take the cities whose inhabitants opened the gates in their minds before the armies arrived.
Misreadings this article corrects
The subject accumulates errors in popular treatment with unusual efficiency, and five of them are worth correcting in one place, since each has appeared at some point in the preceding sections only in passing. The first is that the party’s rural base was a doctrinal preference, a choice made from theory: it was a refuge imposed by the destruction of 1927, theorised afterward into a strategy, and the distinction matters because it reverses the causal arrow between the doctrine and the war. The second is that the war with Japan was incidental to the party’s growth, a backdrop against which an essentially domestic contest proceeded: the war was the growth’s enabling condition, the event that removed the Nationalist state from the northern countryside, legitimised the party’s armed presence, and supplied the nationalism the first explanation describes, and no account that treats it as scenery can explain the chronology. The third is that the Comintern’s guidance was a help the party outgrew: the record shows advice consistently unsuited to conditions, obeyed at the price of catastrophe in 1927 and again in Jiangxi, and the party’s independence was not a graduation but an escape. The fourth is that peasant support for the movement was uniformly voluntary, a plebiscite conducted in the party’s favour: the archival scholarship has shown the organisational and coercive machinery by which support was built and held, and the honest account holds enthusiasm and compulsion in the same frame. The fifth, and the most consequential for readers of the theory, is that the outcome settles the theoretical question, that because the party won and built a socialist state, the revolution was simply a Marxist revolution and the objection merely pedantic: the criteria problem survives the outcome, because the classification still depends on which criterion is doing the work, and the objection, answered on the ground of outcomes, remains correct on the ground of the agent and the wartime programme. A reader who can correct all five errors unprompted has the subject in proper shape.
Was peasant support for the party uniformly voluntary?
No. The archival record shows support built through organisation that combined genuine benefits, rent reduction, protection, and honest administration, with classification, surveillance, and coercion against those who resisted the new order. Villages and households divided in their responses, and the proportions of enthusiasm, calculation, and fear varied by place and period and are seldom recoverable.
Underlying all five misreadings is a single habit, the habit of treating the revolution as the unfolding of an essence, whether the essence is located in the doctrine, the nation, or the peasantry’s eternal grievances. The scholarship this article has followed is united, beneath its disputes, in rejecting that habit. Johnson’s peasants respond to an invasion; Selden’s to a government; Chen’s to an organisation; and the actors in all three accounts are men and women making choices under pressures, in villages whose particular configuration of occupation, landholding, and party presence decided which pressure bore hardest. The revolution looks inevitable only in retrospect, from the vantage of 1949. From the vantage of 1937 it was one possibility among several in a country at war, and from the vantage of 1927 it was no possibility at all. Keeping those vantages in view is the difference between studying the revolution and commemorating it, and it is the habit the three-factor settlement is designed to enforce: a settlement about weights and mixtures is, among other things, a refusal to let any single essence, nation, reform, or party, swallow a history that was made by all three in collision.
Violence and the record: what can responsibly be said
A word is owed about this article’s handling of violence and numbers, since the policy has been visible throughout and should be defended rather than merely practised. The revolution’s history is saturated with violence: the suppression of the communists in 1927 and the purges that followed it, the Japanese counterinsurgency in the north with its campaigns of burning and killing, the Nationalist state’s conscription and its repression of dissent, the communist movement’s own violence in the rectification’s coercive phase, in the security campaigns of the base areas, in the land reform’s struggle meetings, and in the suppression campaigns that followed the founding of the new state. This article has reported the violence of every party plainly, attributed where the record attributes, and without narrative detail, because detail in such matters serves neither explanation nor judgement and because the sources’ detail is itself among the most contested matter in the field. For the same reason the article has stated no casualty figures for the civil war, for the occupation, or for the land reform. Estimates in the literature diverge not by fractions but by multiples, resting on incompatible bases, wartime claims, retrospective reconstructions, and definitions that shift between authors, so that to print any single figure would be to take a side in a dispute the article has no commission to settle and to lend the figure an authority its provenance cannot support. What can responsibly be said, and what the explanations actually require, is the direction and the function of the violence: which party applied it, against whom, and what work it did in building or breaking allegiance. On those questions the record is comparatively firm, and it is on those questions that the third explanation, the one in which coercion is a mechanism rather than an atrocity count, properly rests.
Why does the study of this revolution proceed without agreed casualty totals?
Because the estimates in the literature rest on incompatible foundations: wartime claims by the belligerents, retrospective reconstructions from partial records, and definitions of the categories that differ from author to author, producing totals that diverge by multiples rather than by fractions. A figure printed without its provenance borrows authority the evidence cannot give it.
The same discipline governs the article’s treatment of the party’s growth, which has been described throughout in the qualitative form the evidence best supports: a multiplication, across the war years, to well beyond ten times the scale of the remnant that began them. Precise membership totals exist in the literature, drawn from the party’s own congress reports, and they are not suppressed here from doubt about their existence but from a judgement about their use: the dispute among the explanations does not turn on whether the party numbered one figure or another at a given congress, but on the causes of a growth whose scale no party to the dispute denies, and the article has therefore kept the number out of the argument where it would only masquerade as precision. Readers who require the totals will find them in the scholarship cited in the following section, stated there with their provenance, which is where numbers of this kind belong. The general principle is worth stating once, since it governs the whole series’ treatment of contested quantities: a figure is evidence only when its basis can be stated with it, and where the basis cannot be stated, the honest sentence is the qualitative one.
The border region’s finances and the question of the special trade
One passage of the border region’s record was long omitted from the celebratory literature and has been restored by the archival scholarship, and honest treatment requires its inclusion at the weight the evidence supports. The Shen-Gan-Ning government’s finances, under blockade, rested in part on the export of what its accounts called special goods across the lines into Nationalist and occupied territory, and Chen Yung-fa, in research that drew the predictable controversy, identified the commodity in substantial part as opium, grown in or traded through the region the party governed. The finding is contested in its scale and its interpretation: the party’s defenders point to its campaigns against opium smoking within the base areas, to the distinction between taxing a trade and conducting it, and to the universal practice of the China’s armed powers of the era, Nationalist and warlord governments among them, which financed themselves from the same traffic without apology. The episode’s place in this article is not as an atrocity exhibit but as a discipline for the second explanation. The Yan’an Way’s portrait of a government uniquely clean among China’s powers requires amendment at exactly the point where the archival method is strongest, the fiscal record, and the amendment illustrates the general law the settlement embodies: each explanation is most vulnerable where its own archive is silent, and the reform thesis, built on the border region’s publications, could not see the border region’s accounts. A government that financed its schools and its army partly from a trade it publicly condemned was practising the arts of survival common to every Chinese power of the period, and the recognition does not dissolve the real differences in its conduct toward the peasantry. It restores those differences to their proper scale, which is the scale of a wartime state under blockade, not of a utopia under glass.
The production campaigns: the economics of survival in the border region
The border region’s economy under blockade supplies a concrete test of the reform explanation, because it shows the Yan’an government solving, by organisation, a problem that would have dissolved a lesser administration. When the Nationalist subsidy ceased and the blockade tightened from 1941, the border region faced a fiscal crisis of the first order: a poor region, swollen with soldiers, cadres, students, and refugees, whose grain and cloth had to be found within its own borders or not at all. The government’s first expedient, heavier taxation of the peasantry, produced the discontent that taxation without consent always produces, and the correction that followed became the centrepiece of the Yan’an legend: the Great Production Campaign, under which the army, the administration, and the schools took up farming, spinning, and manufacture themselves, most famously in the reclamation of the wasteland at Nanniwan by a brigade that became the model of self-sufficiency. The campaign’s economic substance was real: the institutions of the border region fed and clothed a large part of themselves, the tax burden on the peasantry was stabilised and then lightened, and the government acquired the standing of a producer rather than a parasite, no small asset in a countryside schooled to regard every uniform as a mouth. Its political substance was larger still. Production was organised as a movement, with labour heroes, emulation drives, and the cooperative forms, mutual aid teams exchanging work at the seasons’ peaks, that carried the party’s presence into the agricultural calendar itself. The episode refines the second explanation rather than merely illustrating it. What won support in the border region was not a schedule of reforms handed down but a demonstrated capacity to solve the community’s material problems by organising the community, and the distinction between reform as policy and organisation as method, which the dispute’s third term insists on, is visible here in miniature, inside the reform explanation’s own home ground.
Religion in the base areas: forbearance as policy and as evidence
The party’s treatment of religion in the base areas is a small subject with a disproportionate evidentiary value, because it shows the united front’s discipline operating where doctrine pressed hardest in the opposite direction. The movement’s theory regarded religion as a consolation of the oppressed destined to fade with the conditions that required it, and its soviet period practice had included the confiscation of temple property and campaigns against superstition that alienated the very peasants the soviets courted. The wartime line reversed the practice: freedom of religious belief was proclaimed in the base area constitutions, temples and churches within the party’s territory were in the main left to their congregations, and the party’s organisers were instructed to work with, rather than against, the religious calendars and authorities of the villages, enrolling monks and priests in the patriotic associations where they would serve. Missionaries who visited or resided in the base areas reported, in letters that reached the outside world, a forbearance they had not expected from an atheist party, and their testimony entered the wartime record as independent corroboration of the border region’s moderation. The forbearance was instrumental, and the article’s method requires the word: it cost the party nothing it valued in wartime, it purchased the neutrality or support of constituencies the enemy might have mobilised, and it was withdrawn in later periods when the party’s situation no longer required it, a sequel outside this article’s period. But instrumentality is not the same as insincerity at the level of the individual cadre, many of whom administered the policy as a genuine pluralism, and the episode’s real lesson is the one the whole study of the base areas teaches: the party’s moderation in the war years was a system, consistently applied across rent, religion, elections, and culture’s limits, by an organisation that understood its own appeal as a wartime construction and policed the construction’s boundaries with the same discipline it applied to its armies. Systems of that coherence are not accidents of enthusiasm. They are what the third explanation means by organisation, wearing its most attractive face.
Women in the base areas: emancipation within limits
The position of women in the base areas tests every explanation against a constituency that was half the population and whose mobilisation the party actively sought. The movement’s programme, from the soviet period onward, promised women emancipation: the marriage law of the Jiangxi Soviet had proclaimed freedom of marriage and divorce among the most radical provisions enacted anywhere in the China of its day, and the base area governments of the war years maintained women’s associations, literacy work, and the spinning cooperatives that gave peasant women an income handled by their own hands, often the first money of their lives that was theirs. The practice was everywhere tempered by the party’s dependence on the peasant household, whose male head it could not afford to alienate in wartime. Divorce, the sharpest edge of the emancipation programme, was restricted in application where soldiers’ marriages were concerned, and cadres learned to mediate family conflict toward reconciliation rather than liberation when the two conflicted; the women’s association in the village spent much of its energy on production and on the support of soldiers’ families, the roles the war required, rather than on the transformation of the family itself. The scholarship on the subject, from Kay Ann Johnson’s study of women and the family revolution to Gail Hershatter’s oral histories of the generation that lived the transition, converges on a judgement that parallels the article’s larger settlement. The changes were real: women spoke in public, organised, earned, and in the land reform acquired land certificates in their own names in many districts. The changes were also instrumental and reversible, advanced where they served production and mobilisation and restrained where they threatened the household order on which the party’s rural power rested. A constituency courted as producers and organisers, and emancipated exactly as far as the court’s purposes ran, is a fair emblem of the base area system as a whole, and the women’s record is among the strongest evidence that the system’s benevolence and its instrumentality were not two facts but one.
Medicine in the base areas: the surgeon from Canada and the infrastructure of care
The base areas’ medical services supply the second explanation with some of its most affecting evidence, and the story’s central figure is an unlikely one. Norman Bethune, a Canadian thoracic surgeon of communist conviction, reached the border region in 1938 and spent the remaining months of his life building surgical services for the Eighth Route Army in the Shanxi mountains: mobile operating units that worked close behind the fighting, training courses for the orderlies and village doctors who were the army’s real medical corps, and a ferocious insistence on standards in conditions that mocked them. He died in November 1939 of an infection contracted at the operating table, and Mao’s funeral eulogy, In Memory of Norman Bethune, became one of the most memorised texts of the Yan’an canon, its praise of a foreigner who served the cause without thought of himself held up to every cadre as the standard of conduct. The eulogy’s canonisation was an act of the organisational method, the manufacture of an exemplar; but the medicine beneath it was real, and its structure is the point. The base areas could not produce surgeons; they produced, in quantity, the barefoot infrastructure of care beneath the surgeon, medics in every company, health workers in the associations, campaigns against the flies and the contaminated wells that killed more villagers than any army. A peasantry whose experience of every previous power included no provision for its sickness met, in the communist districts, a state that treated its health as a public charge, however rudimentary the treatment. The historians who weigh the reform explanation count the clinics with the cooperatives and the schools among the deposits of allegiance, and the sceptics of the third school reply, correctly, that the health register was also a census. Both are describing the same clinic, and the villager who attended it did not experience the descriptions as alternatives.
Education in the base areas: the school as an institution of the new order
The school was the base area institution through which the largest number of villagers met the new order in its benign aspect, and its place in the explanations is as the second factor’s most durable deposit. The border region and the larger base areas built primary schooling where the old order had provided almost none, in winter schools that used the agricultural slack season to teach characters to adults, in village primaries whose teachers were often the association’s own literate youth, and in textbooks composed at Yan’an that taught reading through the vocabulary of the new society, the rent receipt, the cooperative, the election, the resistance. The curriculum’s content was the movement’s, and no reader should mistake the school for a neutral charity: it taught the peasant child and the peasant mother to read the notices of the government that had taught them, and literacy campaigns were enumerated among the party’s achievements in its own reports with the pride of a power that knew what it was building. But the instrumentality, once again, does not cancel the substance. A generation of villagers acquired in the base areas the literacy their parents had been denied by every previous power, and acquired it from the party, an association between the movement and the most unambiguous good a poor society can receive that no subsequent disillusion wholly dissolved. The Nationalist state had promised mass education in its programmes and delivered it in its cities; in the villages it governed, the schoolhouse was commonly the tax office’s annex or did not exist. Here as everywhere in this history, the competition was not between the party’s offering and a rival offering of the same kind, but between an organisation that delivered a service and a state that delivered a demand, and the educational record is the second explanation’s evidence in its most concrete form, a form in which even the third explanation’s adherents concede the fact while disputing, as they dispute everything, whether gratitude or the register in which the school enrolled its pupils did the heavier work.
The invisible infrastructure: intelligence, couriers, and the wireless
An insurgent state lives or dies by its information, and the communist movement’s information system, invisible in most narratives, was among its most consequential constructions. At the summit, the party’s intelligence work penetrated its enemies to a degree the Nationalists discovered only in retrospect: communist agents served inside the Nationalist military and police apparatus for years, among them the secretary whose warning in 1931 allowed the party’s Shanghai leadership to escape a raid by hours, and the officers on Nationalist staffs whose reports reached Yan’an through the war and the civil war, in some cases influencing the operational picture on which communist strategy was built. At the base, the system ran on humbler instruments: the courier lines that carried documents between base areas in the seams of the occupation, the wireless sets, few and precious, that linked the major headquarters, and the village intelligence web, the same associations and militias that governed by day, which reported every stranger, every troop movement, every change in the garrison towns to the district committee. The Japanese counterinsurgency understood the web’s importance and aimed its blockhouse lines and its census campaigns at breaking it; the Nationalist armies of the civil war marched, by communist accounts and by their own commanders’ complaints, in a fog of hostile information, their columns observed from the moment they moved, their intentions sold or signalled ahead, while the communist commanders operated with a picture of the enemy’s dispositions that their intelligence services and the population jointly maintained. The point belongs to the weighing because it refines the third explanation’s meaning once more. Organisation, in this register, was not only the power to compel the village but the power to know: to know the village, the enemy, and the war, better and faster than any rival, and in contests between organisations of otherwise comparable courage and numbers, knowledge of that quality is not an advantage among others. It is the advantage that arranges all the others, the one that lets a weaker force be stronger at the point it chooses, which is the whole art of the war the party fought.
The cadres: the human material of the expansion
Between the party’s programme and the village stood the cadre, and no explanation of the victory is complete that treats the cadre as a transparent medium. The wartime party trained its organisers in a network of institutions centred on Yan’an, of which the Anti-Japanese Military and Political University, Kangda, was the emblem: short courses, conducted often in caves and courtyards, that took students, workers, and peasant activists and returned them to the base areas as political workers drilled in the party’s methods, its discipline, and its current line. The cadre who arrived in a new district was typically young, often an outsider to the province, and equipped with a repertoire: the investigation of local conditions, the identification of activists among the poor, the staging of the first rent reduction or accusation meeting, the building of the association and the cell, the patient conversion of a village’s grievances into an organisation that answered to the district committee. The quality of these men and women, their willingness to live at the peasant standard they administered, to share the dangers of the districts they organised, and to submit to a discipline that punished corruption with a severity no rival administration approached, was a resource the party’s competitors could not copy, because it was not a technique but a formation, the product of selection, training, rectification, and belief. The third explanation, organisation, is an abstraction until it is given this human content: organisation meant, in practice, that in a hundred thousand villages there was a person, or a small committee of persons, whose life was the party’s presence, and that the Nationalist state, for all its ministries, had no equivalent person in the village at all, only the tax collector and the press gang who visited and departed. Wars of this kind are decided by who is present when nothing is happening, and on that measure the contest was unequal from the start.
Who carried the party’s programme into the villages?
Trained organisers, many of them young students and activists schooled at Yan’an’s institutions such as the Anti-Japanese Military and Political University and sent out to the base areas with a standard repertoire: investigate conditions, find activists, stage the first reforms, build the association and the cell, and stay, living at the standard they administered.
The armed margins: bandits, secret societies, and the Red Spears
The countryside the party organised was not empty of armed organisation awaiting its arrival, and its dealings with the armed margins of rural society form a chapter the explanations neglect at their cost. Rural China sustained a dense undergrowth of self-armed bodies: bandit bands in the hills, the lodges of the secret societies, and the village defence associations of which the Red Spears of Henan and the north were the most formidable, peasant militias raised to protect villages against soldiers and bandits alike, animated by a fierce localism and often by religious sanction. Every power in China had to reckon with this undergrowth, and the reckonings differed instructively. The warlords bought or exterminated it as convenience dictated. The Nationalists registered it where they could and feared it where they could not. The Japanese attempted to use it against the communists and were used by it in turn. The communists’ method was absorption where possible and destruction where necessary, conducted with the patience of an organisation that expected to stay: bandit bands were recruited with their leaders honoured and their men re-educated in the ranks, or isolated and annihilated when they preyed on the base areas’ population; the Red Spears were negotiated with, supplied against the Japanese, and gradually drawn into the militia system the party directed, their village defence impulse converted into the base area’s first line. The process was the united front in miniature and the third explanation in the field: the party did not merely appeal to the peasantry over the heads of its existing organisations but digested those organisations, taking their men, their local knowledge, and their legitimacy into its own structure, and leaving the village, at the end, with one armed authority where it had known several. A peasantry that supported the movement had often begun by being organised by somebody else, and the party’s victory in the countryside was in part a victory in a competition among organisers, most of whom were not communists at all when the competition began.
The geography of survival: why the north held
The map of the base areas was not drawn by doctrine, and its asymmetries carry information the explanations need. The movement’s great wartime strongholds lay in the north and centre: the plains and mountains of Hebei, Shanxi, and Shandong, the river valleys of the New Fourth Army’s central China, districts where Japanese garrisons held the towns and railways and the countryside between them was left to whoever could organise it. In the south, by contrast, the party’s wartime presence remained thin and embattled. The southern bases had been the movement’s cradle, but the guerrillas who remained there after the main forces marched north in 1934 had been reduced to scattered bands in the mountain borderlands, and the New Fourth Army incident of 1941 fell on the southern movement’s headquarters at the moment of its reconstruction. Gregor Benton’s scholarship on the mountain bases of the south recovered this theatre and its lesson: the southern guerrillas survived by the same methods as the north and expanded far less, because the conditions differed in kind, a stronger Nationalist presence in the unoccupied south, denser state and lineage organisation in the villages, and no invader’s vacuum of the northern type to fill. Geography in this history is never merely terrain. The north China plain was invadable, occupiable at its nodes, and ungovernable in its immensity by an army that counted its effectives against an empire’s commitments elsewhere; the loess hills of the border region were a fortress the enemy could not profitably storm; the central river country was a chessboard of competing authorities. The party’s method fitted each of these environments because the method was, at bottom, a way of converting the state’s absence into the party’s presence, and the method’s results varied precisely as the state’s absence varied. A theory of the victory that cannot explain why the same party, with the same doctrine and the same calibre of cadre, conquered the north and merely survived in the south is missing its geographical term, and the three-factor settlement, which weights its factors by region, is built to supply it.
The settlement tested: three villages
The settlement can be given a final, concrete form by applying it to three villages of the imagination, each assembled from the documented record of its type. The first is a village of the north China plain in 1939, within reach of a Japanese garrison town, its young men liable to be seized for labour, its grain requisitioned, its former authorities fled or collaborating. Here the party team that slips in by night is received first as the resistance: the nationalist factor opens the door, the association and the militia give the reception a structure, the rent reduction that follows in the next season gives it a material interest, and the register of households that the team quietly compiles gives it a permanence that will outlast the emotions of 1939. The second is a village of the border region in 1943, unoccupied, taxed in grain by a government seated a day’s walk away. Here no occupier threatens, and the party must make its claim entirely in the currency of the second factor: the fair tax assessment, the cooperative’s seed loan, the school, the production campaign’s labour exchange, with the organisation present throughout as the machinery of delivery and the rectification-disciplined cadre as its face. The third is a village of central China in 1944, which has seen the New Fourth Army, a puppet administration, and a Nationalist guerrilla band in successive seasons, and has learned to keep its options open and its grain hidden. Here neither patriotism nor gratitude can be presumed, and the party’s power rests on the third factor nearly unmixed: the cell that never left, the intelligence that knows which household fed which army, the credible promise of protection and the credible threat of punishment, until habituation and the land reform that follows victory convert a garrisoned allegiance into something warmer. The three villages are one country, and the party that won it had to win all three, in the same decade, with the same organisation speaking three dialects of its appeal. That is the three-factor settlement in its working form, and it is why the article has refused, from its first page, to let any single explanation stand for the whole.
The Seventh Congress: the party that emerged from Yan’an
The institutional form of the victor was fixed in the spring of 1945, at the party’s Seventh Congress in Yan’an, and the congress deserves notice as the moment the movement that would take China presented itself to itself in completed form. The congress met in a Yan’an still blockaded and still at war, with delegates who had crossed enemy lines from every base area, and it performed the unifications the rectification and the Resolution had prepared. Mao’s political report, On Coalition Government, set out the programme for the postwar settlement the party proposed, a coalition from which it expected, in the event, to be excluded and against whose exclusion it was preparing the civil war. Liu Shaoqi’s report on the party constitution wrote the thought of Mao Zedong into the party’s guiding doctrine, the formal seal of the sinification the Zunyi Conference had begun in practice a decade before: the universal theory, the constitution declared, had been united with the concrete practice of the Chinese revolution, and the union had a name and an author. The congress elected the leadership that would govern China for a generation, and it did so in an atmosphere, recorded by the delegates themselves, of a confidence wholly unlike the desperation of the Long March a decade earlier: the party of 1945 knew it was a candidate for power, organised its succession, codified its doctrine, and composed its history, in the manner of a state preparing its accession. For the purposes of the dispute, the congress is the organisational explanation’s ceremonial exhibit. Whatever moved the peasantry, nation, reform, or fear, the instrument that converted movement into victory was completed here, a party of more than a million members by its own report, with a doctrine, a history, a leadership, and an army, every element of which had been standardised in the Yan’an years. The other explanations describe the fuel. The congress is a portrait of the engine, assembled and idling, in the last spring of the war.
The questions the victory left open
The victory closed the military question and left the theoretical ones open, and a short inventory of the openings is the right preparation for the personas and the questions that close this article. The stage theory under which the party had fought promised a new democracy of some duration, a mixed economy and a coalition politics within which the transition to socialism would mature; the transition, when it came in the middle 1950s, came with a speed that surprised the theory’s own custodians and that historians cite on both sides of the classification debate, as evidence that the socialist destination had governed the journey all along, or as evidence that the party-state, once built, obeyed its own logic of expansion regardless of the stages it had promised. The peasantry, the revolution’s main force, found within a decade that the land it had received was being gathered into collectives, a sequel that casts its own retrospective light on the wartime bargain and that the second explanation, with its emphasis on the party’s responsiveness to peasant interest, has the most difficulty absorbing. The party’s relationship to the doctrine it professed remained the one the Zunyi Conference had settled in practice: the theory was universal, its application was to be Chinese, and the authority to say what the application required rested with the party’s own leadership, an arrangement whose strengths the war had demonstrated and whose dangers the warning about substitutionism had named. And the historiographical dispute itself remained open in the form this article has mapped, because the archives that might settle the local weights of the three factors continue to be opened unevenly, and the dispute’s persistence is not a scandal but a measure of how much of the evidence a single human lifetime of scholarship can hold. The reader who has followed the argument to this point possesses the dispute in its working shape, the settlement the article defends, and the criteria by which the theoretical question must be argued, which is the equipment the subject, honestly taught, exists to supply.
One further question, of a different order, is sometimes asked of this history and should be answered before the article closes: why a dispute among scholars about a war concluded generations ago retains its heat. The answer is that the event under dispute is not only an event. It is the founding act of a state and the standing test case of a doctrine, so that every generation’s argument about why the communists won is also, inescapably, an argument about what their victory meant and what it licensed. The historian cannot drain that charge from the subject and should not pretend to have done so. What the historian can do, and what this article has attempted, is to keep the charge visible while the evidence is weighed, to attribute every claim to its maker and its archive, and to leave the reader in possession of the means of disagreeing. A subject that founded a world power and refounded a theory will not be argued about quietly. It can still be argued about honestly, and the difference between the two is the whole of the method on display in these pages.
The official history and its custodians: the revolution remembered by its victors
The historiography of the revolution inside China followed the arc the Resolution of 1945 had laid down, and a brief account of its custodianship completes the article’s map of the evidence. In the decades after 1949 the party’s history institutes and its university departments produced the authorised narrative in successive codifications: the history of the movement as the history of the correct line’s triumph, documented from the party’s own archives, taught in the schools, and guarded by the professional boundary that treated deviation from its judgements as a political act. The narrative was not static within its frame: the emphases shifted with the politics of the custodians, the role of this or that leader enlarged or diminished as his standing rose or fell, and whole episodes, the Futian purges among them, receded from the record the institutes published. Foreign scholarship worked in the space this custodianship left: denied the archives, it built the dispute this article maps from the hostile and the published records, and its findings, Johnson’s and Selden’s in their different ways, were received in China, when they were received at all, as specimens of foreign misunderstanding. The archival turn of the 1980s, on which the third explanation rests, was itself a product of a change in the custodians’ calculations, a controlled opening in which local records became available to Chinese and foreign scholars alike while the central archive remained guarded, and the unevenness of that opening is written into the third explanation’s geography, strongest where the local record was richest. The student of the dispute should hold the whole custodial history in mind as a single caution: the evidence for every explanation of this revolution passed, at some point, through the hands of an institution with an interest in the answer, and the weighing this article performs is possible only because the interests were plural, the party’s, the fallen regime’s, the occupier’s, the observers’, and the custodians never succeeded, among them, in making the record speak with one voice.
The Fairbank frame: the revolution as a century’s organising idea
The dispute this article maps was conducted within a frame set by the dean of American China studies, and the frame itself deserves a paragraph of historiography. John King Fairbank, who built the Harvard school and trained a generation of the field, proposed that the Chinese Revolution be understood not as the event of 1949 but as the name for China’s whole modern transformation, the century-long process by which an ancient civilisation met the modern world and remade itself under the impact, with the communist victory as the process’s most radical phase rather than its essence. The frame had consequences for the dispute’s conduct. It directed attention from the party’s victory to the society’s motion, which is the intellectual space in which Johnson’s peasants and Selden’s villagers could both be protagonists; it licensed the social history from which the archival turn later grew; and it drew the criticism, from the field’s more political scholars, that a revolution dissolved into a century loses the event in which men chose, fought, and killed, and with it the question this article has kept at its centre, why these men won. Fairbank’s own wartime service in China and his later defence of his persecuted China Hands gave the frame a biographical authority in America that it never quite acquired elsewhere, and the European and Chinese scholarships of the revolution proceeded on partly different tracks, the French tradition from which Bianco wrote being readier with the peasant war’s vocabulary and the Chinese official tradition being bound, as the Resolution’s section showed, to the victor’s brief. The article’s settlement is offered in awareness of the frame and partly against it: the revolution is treated here as an event with causes that can be weighed, not only as a phase of a civilisation’s long adjustment, because the men and women of the base areas experienced it as an event, with dates, decisions, and dangers, and because an explanation that cannot say why the event happened when and as it did has explained the century at the price of the war.
The missing voice: what the peasants’ own testimony can and cannot say
Every explanation in the dispute claims to know what the peasantry thought, and the peasantry’s own testimony is the scarcest evidence in the field, a scarcity that deserves a section because it disciplines all three. The villager of the 1940s left almost no direct record: illiterate in the main, unconsulted by the archives that preserved his rulers’ papers, he enters the documents as the object of other people’s reports, the Japanese intelligence officer’s estimate, the cadre’s work summary, the observer’s anecdote. Two bodies of testimony partially repair the silence, each with its own distortion. The first is the participant account written close to the events, of which William Hinton’s Fanshen, his record of the land reform in a Shanxi village where he served as an observer with a United Nations relief team in 1948, is the most substantial in any language: a village-level narrative of the classification, the struggle meetings, and the redistribution, written by a sympathiser whose access was extraordinary and whose sympathies the reader must price in. The second is the oral history gathered decades afterward, the life stories recorded by scholars such as Gail Hershatter from the women of the revolution’s generation, testimony of irreplaceable texture about work, marriage, and the arrival of the new order in a household, and subject to the double filter of memory’s own revisions and of the authorised narratives within which the remembering was done. Neither body of testimony can settle the dispute, since testimony gathered by a sympathiser in 1948 and testimony gathered from survivors in their old age will weight gratitude and grievance differently by construction. What they jointly exclude is the condescension of supposing the peasantry’s choices were reflexes. The villagers of these accounts calculate, hedge, conceal grain, volunteer sons and withhold them, embrace the association and evade the register, with a tactical intelligence fully equal to the organisations competing for them, and any explanation of the victory that cannot accommodate a peasantry thinking is refuted by the testimony even where the testimony cannot crown a winner.
The evidence problem: an archive built by the belligerents
A final methodological section is owed before the shelf itself, because the dispute’s shape cannot be understood without knowing how its evidence was made and kept. The documentary record of the Chinese Revolution is, to an unusual degree, the creation of the war it records. The party’s internal documents survive in three great custodies, each with its politics: the party’s own archives, opened selectively and late; the captured records assembled by the Nationalist security organs and carried to Taiwan, where they served first as intelligence and then as scholarship; and the collections microfilmed or purchased by foreign institutions, the Japanese army’s records chief among them, which gave Johnson his evidence and set the first explanation’s terms. The border region’s publications, Selden’s base, were the output of a government conducting a war of legitimacy as well as of arms, and were written to be believed. The county archives that powered the third explanation became available when a later Chinese state’s calculations permitted, and their coverage is densest exactly where the party’s administration was densest, so that the archival turn has its own geography of light and shadow. Memoir literature, abundant on all sides, was written by survivors arranging their pasts in the light of outcomes, and the Resolution of 1945 hung over the communist side’s recollections as an authorised version against which memory was checked. None of this makes the history unknowable; the three custodies corroborate one another at enough points to sustain the narrative this article has given, and hostile archives agreeing are the strongest evidence social history possesses. It does mean that the dispute’s persistence is partly an artefact of custody, and that the settlement’s humility about weights and mixtures is not a rhetorical posture but the only epistemology the record affords. The reader who has absorbed this has the last piece of equipment the subject requires: a sense of why honest scholars, reading honestly, have produced three explanations of one victory, and why the fourth thing, the weighing, must be done by each reader on the evidence as it stands.
Sources and their standing: the shelf behind the dispute
The dispute this article maps rests on a shelf of works whose differing evidence bases explain their differing conclusions, and a reader who wishes to go further should know the shelf in order. Chalmers Johnson’s Peasant Nationalism and Communist Power of 1962 stands first, the nationalism thesis in its classic form, built from Japanese records and indispensable for its question even where its answer is resisted. Mark Selden’s The Yenan Way in Revolutionary China of 1971 is its great rival, the reform thesis built from the border region’s record, and it should be read in its revised edition of 1995, where the author’s own reconsiderations, on nationalism and on coercion alike, are part of the evidence for the settlement. Chen Yung-fa’s Making Revolution of 1986 opens the archival era, and the regional studies that followed it, by Kathleen Hartford, Odoric Wou, David Goodman, and others, supply the local variation on which any honest weighing depends. Lucien Bianco’s Origins of the Chinese Revolution remains the strongest scholarly statement of the objection, that the revolution’s energy was the peasantry’s and its direction the party’s. For the Nationalist side, Lloyd Eastman’s studies of the Nationalist regime’s decay are the standard, severe from within the regime’s own record. Hans van de Ven’s War and Nationalism in China re-examined the war itself and restored the Nationalists’ military effort to the picture, while Gregor Benton’s work on the New Fourth Army and the mountain bases of the south recovered whole theatres the northern-centred literature had neglected. Suzanne Pepper, on the civil war’s urban dimension, traced the contest for the cities and the educated public that accompanied the military collapse. The memoirs and the wartime reportage, Edgar Snow’s account of his 1936 visit to the border region above all, are sources of a different order, evidence for what the movement wished to seem and for what a sympathetic observer could see, to be used with the cautions their genre requires. A shelf of this kind, in which every major work announces its archive in its argument, is the real syllabus of the subject, and the dispute is best learned, as this article has tried to teach it, by reading the works against one another rather than by memorising the verdicts extracted from them.
For students and exam candidates
The examinable core of this article is the dispute, and the grid is the instrument for mastering it. Build every answer on this topic in the grid’s shape: state the three explanations with their authors and dates, Chalmers Johnson in 1962 for peasant nationalism, Mark Selden in 1971 for the Yan’an Way, and the archival scholarship associated with Chen Yung-fa’s study of 1986 for organisation and coercion; assign each its evidence; then weigh them by region and period rather than electing one. On the theoretical question, state your criterion before your verdict, since the classification turns entirely on whether the marker is being shown the agent, the programme, the leadership, or the outcome. Keep the three explanations and the four criteria as permanent study notes in the VaultBook Marxism study notes, where the grid can be stored beside the rest of this cluster’s frameworks, and test the question types against past papers through the ReportMedic UPSC previous-year question explorer, which lets a candidate see how often, and in what wording, the examiners return to the revolution’s causes and its character.
For teachers and seminar leaders
The three explanations grid is designed to be taught from directly. Set the three classic works against one another in sequence, Johnson’s occupier’s-eye view, Selden’s border region, and a chapter of the archival scholarship, and have the seminar identify, for each, the archive beneath the argument before debating the argument itself; students discover faster from the evidence’s provenance than from any summary that the dispute is about records as much as about China. The classification question makes the strongest single discussion in the cluster: divide the room by the four criteria, agent, programme, leadership, and outcome, require each group to argue the verdict its criterion dictates, and let the room discover that it has produced four defensible answers to one question. The materials in this article, the three villages exercise above all, convert readily into an assessed essay: apply the settlement to one base area and defend the weighting the local record supports.
For researchers, historians, and journalists
The open frontier of this subject is local and archival. The weights of the three factors remain unspecified for most base areas and most phases of the war, and the county-level record, where it survives and is accessible, is the only evidence that can specify them; work that reconstructs a single district’s wartime passage through all three custodies of the archive, the party’s, the Nationalist security organs’, and the occupier’s, will do more for the dispute than another general restatement of it. Two cautions govern reporting on this history. The casualty literature is unusable without its provenance, for the reasons the article’s figures section gives, and any figure carried into print should carry its basis with it. And the revolution’s categories, peasant nationalism, the Yan’an Way, the mass line, are terms of art with scholarly owners and dates, not synonyms for the phenomena: precision about which scholar’s term is in use, and about the criteria behind any classification of 1949, is the difference between writing about the dispute and merely re-enacting it.
Frequently Asked Questions
Q: Why did the Chinese Communists win?
They won because three factors worked together in proportions that varied by place and period. Japanese occupation generated a nationalist response, and the party, entrenched behind the enemy’s lines with forces and governments of its own, was positioned to organise and lead it. In its base areas the party governed with a programme of rent and interest reduction, cooperatives, and comparatively honest administration that gave peasants material reasons for support. And its organisation, cells, associations, classification, surveillance, and coercion, built and held allegiance where sympathy alone would not have survived. Against a Nationalist state weakened by inflation, corruption, and predatory conscription, the combination proved decisive.
Q: Was the revolution driven by nationalism or social reform?
The classic dispute poses the two as rivals: Chalmers Johnson argued in 1962 that peasant nationalism created by the Japanese invasion drove the party’s growth, while Mark Selden argued in 1971 that the party’s social and economic reforms in its base areas earned its support. The settlement this article defends refuses the choice. Nationalism predominated in the occupied districts of the north, reform in the secure border region around Yan’an, and both were delivered to the village through an organisation that neither scholar foregrounded. The honest answer is that the proportions changed by region and by phase of the war.
Q: What was the Long March?
The Long March was the Red Army’s strategic retreat of 1934 to 1935, from the Jiangxi Soviet, which Nationalist encirclement campaigns had made untenable, across western China to northern Shaanxi, where Yan’an became the survivors’ headquarters. The remaining communist forces reached the region by 1936. The march destroyed much of the army that undertook it, but it saved the movement’s core, and in its course, at Zunyi in January 1935, the leadership was remade, Mao Zedong rising to effective command and the doctrine of protracted rural war becoming the settled line the party followed to victory.
Q: How did the party grow during the war with Japan?
Across the war years the party’s membership multiplied to more than ten times its former scale, growing from the Yan’an remnant into a movement administering base areas across the northern and central provinces. The mechanism was the second united front’s shelter combined with the occupation’s geography: Japanese forces held cities and railways but could not administer the countryside, and into that vacuum the party sent organisers behind its guerrilla units to erect administrations, enrol the villages in associations, raise militias, and levy regular taxes. Growth ran fastest where the occupation pressed hardest and slowest where the party’s own presence was thinnest, and it survived a severe contraction in 1941 and 1942.
Q: Was the Chinese Revolution a Marxist revolution?
The answer depends on the criterion, and the article’s discipline is to state the criterion first. By the agent, no: the force that made it was a peasant army, and the industrial working class in whose name it acted was a small fraction of the nation. By the wartime programme, only in part: the programme of 1937 to 1945 was national resistance and moderate reform, not socialism, and was not presented as such. By the leadership’s doctrine, yes completely. By the outcome, yes in the end: the victors abolished the landlord class, built a party-state, and moved property toward socialised ownership within a decade. The classification is a choice among criteria, to be defended as such.
Q: What role did the Comintern play in China?
The Comintern midwifed the party’s creation, financed and trained its early cadres, gave it the Leninist party form that proved the decisive import, and directed its strategy for the first half of its life, frequently to catastrophic effect. Its line imposed the first united front and maintained it into the catastrophe of 1927; its insurrectionary orders afterward repeated the error in reverse; its returned students presided over the loss of Jiangxi. The party’s recovery began with its emancipation from that direction at Zunyi in 1935, and its doctrine of the sinification of Marxism was the formal record of the escape.
Q: What was the united front strategy?
The united front was the party’s strategy of alliance with wider forces against a principal enemy, practised twice in opposite forms. Under the first arrangement, from 1923, party members joined the Nationalists as individuals and built the alliance’s mass movements, only to be destroyed when it collapsed in 1927, having neither forces nor ground of their own to fall back on. Under the second, from 1937 against Japan, the party kept its own army, governments, and base areas inside the alliance, moderated its programme to rent and interest reduction, and used the front’s shelter to spread through the countryside behind the Japanese lines.
Q: Did coercion play a role in rural support?
Yes, and the archival scholarship has made the role impossible to omit. Alongside the benefits the movement delivered, reduced rents, protection, comparatively honest administration, stood the apparatus that classified every household, watched the village through its own associations, controlled the distribution of scarce goods, and punished resistance, whether from collaborators, from landlords who fought the settlement, or from villages that refused the tax and the recruiting officer. How much of any district’s allegiance was enthusiasm, how much calculation, and how much fear cannot now be separated, and the record excludes the simple picture in either direction: neither a plebiscite of gratitude nor a countryside held down by terror alone.
Q: How did the Chinese Revolution influence other movements?
Its largest influence was as a method. The victors’ doctrine of people’s war, a disciplined party, a peasant army, rural bases run as embryonic states, a broad united front, and protracted war trading space for time, was studied as a manual by insurgencies across the agrarian world. It was applied with success in Vietnam, where Chinese advisers and supplies supported the victors of Dien Bien Phu, and it failed in Malaya, where the party could not root itself in the majority peasantry. Its deeper effect was doctrinal: it broke the monopoly of the Russian road and taught movements that doctrine was to be fitted to local conditions.
Q: What do historians disagree about regarding the revolution?
They disagree principally about causes and about classification. On causes, the dispute runs among three explanations of the party’s wartime growth: peasant nationalism, the answer Chalmers Johnson built from Japanese records in 1962; social and economic reform, Mark Selden’s answer of 1971 from the border region’s record; and organisation with coercion, the answer the local archives supported from the 1980s. On classification, they disagree about whether a revolution made by peasants under a moderate wartime programme counts as a Marxist revolution, a question that turns on the criterion chosen. They also dispute the weight of Soviet assistance in Manchuria and the scale of the war’s violence, on which this article states no figures.
Q: How does the three-factor settlement grade an explanation of the victory?
It grades an explanation by its coverage of the record’s variation. An explanation that accounts for the occupied north China plain but not for the unoccupied border region, or for the expansion of the late 1930s but not for the contraction of 1941 and 1942, or for the civil war that followed the occupier’s departure, is marked down for the territory it cannot reach. The settlement itself claims no single mechanism: it holds that nationalism, social reform, and organisational coercion all operated, in weights set by region and period, and it grades rival accounts by how much of that variation they can absorb without strain.
Q: What should a student separate between a peasant war and a revolution led by a Marxist party?
The student should separate the force from the direction. A peasant war is identified by its agent and its horizon: peasants in arms for land, tax relief, and the removal of oppressors, issuing historically in failure or in a new dynasty over an unchanged social order. A revolution led by a Marxist party is identified by what its leadership does with victory: the landlord class abolished outright as a category of the social order, a party-state constructed that reached into every village, and property relations transformed. The Chinese case was the first by its agent and the second by its direction, and confusing the two descriptions is the standard error.
Q: Why did the party expand fastest during the war of resistance?
Because the war created the three conditions of its growth simultaneously. The Japanese occupation destroyed the Nationalist administration across the countryside of the north and centre while lacking the manpower to replace it, opening the vacuum the party filled. The invasion generated the nationalist response that made the party’s organisers welcome as the resistance’s local form. And the second united front gave the expansion legal and political shelter, while the party’s moderation of its programme to rent and interest reduction removed the obstacle that its earlier land revolution would have placed in the way of the village’s propertied strata.
Q: How does the three explanations grid order the causes?
The grid does not rank the causes in a single order, because the settlement behind it holds that their order changes with the evidence’s address. It arranges the three explanations side by side, nationalism, reform, and organisation with coercion, and for each one it specifies how the cause is supposed to work, what records the claim rests on, where the claim runs strongest, and what it leaves out. Used properly, the grid converts a quarrel about which cause was real into a specification of where and when each cause bore the weight, which is the form in which the record can actually decide among them.
Q: What separates a wartime programme from a socialist programme?
Their objects and their promises. The wartime programme pledged resistance to the invader, reduced rents and capped interest, honest administration, and a democratic order that protected private capital and left the landlord’s title intact; it was built to unite the widest possible alliance, and it postponed socialism by design to a later stage. A socialist programme promises the socialisation of the means of production and the abolition of the exploiting classes, and the party adopted measures of that kind only after its victory, in the land reform’s completion and the collectivisation and nationalisations of the following decade. The distinction is the party’s own, drawn in its Yan’an texts.