The Paris Commune lasted seventy two days, and it has been argued about for a century and a half. Between the middle of March 1871 and the end of May, an elected council governed the largest city in France in defiance of the national government at Versailles, passed a body of decrees that ranged from the remission of rents to the separation of church and state, and was then destroyed in a week of street fighting and mass execution that remains one of the most contested episodes in European history. Almost everything about the episode is disputed, and the disputes are not ornaments on the story but the story itself. It is disputed who governed Paris, since the council that ruled the city was dominated by Blanquists and Proudhonists rather than by members of Karl Marx’s International. It is disputed what Marx himself thought, since he warned against the rising before it happened, defended it completely once it existed, and described it with cool detachment a decade after it fell. It is disputed how many people died in its suppression, since the figure repeated for a hundred years has been challenged by archival research that argues for a far lower toll. An account of the Commune that smooths over those three disputes has discarded the most valuable material the episode offers, and this article is organised to keep all three in view from the first page to the last.

One organising claim runs through everything that follows, and it can be named at the outset. The three-position sequence is the pattern formed by Marx’s movement from warning to defence to sober reassessment, and it is the key to every argument that has since been built on the Commune in his name. Before the rising, in an address issued by the General Council of the International in September 1870, Marx told the workers of Paris that an insurrection against the government then besieged by Prussia would be desperate folly. During the rising and in the days after its fall, he wrote The Civil War in France, the most passionate defence of the Commune produced by any major thinker, and drew from the episode a conclusion about the state that he and Engels inserted into the preface of the Communist Manifesto the following year. A decade later, writing privately to a Dutch correspondent in 1881, he described the Commune as the rising of one city under exceptional conditions and its majority as in no sense socialist. Each of the three positions is genuine, each is documented, and each is regularly quoted in isolation by writers who want Marx as either the Commune’s prophet or its embarrassed critic. Quoted together, they show something more interesting than either portrait: a theorist watching an event he had not chosen, revising his account of the state in response to it, and refusing, when the commemorative mood arrived, to pretend the event had been what it was not.

The episode also carries a second discipline that this article observes throughout. The Civil War in France is Marx’s account of the Commune, written while the fighting’s outcome was still fresh and intended in part as an act of solidarity with the defeated. It is a remarkable document, and parts of it rest on reports that later research has corrected. It is therefore treated here as a source to be read against the record rather than as the record itself. What the Commune actually decreed, who actually sat on its council, what its armed forces actually did, and how many people actually died are questions for the archives and for the historians who have worked in them, from Prosper Olivier Lissagaray, who fought on the barricades and wrote the first great history, to Robert Tombs, whose study of burial records reopened the casualty question in the early twenty first century. Where the evidence divides, the division is reported. Where a figure is contested, both figures and the grounds of each are given. The Commune has suffered more than most events from writers who knew in advance what it had to mean, and the surest way to repeat their error is to decide the verdict before presenting the case.

The road to the eighteenth of March

The Commune began in a lost war. In July 1870 the empire of Napoleon III declared war on Prussia, and within six weeks the French field armies had been surrounded and destroyed. The emperor himself was captured at Sedan at the start of September, and two days later Paris proclaimed a republic and formed a Government of National Defence, which inherited a war it could not win and a capital it could not feed. Paris was encircled by the middle of September and besieged for more than four months. Inside the walls, the burden of the siege fell with savage unevenness. The wealthy kept their cellars and their credit; the working population, which filled the ranks of the National Guard, queued for bread, ate through the animals of the zoo, and watched the price of every necessity climb beyond a labourer’s wage. The Guard became in those months something more than a militia. It was the armed organisation of the Parisian districts, officered by election, paid a small daily sum that kept many families alive, and increasingly convinced that the government inside the city feared Paris more than it feared the Prussians outside it.

That conviction hardened into certainty during the winter. The government’s sorties against the besieging armies failed with heavy loss. An armistice at the end of January 1871 opened the way to national elections, and the elections of February returned an assembly with a strong monarchist majority, meeting first at Bordeaux and then at Versailles, committed to peace on Prussian terms and deeply hostile to the capital that had endured the siege. Adolphe Thiers, the veteran politician placed at the head of the executive, made the disarmament of Paris a priority of the settlement. Peace required an army able to face France’s own capital, and the Paris National Guard, with hundreds of battalions and hundreds of cannon, was in the eyes of Versailles an armed republic lodged inside a country that had just voted for order. The cannon mattered in a particular way. Many of the guns parked on the heights of Montmartre and Belleville had been paid for by public subscription among Parisians during the siege. They were, in the feeling of the districts that had bought them, the property of the city, and their removal would be both a confiscation and a humiliation.

The confrontation came before dawn on 18 March. Troops sent by the government climbed Montmartre to seize the guns, and the operation failed for want of horses to draw the cannon away and for want of will among soldiers who fraternised with the crowds that gathered against them. Two generals, Lecomte, who had ordered his men to fire on the crowd, and Thomas, a former commander of the National Guard hated for his severity, were seized and shot by soldiers and guardsmen in a garden in the rue des Rosiers. The killings announced that the crisis had passed beyond negotiation. Thiers ordered the entire administration of the state to withdraw to Versailles, expecting Paris to collapse into disorder and be retaken cheaply, and the Central Committee of the National Guard, a body of mostly obscure battalion delegates, found itself in possession of the Hôtel de Ville and of a capital abandoned by its government. The Committee’s first significant act set the pattern for everything the Commune would become. Instead of governing in its own name, it called elections. Paris voted on 26 March, and on 28 March the Commune was proclaimed beneath the flags of the National Guard to an enormous crowd, with the names of the elected read out one by one.

What Paris had done, in its own understanding, was to revive an old claim rather than invent a new one. The commune was the traditional form of municipal self government in France, and the crowd’s demand had been framed through the siege in exactly those terms: that Paris should elect its own council, control its own militia, and manage its own affairs as other French towns did in law if not always in fact. Layered over that municipal claim were others that pulled in different directions. Republicans remembered 1793 and wanted a revolutionary government in the Jacobin mould. Followers of Auguste Blanqui wanted a disciplined insurrectionary authority that would save the republic by energetic dictatorship. Followers of Pierre Joseph Proudhon, dead six years but present in every workshop argument, wanted a federation of communes that would dissolve central power altogether. Members of the International Workingmen’s Association, Marx’s organisation, wanted measures for labour, though they disagreed among themselves about nearly everything else. The Commune was all of these inheritances at once, wearing one name, and much of its history is the story of those inheritances competing inside a single council chamber while an army assembled against it outside the walls.

How brief was the Commune’s life?

Seventy two days, counted from the rising of 18 March 1871 to the fall of the last barricades on 28 May. The elected Commune itself sat for a shorter span, from its proclamation on 28 March to the collapse of organised resistance in the Bloody Week, and every measure it passed belongs to that compressed and besieged interval.

Measured against the movements that have claimed it as an ancestor, the brevity is startling. The Commune governed for roughly ten weeks, less time than a single modern parliamentary session, and it governed under conditions no successor regime would have recognised as normal: a hostile national government twenty kilometres away, a victorious foreign army holding the forts to the north and east and watching through field glasses, a treasury dependent on whatever the Bank of France could be persuaded to release, and a population divided between districts that manned the barricades and districts that waited for deliverance. Later revolutionaries measured themselves against its span with an almost superstitious attention. Lenin, it is said, danced in the snow on the morning his own government had outlasted the Commune’s seventy two days, because survival past that mark had become the tradition’s private unit of account. The anecdote survives because it captures something real about the Commune’s afterlife: an episode so short that its duration itself became a benchmark, and so dense with decree and argument that longer lived governments have produced less for historians to dispute.

Brevity also shapes the evidence. Ten weeks of government generated an official journal, minutes of council sessions, thousands of posters and proclamations, and a flood of hostile reporting from Versailles, but it did not generate the sediment of implementation that a longer regime leaves behind. Many decrees were announced in the final fortnight and never operated at all. Many were drafted for districts the council no longer controlled. A decree of the Commune is therefore evidence of intention and of the balance of argument inside the chamber, and only sometimes evidence of an accomplished change in the life of the city. That distinction, between the Commune of the decree book and the Commune of the streets, is one the ledger later in this article is built to preserve, because the two are habitually merged in commemorative writing and the merger flatters everyone. A government that existed for ten weeks under siege should be judged first by what it attempted in full knowledge of its position, and only then by the far harder question of what its measures might have become in a peace it was never going to be granted.

Who sat on the Council of the Commune

The single most consequential fact about the Commune is also the one most often omitted from popular accounts: it was not a Marxist government. The council elected on 26 March 1871, and enlarged by supplementary elections in April after refusals and resignations, numbered around ninety members over its life, and its political composition reflected the Paris of the workshops and the clubs rather than any single doctrine. The largest organised currents were the Blanquists, disciplined revolutionaries who looked to the imprisoned Blanqui as their chief and who thought in terms of insurrection, dictatorship and the revolutionary seizure of the state, and the Proudhonists, federalists and mutualists whose instinct ran in precisely the opposite direction, toward the dissolution of central authority into self governing communes and associations. Around and between them sat neo Jacobins who modelled themselves on the Committee of Public Safety of 1793, radical republicans whose socialism was rhetorical, journalists, clerks, artisans, and a handful of professional revolutionaries of no fixed school. Delescluze, the ageing Jacobin who became the Commune’s delegate for war in its final weeks and died on a barricade, belonged to a tradition that predated Marx’s influence in France entirely.

Members of the International formed a small minority of the council. Historians counting affiliations usually arrive at a figure somewhere between seventeen and twenty five International members among the roughly ninety elected, and the counting is genuinely difficult, since membership in the French sections was fluid and some councillors carried overlapping loyalties. Their importance inside the Commune was nonetheless real, and it was practical rather than commanding. Léo Frankel, a Hungarian jeweller and correspondent of Marx, served as delegate for labour and presided over the commission that produced the Commune’s most substantial social measures. Eugène Varlin, a bookbinder who had built the International’s French organisation, worked on finance and on the supply of the city. Benoît Malon, Auguste Serraillier and others carried administrative burdens out of proportion to their numbers. Yet the International members were themselves divided, many of them formed by Proudhon’s teaching rather than by Marx’s, and the Paris sections of the International had spent the years before the war in factional argument in which Marx’s London General Council was a distant and sometimes resented authority. Marx held no seat, issued no instructions that the council obeyed, and learned of most decisions from newspapers and letters like any other observer in London. The Commune’s most famous measures, as the ledger will show, can be traced to identifiable tendencies inside the chamber, and the tracing rarely leads to Marx.

The council’s social composition was as striking as its political mixture. It was dominated by men of the skilled trades and the lower middle class: metalworkers, printers, bookbinders, shoemakers, building workers, clerks, teachers, journalists, a minority of professionals and intellectuals, and almost no one from the great bourgeoisie that had fled to Versailles or the western districts. Women, who organised clubs, staffed canteens and ambulances, founded the Union des Femmes pour la défense de Paris under Elisabeth Dmitrieff, and fought in the final week, were excluded from the vote and from the council itself, a limit the Commune shared with every French institution of its century and one its admirers have often preferred to pass over. Dmitrieff herself illustrates the episode’s tangled allegiances: a Russian member of the International in direct correspondence with Marx, she built in Paris an organisation of working women concerned with wages, workshops and the restructuring of labour, an achievement that owed something to the International’s networks and nothing to any directive about how the Commune should govern.

Understanding the composition matters because it governs the meaning of everything Marx later said. When Marx praised the Commune’s measures, he was praising the improvised work of Blanquists, Proudhonists, Jacobins and unaffiliated artisans, most of whom had never read him and some of whom belonged to traditions he had spent years attacking in print. When he drew conclusions about the state from their experiment, he was generalising from a government whose members would have disputed his interpretation of their own actions, in several cases from prison or exile. And when later movements claimed the Commune as the founding deed of their own doctrine, they were annexing an event whose makers had been annexed by nobody. The composition question is therefore not pedantry. It is the difference between a case study and a foundation myth, and the case study is the more instructive of the two. Readers who want the general theory that Marxists built on this episode will find it set out in the marxist theory of the state guide, while the present article keeps its attention on the narrower ground of what happened in Paris and what one observer in London made of it.

What the Paris Commune actually decreed

The Commune legislated at speed, in a chamber that sat while guns were being positioned against it, and its decrees divide into three families that are worth keeping apart. The first family was defensive and economic in the immediate sense: measures to keep a besieged population fed, housed and solvent. The second concerned the form of government itself: how officials were chosen, what they were paid, and how they could be removed. The third reached toward a reordering of property and labour: workshops, church property, the organisation of work. The families matter because posterity has tended to judge the Commune by the third while the Parisians of 1871 mostly lived under the first, and because Marx’s theoretical harvest came overwhelmingly from the second. A government remembered as a socialist experiment spent the greater part of its energy keeping rents unpaid, bakeries staffed, pawned blankets returned and salaries capped, and that distribution of energy is itself a finding about what a workers’ government, improvised under fire, actually does with its days.

Was the Paris Commune a socialist government?

Not in the ownership sense, and not a Marxist one. Its leadership was dominated by Blanquists and Proudhonists, and members of Marx’s International were a small minority. Marx defended it after it began and drew theoretical conclusions from it, but he later described its majority as in no sense socialist.

The rent decree shows the first family at its most characteristic. On 29 March the council remitted the rents that had accumulated through the siege months, cancelling nine months of obligation for tenants who had been shelled, starved and in many cases unpaid since the previous autumn. Landlords protested that the decree was confiscation wearing a municipal costume, and in a narrow sense they were right, since the Commune had no power to compensate them and no intention of trying. Yet the measure was less an attack on property as such than a refusal to let the siege’s costs settle where the market would have dropped them, and it was wildly popular in the districts that formed the Commune’s base. Companion measures extended the same logic: a moratorium on commercial debts, the return from the municipal pawnshops of pledged tools, bedding and clothing below a set value, so that a workman could recover the instruments of his trade, and the reopening of the workshops and canteens that kept the Guard’s families fed. None of this expropriated an industry. All of it announced that the ordinary operations of debt and rent would be suspended where they collided with the survival of the population, and the announcement, repeated across dozens of small decrees, was the socialism most Parisians experienced: not common ownership, but a government that treated subsistence as prior to contract.

The second family went deeper, because it redesigned the machinery through which the city was ruled. Every public official of consequence was to be elected, and every elected official was subject to recall by the electors who had chosen him, at short intervals and on demand. Administrative and judicial posts alike were opened to election where appointment had ruled before. Officials were to be paid at workmen’s wages, a principle the council applied to itself with a ceiling of six thousand francs a year at a time when senior servants of the empire had drawn salaries and perquisites many times that sum. The standing army was abolished in principle, its place taken by the National Guard, the armed people in the Commune’s own formula. Church and state were separated by decree on 2 April, religious budgets suppressed, and church property declared national, a measure carried by anti clerical conviction and by the memory of priests blessing the cannons of previous coups. Taken together, these decrees describe a coherent constitutional idea, assembled from Jacobin memory, Proudhonist suspicion of authority and workshop practice: that delegation is a loan and not a gift, that office should pay like labour and answer like labour, and that the distance between the governed and the governing is a political choice rather than a law of nature. It was this family, far more than any economic measure, that Marx seized upon, and the seizure is examined in its own section below.

The third family was the smallest, the last to arrive and the most argued over, then and since. On 20 April the Commune abolished night work in the bakeries, a reform demanded by the bakers themselves, whose trade condemned them to begin labour when the city slept, and resisted by master bakers whose protests forced partial retreats in enforcement. On 16 April came the decree with the longest afterlife: workshops abandoned by owners who had fled the city were to be inventoried and handed over to cooperative associations of the workers who had staffed them, with provision for compensation to be settled later. A commission under Frankel began the census of abandoned ateliers. A handful of workshops, including work for the Louvre and for military supply, passed into cooperative management before the end. The measure’s reach was therefore modest and its significance mostly prospective, a sketch of an industrial order rather than its construction. The Commune undertook no general expropriation of factories, banks or railways, seized no great fortunes, and left the structure of Parisian property standing almost everywhere it found it. Its boldest economic instinct was protective rather than transformative, and its boldest transformative instinct was constitutional rather than economic, which is nearly the reverse of the reputation that later attached to it.

The Commune ledger

The table below is the article’s findable artifact, and it exists to enforce the distinction this whole account depends on: the distinction between what the Commune enacted and what legend afterwards assigned to it. Each row takes one measure, states what it did in the plainest terms available, names the tendency inside the council that chiefly proposed it, records whether Marx cited the measure as a model in his writings on the Commune, and follows the idea into its later career. Read down the third column and a pattern emerges that no summary of the episode usually admits: Marx’s citations cluster on the constitutional measures, the ones about election, recall and pay, and barely touch the economic decrees that dominate the Commune’s popular image. Read down the second column and a second pattern emerges: the authorship of the Commune was plural in a way its later custodians found inconvenient, with Proudhonists drafting labour measures, Jacobins supplying the forms of revolutionary government, Blanquists pressing the logic of emergency, and International members doing much of the administrative work that held the parts together.

Measure	What it did	Tendency that proposed it	Cited by Marx as a model	What became of the idea
Remission of rents	Cancelled rents accrued through the siege months and barred eviction for arrears	Radical republicans and International members responding to district pressure	No	Became the template for emergency rent moratoria adopted by later governments in wartime
Return of pledged goods	Released tools, clothing and bedding held in municipal pawnshops below a set value	International members and district delegates	No	Remembered as the Commune’s signature act of immediate relief and rarely repeated as policy
Election and recall of officials	Made public posts elective and every mandate revocable by the electors at any time	Proudhonist federalists with Jacobin support	Yes	Entered Marxist and council traditions as the defining institution of workers’ government
Officials paid at workmen’s wages	Capped public salaries at six thousand francs a year and applied the cap to the council itself	Shared across tendencies, driven by workshop practice	Yes	Became a standing demand of revolutionary movements and a test applied to later regimes
Abolition of the standing army	Replaced the professional army within the city with the National Guard of armed citizens	Jacobins and Blanquists drawing on the levée tradition	Yes	Recast by later theorists as the suppression of a separate armed body over society
Separation of church and state	Ended religious budgets and declared church property national	Anti clerical republicans across factions	No	Achieved in French law a generation later by a republican parliament, not by revolution
Abolition of night baking	Ended night labour in the bakeries at the demand of the bakers’ organisation	International members through the labour commission	No	Absorbed into the general history of working time regulation in the baking trade
Abandoned workshops to cooperatives	Ordered inventories of deserted workshops and their transfer to workers’ associations	Proudhonists and International members, drafted under Frankel	Partly, as tendency rather than text	Supplied the precedent invoked by later advocates of workers’ control and self management
Moratorium on debts	Postponed enforcement of commercial obligations contracted before and during the siege	Proudhonist mutualists concerned with small trade	No	Treated by historians as siege economics rather than doctrine
Secular and free education	Committed the city to free, compulsory and secular schooling and began reorganising schools	International members around Vaillant with republican teachers	No	Realised nationally under the Third Republic within a decade by ordinary legislation

Two rows of the ledger deserve a warning attached to them, because they are the rows where legend has done the heaviest work. The workshops decree is habitually described in commemorative writing as the socialisation of production, and the ledger’s wording is chosen to resist that description. The decree addressed abandoned workshops only, it contemplated compensation, it was issued with five weeks of the Commune’s life remaining, and the number of enterprises actually transferred was small enough to be counted without effort. Its importance is genuine but prospective: it showed a government of working men reaching, under Frankel’s commission, for a cooperative settlement of the ownership question, and stopping, whether from principle or from time, at the threshold. The recall row carries the opposite warning. Recall is sometimes treated in later writing as a decorative flourish, a garnish of direct democracy on a conventional revolutionary government. The record shows the opposite. The Commune’s electors used recall and replacement, the council’s composition shifted through by elections and removals, and the principle structured the Commune’s whole claim to be a different kind of state. When Marx came to write his account, it was this row, and the rows on pay and the standing army, that he lifted out of the decree book and into theory. The economic rows he largely left where he found them. That selectivity is a clue to what he thought the episode had proved, and it disciplines any account of the lessons drawn: the discovery Marx announced was constitutional before it was economic, a discovery about the form of a government rather than about the management of production.

The first position: the warning

Marx’s engagement with the Commune began half a year before the Commune, and it began with a prohibition. In the autumn of 1870 the International’s General Council in London, with Marx drafting, issued two addresses on the Franco Prussian war. The first, published in July as the war began, had warned the workers of Europe against the dynastic adventure and predicted that a war on France would ruin the empire that launched it. The second, dated 9 September 1870, five days after the proclamation of the republic in Paris, faced the new situation directly. The French workers, it argued, should use the republic to build their organisations and press their claims in calm and resolute fashion. What they should not do was rise. Any attempt to overthrow the government at that moment, with the Prussian armies advancing on the capital and the country unready, would be desperate folly, and the address used exactly that register of warning. The enemy was at the gates, the provinces were conservative, the peasantry distrusted Paris, and an insurrection in the capital would hand the reaction the pretext and the means to crush the workers’ movement for a generation. The advice was strategic rather than squeamish. Marx had supported insurrections before and would defend this one soon enough. His point was about timing and correlation of forces, and the events of May 1871 would vindicate the analysis with terrible fidelity.

Did Marx support the rising before it began?

He did not. The International’s Second Address of September 1870, drafted by Marx, warned Paris workers that an attempt to overthrow the government while Prussia besieged the city would be desperate folly, and his private correspondence through the winter repeated the warning in blunter terms to correspondents in France.

Nor was the public warning a pose adopted for respectability. Marx’s private letters from the siege months carry the same judgement without the diplomacy. He worried that the Paris sections of the International, in which Proudhonist and Blanquist influences contended, would force a premature explosion; he followed the failed risings of October 1870 and January 1871, when crowds and Guards battalions moved against the Government of National Defence after its humiliations, with the exasperation of a man watching his analysis confirmed in the worst way. Even after 18 March, when the insurrection he had counselled against existed and had won the city, his correspondence held reservations that the public record would later conceal. Writing to Ludwig Kugelmann in April, at the very moment his praise of the Communards was most ardent, Marx listed their errors as he saw them: the Central Committee had surrendered its power too soon by rushing to elections instead of consolidating the victory, and the Commune had failed to march at once on Versailles while Thiers’s government was weak and disorganised. History, he observed in the same letter, would have been far easier had the moment been seized. The letter is indispensable evidence, because it shows the defence and the criticism running in the same mind at the same time, and it forbids the simple story in which Marx either fathered the Commune or merely applauded it. He judged it a rising made at the wrong moment by forces he did not control, and he judged its makers magnificent in the making. Both judgements were sincere, and the sequence they begin is unintelligible unless both are kept.

The warning also deserves a fair statement of its limits. It was addressed to a situation that vanished within months. By March 1871 the republic Marx had told workers to use calmly was being dismantled by an assembly intent on monarchy and on the disarmament of Paris, and the rising of 18 March was not the planned insurrection the address had warned against but a spontaneous defence of the city’s cannon that escalated beyond anyone’s design. Whether the warning still applied to that different event is a question Marx never answered in so many words, and honest accounts should leave the question open rather than conscript the September address into a verdict on March. What can be said is that Marx, when the unforeseen rising succeeded, did not stand on his earlier counsel. He crossed to its side entirely, and the crossing, precisely because it was a crossing, is the pivot on which the three-position sequence turns.

The second position: the defence

Marx followed the Commune from London through newspapers, through the reports of the International’s network, and through correspondence with participants including Frankel, to whom he sent advice on matters as practical as the treatment of the Bank. His public defence took shape in stages. Through April and May he drafted and redrafted an address of the General Council on the civil war in France, working through early versions that the surviving manuscripts show him expanding as news arrived. On 30 May 1871, two days after the last barricades fell, he read the finished text to the General Council, which adopted it. The Civil War in France appeared in print in London in the middle of June, over the names of the Council’s members, and it made Marx, to the fury and fascination of the European press, the public face of the Commune’s defence across the continent. Newspapers that had never printed his name now denounced him as the red doctor directing revolution from exile, a reputation built on a pamphlet defending a government in which his followers had been a minority and his instructions largely absent. The irony is structural and worth pausing on: the defence was so much more powerful than the influence that the defence itself created the myth of the influence, and Marx spent part of the following decade living inside a legend his own prose had manufactured.

The pamphlet’s argument moves through three registers. The first is exposure: a savage account of Thiers and his government, of the assembly of landlords and monarchists at Versailles, of the lies by which the provincial press prepared France to accept the reconquest of its capital. The second is vindication: an answer, charge by charge, to the atrocity stories attached to the Commune, several of which, including tales of organised bands of women incendiaries, the pétroleuses, were fabrications or wild exaggerations of the Versailles press. The third register is the one that outlasted the quarrel, and it is analysis. In the pamphlet’s third part Marx set out what the Commune had been: not a conspiracy or a riot but a form of government, the political form at last discovered, in his phrase, under which the economic emancipation of labour could be worked out. The famous sentences belong to this part. The Commune was formed of councillors elected by universal suffrage and responsible and revocable at short terms. It was a working body, executive and legislative at the same time, not a parliamentary talking shop. It suppressed the standing army and replaced it with the armed people. It stripped the police of political functions and made its agents responsible and revocable. It opened office to election, paid office at workmen’s wages, and in doing so, Marx argued, it had solved the riddle that all previous revolutions had left untouched: how to prevent the machinery of the state from standing over society as its master rather than serving as its instrument.

Reading the pamphlet now requires the double vision this article has promised. As polemic it is magnificent and as reporting it is uneven, and the unevenness has a pattern. Marx wrote at speed, from exile, dependent on Paris newspapers and on correspondents whose accounts he could not verify, about events in a city he could not enter. Where the Commune’s decrees were concerned his reporting is exact, because the decrees were printed and could be checked against the official journal. Where the fighting and its atrocities were concerned he relied on the same flood of rumour that misled everyone, and some of his vindications have not survived archival scrutiny intact, while some of the charges he dismissed as invention, including organised incendiarism in the final week, turn out to have had a basis in the deliberate burning of public buildings by retreating Communard units, though not in the myth of the pétroleuse. None of this diminishes the pamphlet’s analytic core, which did not depend on the daily news. It does mean that The Civil War in France cannot be quoted as a neutral chronicle, and that its author, who knew the difference between evidence and advocacy better than most, would have been the first to insist on the distinction had the pamphlet been written by an opponent.

What Marx drew from the episode

The conclusion Marx carried away from the Commune can be stated in one sentence, and he stated it himself with unusual directness. The working class, he wrote, cannot simply lay hold of the ready made state machinery and wield it for its own purposes. The sentence appeared first in The Civil War in France and then, in the preface that Marx and Engels wrote for the German edition of the Communist Manifesto published in 1872, it was elevated into a correction of the Manifesto itself, the authors declaring that the programme of 1848 had become antiquated in that one particular, proved so by the Commune. Readers tracing the amendment can follow its textual path in the communist manifesto readers guide, where the prefaces are treated as a layered record of the theory’s revisions. The claim embedded in the sentence deserves slow unpacking, because it is the point at which the Commune stops being an episode and becomes a doctrine. Every previous revolution, on Marx’s reading, had perfected the bureaucratic and military machine of the state and transferred it to new hands, which then used it against the society that had made the revolution. The Commune had done the opposite. It had begun to dismantle the machine’s separateness: fusing the making of law with its execution, subjecting every delegate to recall, paying officials as workers, dissolving the standing army into the armed population. The lesson was therefore negative in form and institutional in content. A workers’ government could not be the old state under new management. It had to be built on the Commune’s pattern, as a working body whose delegates remained inside the class that sent them.

What did Marx conclude about the state?

That a workers’ movement could not take over the existing state machine and use it unchanged. Reading the Commune’s elective, revocable and workman paid offices as the alternative, he and Engels wrote the conclusion into the 1872 preface to the Manifesto as the one correction the episode had proved necessary.

Around that central sentence gathered the institutional details that later theory would elaborate into a model. The delegate bound by a mandate and subject to recall answered the question of representation: the representative was to be a delegate in the strict sense, carrying instructions and liable to instant dismissal, not a trustee freed by election from those who elected him. The workman’s wage for officials answered the question of privilege: the material separation between rulers and ruled, which Marx took to be the seed of bureaucracy, was to be cut away at its root in pay and perquisite. The fusion of executive and legislative answered the question of parliamentarism, which the Commune’s practice treated as a screen behind which ministerial and official power operated unobserved. The replacement of the standing army by the armed people answered the question of coercion’s ownership: the separate body of armed men, available for use against the population, was the state’s hard core, and the Commune had aimed at exactly that core. The term later fixed to this whole arrangement, the dictatorship of the proletariat, belongs to another part of the series and is examined in the article on the dictatorship of the proletariat; the present point is the narrower one, that in 1871 Marx attached the term’s institutional content to a specific, short lived, largely non Marxist municipal government, and treated its constitution rather than its economic programme as the discovery.

Which measures, then, did Marx treat as the real discovery? The ledger already supplies the answer, and the texts confirm it line by line. In The Civil War in France the passages written with the greatest theoretical excitement concern election, recall, pay, the working body, the army and the police. The economic decrees receive honourable mention and brief treatment; the workshops decree is praised as a tendency, a pointer showing the direction in which the Commune’s logic ran, rather than analysed as an achievement. When Engels returned to the pamphlet twenty years later, in the introduction written for the edition of 1891, he sharpened the same emphasis and directed it against the state superstition he found among socialists of his own day, holding up the Commune’s safeguards against the transformation of officials from servants of society into its masters as the episode’s permanent bequest. The consistency across the two decades matters. What the tradition’s founders thought the Commune had proved was a proposition about political form. What later popular memory thought it had proved was a proposition about property. The founders’ version is less stirring on a banner and considerably more precise, and recovering it is one of the services the ledger performs.

The Bank of France, or the revolution that paid its enemy

No decision of the Commune has been more second guessed than its treatment of the Bank of France, and none better illustrates the distance between the Commune’s self understanding and the revolutionary logic later imposed on it. The Bank stood in the rue de la Vrillière holding the reserves of French finance: notes, gold, securities and the machinery of credit on which the Versailles government depended for the sinews of its reconquest. A revolutionary government in the Jacobin or Blanquist style might have seized the reserves and used them as a hostage, threatening the entire propertied order of France with ruin unless Paris was left alone. The Commune instead appointed as its delegate to the Bank Charles Beslay, an elderly Proudhonist engineer and the council’s most committed apostle of legality in finance, and Beslay protected the Bank. He negotiated with its deputy governor, drew out sums belonging to the city and to the Commune’s accounts, in total somewhere around sixteen million francs across the whole episode, and left the vaults untouched. Through its provincial branches, meanwhile, the Bank supplied Versailles with funds on an entirely different scale, commonly reckoned in the hundreds of millions, financing the army that destroyed the Commune. The contrast has the shape of a parable, and historians have rarely resisted telling it as one.

Why did the Commune leave the Bank of France untouched?

Because its delegate there, the Proudhonist Charles Beslay, believed the Bank’s reserves belonged to the whole of French commerce and that seizing them would destroy credit, ruin small savers and discredit the Commune as theft, a conviction the council, divided and fearful of provincial opinion, never overruled.

Beslay’s reasoning was not cowardice, and stating it fairly is necessary before judging it. For a Proudhonist, the Bank’s offence was its monopoly and its privilege, not its existence, and the remedy was a people’s bank organised on mutual credit, not a confiscation that would make the Commune the proprietor of the old machine. Beslay also believed, and said, that the Bank was the property of France rather than of Paris, held on behalf of commerce in every province whose opinion the Commune was courting, and that to loot it would confirm every Versailles poster depicting the Communards as thieves. There was a further consideration that the council’s moderates pressed throughout: the Commune presented itself as a lawful municipal authority defending republican legality against a coup, and a government resting its claim on legality hesitates before the most spectacular illegality available to it. The hesitation was coherent. It was also, in the judgement of many later critics inside the tradition, fatal. Engels, writing the introduction of 1891, counted the respectful treatment of the Bank among the Commune’s cardinal errors and laid it at the door of the Proudhonist influence, arguing that the reserves in the Commune’s hands would have bought concessions no army could have won. Marx’s own contemporary letters criticise other errors more sharply than this one, and the difference of emphasis between the two founders is a small but genuine complication for any account that speaks of the Marxist verdict on the Bank as if it were single and settled.

The counterfactual deserves a cooler assessment than it usually receives. Seizing the Bank might have strengthened the Commune’s negotiating hand; it might equally have unified propertied France, terrified the provinces whose neutrality the Commune needed, collapsed the credit on which the city’s own food supply depended, and invited the Prussian command, which held the ring around Paris and cared intensely about French financial stability, to shorten the leash it kept on Versailles. Counterfactuals of this size cannot be settled, and the honest conclusion is double. The Commune’s financial restraint flowed from the convictions of the men who held the relevant posts, and those convictions were Proudhonist rather than Marxist, which is one more reason the episode resists annexation by any single doctrine. At the same time, the restraint had a cost that can be stated without romance: the institution the Commune declined to touch financed its destruction, and the asymmetry between sixteen million drawn and hundreds of millions supplied remains the starkest single illustration of the difference between governing a city and making a revolution against a state.

The third position: the sober reassessment

Ten years after the fall of the Commune, Marx was asked by Ferdinand Domela Nieuwenhuis, a leader of the Dutch socialists, whether the Commune offered a model for the movement’s conduct. The letter Marx wrote in reply, dated February 1881, belongs to a different world from The Civil War in France, and the difference is the point. The Commune, he wrote, was merely the rising of a city under exceptional conditions. Its majority was in no sense socialist, and could not have been. With a small amount of common sense, he added, it might have reached a compromise with Versailles useful to the whole mass of the people, and that compromise was the only thing achievable at the time. The sentences have been quoted ever since by critics of the tradition as a recantation and by defenders as a private note of no public standing. Both treatments mistake the document. Marx was answering a practical question about imitation, and his answer applies his whole method to the case: an event’s meaning is fixed by its conditions, a rising of one besieged city in a defeated and monarchist country could not have carried a socialist transformation whatever its members intended, and copying its forms in different conditions would be theatre rather than strategy. The warmth of 1871 is absent. So is any withdrawal of the theoretical conclusion drawn in 1871. The letter says nothing against the analysis of the state that the Commune had prompted, and its target is not the Commune but the impulse to treat the Commune as a transferable script.

Set the three positions side by side and the sequence resolves into something more disciplined than either conversion or apostasy. The warning of September 1870 concerned timing: a rising in a besieged capital, in a country whose provinces and peasantry stood against it, would be crushed, and the workers should not spend themselves in it. The defence of 1871 concerned obligation and analysis: given that the rising existed and that Versailles was drowning it in blood, the International’s duty was solidarity, and the analyst’s duty was to extract what the experiment had revealed about government, which turned out to be a great deal. The letter of 1881 concerned imitation: the Commune belonged to its conditions, its majority had not been socialist, and no movement should mistake a heroic municipal defence for a general model of revolution. Each position answers a different question, and the appearance of contradiction among them arises largely from flattening the questions into one. Did Marx support the Commune? Before it existed, no. While it existed, completely. As a template afterwards, no again, though he continued to hold the lesson he had drawn from it. The sequence is not a scandal to be explained away. It is the most instructive part of the whole inheritance, because it shows a theorist’s categories being formed by an event in public, tested against the event’s failure, and retained only in the precise form the evidence would carry.

The sequence also explains the Commune’s strange double career in later politics, claimed with equal confidence as a Marxist model and as proof that Marxism misreads history. Both claims are made by selecting one position and discarding the other two. The model claim selects the defence: Marx praised the Commune’s constitution, therefore the Commune’s constitution is the Marxist model of the state, and the warning and the letter are passed over as tact. The misreading claim selects the letter, or the warning: Marx admitted the majority was not socialist and that he had opposed the rising, therefore the entire edifice of Commune theory rests on a misdescription, and the defence is passed over as propaganda. A reader equipped with the full sequence can refuse both shortcuts. The defence was genuine advocacy written under the pressure of a massacre, and its analytic claims about the state machine were held by Marx for the rest of his life and repeated by Engels two decades later, which disposes of the notion that they were a pose. The warning and the letter were also genuine, and they mark the limits Marx himself set around the episode’s meaning, which disposes of the notion that he canonised it. What remains, when the shortcuts are refused, is a case study in how theory and event correct one another, and it is more valuable to students of the tradition than either of the legends. Anyone keeping study notes on the sequence, the ledger and the casualty dispute in one place will find the Marxism study notes a practical companion, since the three positions repay being set side by side on a single page rather than met separately in three different books.

Why the Commune fell

The Commune’s defeat was overdetermined, which is a historian’s way of saying that it had more causes than it needed. Some were structural and probably decisive on their own. Some were choices, made by identifiable men in identifiable rooms, and the argument over how much weight the choices could have carried against the structure is where the historiography lives. Begin with the structure. The Commune was one city. France outside the walls had elected a monarchist assembly, associated Paris with the war’s prolongation and the siege’s miseries, and wanted peace, order and the resumption of trade. The provincial communes that flared in Lyon, Marseille, Saint Étienne, Narbonne, Le Creusot, Toulouse and Limoges in the days after 18 March were each suppressed within days, leaving Paris in an isolation that no proclamation could break. The peasantry, the largest class in France, received the Commune’s appeals, including an address to the rural workers distributed by balloon, through the filter of priests, landlords and Versailles newspapers that described a Paris of arsonists and atheists dividing property. Against the city stood a national government that could draw on the resources of the whole country and on an army being rebuilt with Prussian cooperation, since Bismarck released French prisoners of war in growing numbers precisely so that Thiers could finish the Commune and stabilise the peace settlement. By May the Versailles forces under Marshal MacMahon numbered well over one hundred thousand men, trained, equipped and artillery heavy, against a National Guard whose paper strength dissolved, at the barricades, into far smaller numbers of actual fighters, many of them defending their own streets rather than a coordinated plan.

The choices sit inside that structure, and three of them have carried the weight of later argument. The first was the failure to march on Versailles in the days immediately after 18 March, when the government’s forces were few, demoralised and in retreat, and the road west stood open. The Central Committee of the National Guard, holding Paris by acclamation, chose instead to legitimise the revolution by holding elections, a scruple Marx criticised in the Kugelmann letter as surrendering power’s substance for its forms, and one the Blanquists, who might have been expected to demand the march, failed to override. The April sortie, when it came, was improvised and bloody, costing the Commune the popular commander Flourens, killed after capture, and the illusions that accompanied it. The second choice was the Bank, examined above: the revolution financed its restraint and starved its defence. The third was the Commune’s failure to build, in its ten weeks, a military organisation equal to its danger. Command was divided among delegates for war who succeeded one another, Cluseret, then Rossel, then Delescluze, each inheriting a Guard whose battalions elected their officers, guarded their own districts, and resisted every attempt at the discipline a siege required. Rossel, the one professional soldier of the three, resigned in despair at the council’s interference and the Guard’s looseness, and his letter of resignation reads like a staff officer’s autopsy written before the death. Artillery went unserved, the enceinte’s defence went uncoordinated, and when the end came the city fought as a federation of neighbourhoods rather than as an army, which produced heroism in abundance and a defence in depth almost nowhere.

Historians divide over the counterfactual with unusual passion, and the division should be reported rather than resolved. One tradition, running from Lissagaray through many later writers, holds that the Commune was doomed by its scruples and its delays, that a march on Versailles in March, a seizure of the Bank, and a disciplined command might have changed the outcome, at least to the extent of forcing a negotiated settlement. Another tradition answers that the correlation of forces made the outcome a matter of time and method rather than of choice: a city cannot defeat a nation that has the countryside, the railways, the treasury, foreign recognition and a released army, and the Prussian command, which permitted Versailles to move troops through its lines and handed over the northern forts, ensured that Paris could neither escape nor be relieved. Between the two stands a quieter judgement that has much to recommend it. The Commune’s choices may have shortened or lengthened its life and certainly shaped its character, and its character is what history kept. Its survival would have required the provinces, the peasantry and the European situation to be other than they were. A government can be responsible for its defeats without being the author of its impossibility, and the Commune’s responsibility and its impossibility both belong in the record.

The violence, and the question of the numbers

An honest account of the Commune’s fall has to describe two violences, not one, and has to resist the gravitational pull of both traditions that would keep only half the description. The Commune’s violence was real, and it culminated in the killing of hostages during the Bloody Week. From early April the council had held prominent prisoners as security against Versailles executions of captured Guardsmen, under a decree that promised reprisal in kind, three hostages for every Communard shot. Among the prisoners were Georges Darboy, Archbishop of Paris, seized in the first days of April with other clergy, gendarmes and officials. Attempts to exchange Darboy for the imprisoned Blanqui, the one bargain the Commune wanted most, were refused by Thiers, and the archbishop remained in the prison of La Roquette as the front collapsed. On 24 May, with Versailles troops inside the city, Darboy and five others were shot by a firing squad in the prison yard, an execution ordered in the chaos of defeat and carried out under the authority of the Commune’s apparatus. Two days later, in the rue Haxo in Belleville, a crowd and guardsmen killed around fifty prisoners, gendarmes and priests among them, in an episode that blended orders, panic and revenge. Across the whole episode the number of hostages killed by the Commune is generally reckoned at around one hundred. The figure matters and must not be minimised: these were prisoners, killed in custody, several of them clergy whose offence was their office. It also matters, for proportion, that the Commune’s killings of prisoners across ten weeks were fewer than the Versailles forces would carry out in single locations during the week that followed.

The violence of the suppression operated on a different scale and by a different logic. As MacMahon’s troops fought their way across Paris between 21 and 28 May, they shot prisoners summarily and in organised batches: Guardsmen and suspected Communards were executed at the Lobau barracks, in the Luxembourg gardens, at the École Militaire and at dozens of improvised sites, and on the final day one hundred and forty seven prisoners were shot against the wall of Père Lachaise cemetery that commemoration knows as the Mur des Fédérés. Retreating Communard units set fire to public buildings, the Tuileries Palace, the Hôtel de Ville, the Palais de Justice among them, acts of deliberate destruction in defeat that Versailles propaganda enlarged into a general plot of incendiarism and attached to the mythical pétroleuses. Whole districts were searched house by house. The arrested were marched to Versailles in columns and held in camps and hulks where disease and neglect added to the toll, and the courts martial that followed worked for years: of the tens of thousands arrested, commonly counted at between thirty six and forty three thousand, thousands were imprisoned, and around four thousand five hundred were deported to the penal colony of New Caledonia, from which many returned only under the amnesty of 1880. Children were among the arrested in large numbers. The repression was not a riot of undisciplined troops, though it included such episodes; it was a policy conducted by a government that had announced in advance its intention to make an example of Paris, and its scale was chosen accordingly.

What is the honest answer on the Bloody Week death toll?

Nobody knows, and the honest answer gives the dispute rather than a figure. Estimates repeated for a century ran from twenty to thirty thousand executed. Archival research published in the early twenty first century, built from burial and cemetery records, argues for several thousand, and the dispute between the two remains open.

The casualty question deserves its prominence, because it is the point where the Commune’s history is still being rewritten by evidence rather than merely reinterpreted. The traditional figure entered the record early and from several directions at once. Lissagaray, the Commune’s first historian and a participant, wrote of losses in the region of twenty thousand, and his authority as an eyewitness carried the number into the tradition. Versailles apologists arrived at similar or higher figures by the opposite road, since a Paris imagined as a city of criminals required a repression of matching enormity to appear proportionate, and the monarchist writer Maxime Du Camp retailed numbers of the same order. For a century the agreement of enemies on a vast figure gave it the appearance of settled fact, and it passed into textbooks, speeches and the common stock of reference, usually rendered as twenty thousand or thirty thousand dead in the Bloody Week. The figure did moral work for everyone who repeated it. For the left it measured the crime of the bourgeoisie. For the right it measured the scale of the danger that had been averted. Numbers that serve all sides so conveniently should always be suspected, and in this case the suspicion arrived, late but with documents in hand.

Robert Tombs, a historian of France who had already written a major study of the war against the Commune, returned to the question with a method the earlier literature had never applied systematically: he counted burials. Working through the records of Paris cemeteries and the registers of deaths and interments for the weeks of the suppression, and publishing his findings in the early 2010s, Tombs argued that the number of Communards killed during the Bloody Week itself was far lower than tradition claimed, in the region of six thousand to seven and a half thousand when battle deaths and executions are combined, rather than twenty or thirty thousand. The argument had earlier partial anticipations, notably from Jacques Rougerie, the French historian whose meticulous studies of the Commune’s judicial records had long implied a smaller toll, but Tombs gave it a full evidentiary basis and a number. The revision has not gone unanswered. Critics observe that burial registers may miss bodies disposed of in mass graves outside the municipal cemeteries, in the Seine, or in the fortifications and suburbs beyond the recording system, that deaths in the camps and prisons and on the convoys to Versailles belong in any moral accounting even if they fall outside the week, and that the correction of an inflated figure can slide, in careless hands, into a minimisation of a repression whose character is not in doubt. Tombs himself has been careful to insist that a smaller number of summary executions remains a number in the thousands and a crime of the first order, and that the revision concerns scale, not culpability.

For the reader who needs a citable answer, the defensible form is a sentence with its evidence attached: the traditional estimates of twenty to thirty thousand executed derive from nineteenth century accounts on both sides; archival research based on burial and cemetery records argues for a figure of several thousand killed in the week itself; the dispute is unresolved, and any single figure stated without its source is a marker of advocacy rather than of knowledge. That sentence is longer than the number it replaces, and the length is the price of accuracy on a question where brevity has been the ally of legend for a hundred and fifty years. The episode also offers a general lesson in the handling of atrocity figures, applicable far beyond Paris: the first numbers to circulate are generated by the needs of the moment, they are ratified by repetition across enemy lines, and they can be displaced only by the slow counting of records that were made for other purposes. The method matters as much as the result, and readers interested in how historians rebuild a contested past from administrative residue will find a kindred approach, in a very different archive, traced in the article on indian marxist historians.

The Commune against itself

The council chamber contained the Commune’s contradictions in miniature, and its internal quarrels are part of the evidence about what kind of government it was. The deepest fissure opened in May, over a name. As the military situation darkened, a majority of the council voted to create a Committee of Public Safety, borrowing the title and the implied powers of the body that had governed revolutionary France in 1793 and 1794. The minority, which included most of the International members and the Proudhonist federalists, protested that the new committee recreated exactly the unaccountable central authority the Commune existed to abolish, and issued a public manifesto refusing to attend the council while continuing work in their districts and commissions. The split between majority and minority, as the two blocs came to be called, was never healed, and it maps with uncomfortable precision onto the theoretical division running through the whole episode: Jacobins and Blanquists reaching, under pressure, for concentrated revolutionary dictatorship, and federalists resisting in the name of the mandate and the commune. Anyone seeking the roots of the later quarrel between Marxists and anarchists over the state will find them enacted here in procedural form, a generation before the quarrel acquired its classic texts; the article on marxism vs anarchism on the state follows that dispute into its developed shape, while the Paris record shows the dispute at birth, inside a government rather than between pamphleteers.

The same tension ran through the Commune’s handling of liberty in wartime. A government founded on the revocable mandate and the open club suppressed hostile newspapers as Versailles approached, arrested suspects under the hostage decree, and leaned on the apparatus of the prefecture it had promised to transform, under a prosecutor, Raoul Rigault, whose zeal for revolutionary policing sat oddly beside the council’s constitutional principles. The Commune’s defenders reply that it suppressed less, and killed far fewer, than the government arrayed against it, and the reply is true as far as it goes. The harder question is what the episode reveals about the fate of anti authoritarian designs under the pressure of war, and the Commune’s record here is honestly mixed: it abolished the political police as an institution and rebuilt something recognisably like it under another name within weeks. Marx’s account passes over most of this in silence, a silence that belongs to the pamphlet’s genre as a defence brief. The historians who came after, with the council’s minutes before them, have had no such excuse, and the best of them, from Rougerie onward, treat the majority minority split and the drift to emergency government as central evidence about the episode’s meaning. A Commune that contained both the mandate system and the Committee of Public Safety contained the whole argument of the following century inside its own ten weeks, which is one reason the argument has never been able to leave it alone.

The strongest objection, at full strength

The objection to everything built on the Commune should now be stated as strongly as it can be stated, because the article’s own discipline requires the case against its subject to be heard undiluted. General theoretical conclusions about the modern state were drawn, by Marx and far more by his successors, from a single episode lasting ten weeks, in one city, under siege conditions that no theory would choose as a laboratory. The episode’s government was not composed of Marxists and did not enact a Marxist programme; its majority was, on Marx’s own later admission, in no sense socialist. The experiment was terminated by external military force before its institutions could be tested by the ordinary frictions of peace: by budgets that must balance, by electors who tire, by delegates whose recall becomes a factional weapon, by the slow accretion of expertise and routine that turns servants into masters even under constitutions designed to prevent it. From such an episode, the objection runs, almost nothing general can follow. To conclude from the Commune that the state machine must be smashed and replaced by revocable delegation is to do precisely what Marxists habitually accuse their opponents of doing: mistaking a vivid case for a demonstration, and a selection of agreeable facts for a sample. The canonisation that followed compounds the offence. Later movements carried the Commune as a relic, quoted its seventy two days as a warrant, and built doctrines of the state whose authority derives from an event whose participants would not have recognised the doctrines. The Commune, on this account, is not the foundation of a theory of the state. It is the founding example of the tradition’s tendency to overgeneralise from defeat, beautified by martyrdom and protected from falsification by the impossibility of rerunning the experiment.

The answer begins by separating the conclusions Marx drew into those that depend on the Commune’s success and those that do not. The claim about the inherited state machinery belongs to the second class, and it does not require the Commune to have survived, or to have been socialist, or to have governed well. It requires only that the Commune’s different way of constituting authority was actually attempted: that delegates were elected and recalled in fact, that officials were paid at the capped rate in fact, that the standing army was displaced by the armed population in fact, and that the machinery so constituted functioned, for ten weeks, without collapsing into the usual separation between a governing caste and a governed mass. A single genuine instance is enough to establish a possibility, which is all that conclusion claims: that the ready made machine is not the only form a state can take, and that a movement intending to transform society cannot simply staff the old form and expect new results. The objection is at its weakest exactly here, because possibility proofs are legitimately built on single cases in every field of inquiry, and ten weeks of functioning is a fair amount of evidence for a possibility. Where the objection strikes home is against the larger edifice: the confidence with which later theory treated the Commune’s forms as a general solution rather than a demonstrated possibility, the conversion of a wartime municipality into a timeless model, and the habit of citing Paris 1871 to settle questions about industrial societies of a scale and complexity the Communards never faced. That concession should be made without evasion. The Commune’s canonisation in the tradition far exceeds what the episode can support, and Marx’s own three positions, warning, defence and cooling reassessment, supply from inside the tradition the correctives that its later custodians too often discarded. The case study survives the objection. The foundation myth does not, and the distinction between them has been this article’s argument from the first page.

How later movements used the Commune

The Commune’s afterlife was longer and more consequential than its life, and it began at once. The exiles who reached London, Brussels and Geneva carried the episode into the International’s internal wars, where it became a weapon in the struggle between Marx’s supporters and the anarchist followers of Bakunin, each side claiming the martyred city against the other. The refugees’ relief, the defence of the Communards’ memory and the quarrel over the rising’s lessons ran together through the 1870s. Engels revived the pamphlet in 1891, on the twentieth anniversary, with an introduction that presented the Commune to a new generation of German socialists as the concrete meaning of the dictatorship of the proletariat and as a standing warning against superstition toward the state, and it was through that edition that most later revolutionaries encountered Marx’s text. In Russia the episode acquired a second life. Lenin studied The Civil War in France and Engels’s introduction with close attention, and in The State and Revolution, written in hiding in the summer of 1917, he rebuilt the Commune’s measures into a programme: smash the bureaucratic military machine, replace it with armed workers, elect and recall all officials, pay them workmen’s wages. The article on the state and revolution readers guide examines that rebuilding in detail; here the essential point is the transformation it involved, since Lenin generalised the Commune’s municipal forms into a doctrine for a continental empire and fused them with the soviet, an institution born in Russia in 1905 that owed the Commune an acknowledged debt of inspiration and solved, or claimed to solve, the problem of scale that Paris never faced.

The uses multiplied and diverged across the twentieth century. The Commune became the founding martyrdom of the French left, commemorated each spring at the Mur des Fédérés by parties that agreed on little else. Council movements after the First World War reached back to it as the ancestor of workers’ councils everywhere. Mao’s China invoked it during the Cultural Revolution, when the brief Shanghai People’s Commune of 1967 took its name and its warrant from Paris before being dissolved into other forms. The Paris of May 1968 quoted the Commune in its posters and occupied its symbolic ground. Each appropriation selected the Commune it needed: the martyrdom without the Proudhonists, the mandate system without the Committee of Public Safety, the defence without the warning that preceded it and the letter that cooled it. The three-position sequence is the solvent for all of these selections, since it restores the discomfort each appropriation exists to remove. The Commune also left a quieter legacy in the republic that destroyed it. The Third Republic, built by the men of Versailles, went on to enact through ordinary legislation several measures the Commune had proclaimed: the separation of church and state in 1905, free and secular primary schooling in the 1880s, the legalisation of trade unions, the amnesty that brought the deportees home. A ledger of the episode’s outcomes that counted only revolutionary inheritances would miss this one: the demonstration, unwelcome to every party, that a defeated insurrection can set a legislative agenda its conquerors spend decades completing.

Life inside the second siege

Between the proclamation in March and the catastrophe in May, Paris lived a strange double existence that the political narrative tends to flatten. The city had been besieged by Prussia through the winter; it was now, in effect, besieged by France, with the Prussian armies still holding a ring of forts to the north and east and permitting Versailles to concentrate against the western gates. Inside the walls the spring was remembered by many participants as a season of extraordinary civic energy. The clubs met nightly in churches and halls, debating motions that ranged from the organisation of labour to the renaming of streets. Newspapers multiplied beyond counting. The boulevards filled in the evenings, the National Guard paraded with bands and banners, and the city’s theatres and concert rooms staged benefits for the wounded and for widows. The Commune’s cultural delegates reopened museums, proposed the dismantling of the Vendôme Column as a monument to military glory, which was carried out in May before enormous crowds, and sketched plans for the reorganisation of artistic life under the artists’ federation that the painter Gustave Courbet had helped to found. To read only the council’s decrees is to miss the texture that made participants, decades later, describe the episode as a festival of the oppressed as well as a government, and the texture matters to the history because it explains the ferocity of the grief that followed, in Paris and in exile. Men and women do not mourn an administrative arrangement for fifty years. They mourn a world they briefly inhabited.

The energy had its organised forms, and the most consequential of them were built by women. Barred from the vote, women nonetheless constructed the Commune’s welfare and much of its argument. Dmitrieff’s Union des Femmes enrolled thousands, organised workrooms and production cooperatives for women’s labour, pressed the labour commission on wages and on the restructuring of the garment trades, and demanded, in terms that unsettled the council’s Proudhonists, recognition of women’s work as a public question rather than a domestic one. Louise Michel, a schoolteacher from Montmartre, worked in the vigilance committees, the ambulance services and the firing line in sequence, and became in defeat the repression’s most famous defendant, demanding the firing squad in open court and receiving deportation instead. Canteens, crèches, ambulance stations and the secular schools hurriedly reorganised under the education delegation depended on women’s labour to a degree the official record acknowledges only in passing. Education itself became a field of experiment: teachers dismissed the clergy from municipal schools, secular instruction was declared the rule, and a commission under Édouard Vaillant projected a system of free and compulsory schooling with technical training integrated into it, a programme whose realisation the council did not live to attempt and the Republic later accomplished by other means. The Commune’s social imagination, in short, ran well ahead of its decree book, and both ran ahead of its calendar. That is the charitable reading, and the uncharitable one is also available and should be stated: a government that could find time to topple a column while its artillery went unserved was distributing its energies by the logic of symbolism, and the habits of the festival and the habits of the siege pulled against each other in the same streets, sometimes in the same men.

The Bloody Week

The end came through a door left open. On the afternoon of 21 May a Versailles agent, Ducatel, signalled from the ramparts near the Point du Jour gate that the defences there stood empty, and the troops of Versailles entered Paris without a battle at the wall. The Commune had a week to live. What followed was not a single engagement but a slow, grinding reduction of the city from west to east, street by street and barricade by barricade, as MacMahon’s divisions, fighting with the method of an army that had learned siegecraft in the winter, pushed the defence back toward the working class districts of the east where the Commune had been born. The Tuileries, the Hôtel de Ville and other public buildings burned. The defence, denied any central direction after the council dissolved into district committees, contracted into a series of local stands of extraordinary tenacity, at the Butte aux Cailles, in the rue de la Roquette, around the Père Lachaise heights. On 27 May the Belleville and Ménilmontant districts held the last organised positions; on 28 May the final barricade, in the rue Ramponeau by the usual account, was overrun, and the fighting that had begun with cannon on Montmartre ended in the streets of Belleville seventy two days later. The executions outlasted the battle by days, in the barracks yards, the parks converted to killing grounds and the columns of prisoners on the Versailles road.

What happened during the Bloody Week?

Between 21 and 28 May 1871 Versailles troops reconquered Paris street by street from the western districts to Belleville. Retreating Communard units burned public buildings, hostages including the Archbishop of Paris were shot by the Commune’s side, and the victorious army carried out summary executions on a far larger scale.

The week has been narrated so often in the colours of martyrology and of counter revolutionary triumph that its military character is easily lost. The Versailles army did not storm a fortress; it conducted a methodical urban offensive with overwhelming artillery, secure supply and complete freedom from interference by the Prussian forces holding the northern forts, which allowed French troops passage and denied the Commune any bolt hole beyond the walls. The Commune’s defence suffered from every weakness its history had accumulated: guns without crews, battalions loyal to their own quartiers, delegates for war issuing orders that district committees felt free to disregard, and a council that spent part of the final days disputing the powers of its Committee of Public Safety while the enemy advanced along the boulevards. Individual stands achieved the density of legend. The aggregate achieved nothing that could alter the result. Military historians who have reconstructed the week tend to converge on a judgement the political traditions both dislike: the outcome was decided by the correlation of forces before the first barricade of May was built, and the week’s significance lies not in the fighting’s suspense, of which there was little after the second day, but in the use the victors made of victory. The repression was the message, addressed to Paris and to France, and its scale was a political decision dressed as a military necessity.

What the Commune did not do

The record of omissions is as instructive as the record of acts, and it corrects a different set of legends. The Commune did not proclaim the abolition of private property, did not expropriate the factories of Paris, did not touch the Bank, did not repudiate the national debt, did not separate itself formally from the French Republic whose flag, altered to red in many districts, it continued in law to claim. It did not enfranchise women, did not abolish inheritance, did not collectivise the bakeries whose night work it regulated, and did not attempt the socialisation of the railways or the great department stores then rising in the western districts. Its fiscal policy, such as siege conditions allowed, was orthodox to the point of pedantry: it paid its obligations, kept its accounts, and recoiled from paper money schemes. Its relations with the peasants it sought to win were conducted through manifestos rather than through any transfer of land or cancellation of rural debt that might have given the manifestos weight. Each omission had its reasons, in doctrine, in the balance of the chamber, or in the calendar, and the Proudhonist colouring of the Commune’s economic instinct explains several at a stroke. The list is nonetheless the necessary counterweight to the decree book. A government is what it declines to do as much as what it attempts, and the Commune declined, by conviction or by circumstance, nearly the whole programme that later centuries would attribute to the phrase socialist government. Marx, who read the decree book closely, knew the omissions perfectly well, which is why his praise fastened on the political form and why his private judgements retained their edge. The legend requires a Commune that attempted everything and was crushed. The record shows a Commune that attempted a constitutional revolution and a social minimum, under conditions that would have defeated a far more radical programme even sooner.

The Civil War in France read against the record

Because the pamphlet has carried the Commune into world politics, it deserves an audit of its own, conducted with the same ledger discipline applied to the decrees. Marx drafted the address through April and May 1871, and the surviving early drafts, recovered and published long afterwards, show the argument being built in layers as news arrived from Paris: an opening anatomy of the Versailles government and its origins in the empire’s collapse, a history of French state power reaching back through the century’s coups, and only then the analysis of the Commune itself, expanded in the final version into the text the General Council adopted on 30 May. The drafting history matters because it dissolves a persistent misunderstanding. The pamphlet was not a message sent to Paris, nor instructions for a movement. It was an address from a London council to the workers of Europe and America, written over the sound of a defeat its author could hear only through newspapers, and its purpose was double: to fix the meaning of the event before its enemies fixed it, and to extract from the event the one theoretical correction Marx believed the movement required. Judged as the first kind of act, advocacy, it succeeded beyond any comparable pamphlet of the century. Judged as the second, as analysis, its core claims have proved durable in a way its day to day reporting has not.

What kind of book is The Civil War in France?

It is the address Marx drafted for the International’s General Council, adopted on 30 May 1871 and published in London weeks later. It indicts the Versailles government, answers the atrocity charges against the Commune, and argues that the Commune’s elective and revocable constitution was a new form of working class government.

Three of the pamphlet’s features repay attention by any reader approaching it fresh. The first is its treatment of the French state as an institution with a biography: Marx traces the centralised executive, with its army, police, bureaucracy and clergy, through the revolutions that had each seized and strengthened it, arriving at the empire as the machine’s most complete form, and the whole historical construction exists to make the Commune intelligible as the machine’s negation. The second is the pamphlet’s silence where a defence brief would be expected to shout. It says little about the Commune’s internal divisions, little about its economic hesitations, and nothing about Marx’s own prior warning against the rising, and these silences are not oversights but the structure of advocacy, visible as such once the letters are read beside the text. The third is the pamphlet’s relationship to its facts. Its account of the decrees is precise. Its account of the repression’s scale draws on the same early reports that produced the traditional casualty figures, and it repeats some of the inflation that archival work has since corrected, while its dismissal of Communard incendiarism has required the qualifications described earlier. None of this is an accusation of bad faith. It is a description of a text written inside an information war by a partisan of genius, and it dictates the reading protocol the pamphlet demands: trust its constitutional analysis as theory, verify its chronicle against the archives, and remember at every page which of the two one is holding.

The afterlife of the text is itself evidence. Translated rapidly, pirated, denounced in parliaments, it made its author the most notorious revolutionary in Europe and attached his name permanently to an event he had not led. The International, already strained by the war and by its internal factions, did not survive the association intact; within a year its congress had split against the anarchists, and within five the organisation had dissolved, carrying into exile the quarrel over what the Commune had meant. Marx’s correspondence in those years shows a man alternately proud of the pamphlet’s effect and exhausted by the legend accumulating around it, defending the Communards in public prints, organising relief for the refugees crowded into London, and resisting, in private, the attempts of admirers and enemies alike to make him the Commune’s author. The resistance is worth underlining, because it is of a piece with everything else in the sequence. The man who wrote the most influential account of the Commune declined the authorship that account thrust upon him. Later custodians of the tradition were less scrupulous, and the difference between the founder’s scruple and the custodians’ confidence is a fair measure of what the three-position sequence exists to correct.

How the evidence was made

The Commune is unusually well documented for an event of its brevity, and unusually treacherous in its documentation, because almost every source was produced by a combatant. The official journal printed the decrees and debates under the government’s own supervision. The council’s minutes survived the fires in part and were reconstructed by scholarship. Versailles generated a parliamentary inquiry whose volumes of testimony were assembled to convict the insurrection in the court of respectable opinion, and army records, court martial dossiers and deportation lists that historians have since mined for purposes the compilers would have deplored. Lissagaray wrote his history as a participant determined that the defeated should have their chronicle, gathering testimony from survivors while memories were fresh and passions total, and his book remains the narrative foundation on which nearly everything later rests, indispensable and partisan in equal measure. The foreign press, the illustrated papers and the photographers added a visual record that the repression then curated for its own ends, most notoriously in the manipulated imagery of corpses and ruins circulated as proof of Communard barbarism. Even the barricades acquired a staged double life, photographed in quiet streets for the illustrated market.

Twentieth and twenty first century scholarship built its advances by turning from the narratives to the residues: the trial records counted defendant by defendant, the cemetery registers counted burial by burial, the financial archives of the Bank read against Beslay’s memoirs, the electoral returns of March and April analysed street by street to establish who the Commune’s voters and councillors actually were. Rougerie’s work on the judicial repression set the pattern, demonstrating that a dossier compiled to punish could be read to measure. Tombs’s casualty revision belongs to the same lineage and derives its authority from it. The lesson generalises, and it is the one this article has applied throughout: on the Commune, the administrative record is usually more trustworthy than the eloquent one, because the clerks who made it were counting bodies, francs and votes for purposes that had nothing to do with posterity, while the eloquent were already fighting the war of memory that has never ended. Any reader who wishes to test a claim about the episode should ask first which kind of record supports it. Claims about composition, finance and death tolls that rest on registers deserve confidence in proportion to the registers’ indifference. Claims that rest on the speeches and the histories deserve the respect due to testimony, and the discount due to advocacy, in proportions that the three-position sequence itself helps to calibrate.

A government of commissions

The Commune’s working machinery deserves a closer look than its drama usually permits, because the machinery is where its constitutional ideas met administrative reality. The council governed through commissions, groups of its own members charged with finance, labour, education, provisioning, external relations and the rest, each commission answerable to the council and each headed by a delegate who functioned as a minister while remaining, in the Commune’s theory, a revocable agent of the whole. The arrangement fused powers by design: the council legislated in the morning and executed through its commissions in the afternoon, and Marx fastened on exactly this fusion as the difference between a working body and a parliamentary one. In practice the commissions were overwhelmed. Provisioning a city at war, paying and supplying a militia of tens of thousands, running hospitals, schools, workshops and a treasury on the Bank’s reluctant driblets, all with personnel drawn largely from men whose administrative experience was the workshop and the union branch, produced improvisation of a high order and failures to match. The intendance lurched from shortage to shortage. Pay for the Guard arrived late or not at all. Orders issued by the war delegation were countermanded by district committees whose authority derived from the same elective principle the centre invoked. Every defect of the system can be read two ways, and has been: as the growing pains of a new form of government learning in weeks what older states learned in generations, or as the predictable harvest of dissolving hierarchy in the middle of a siege. The evidence supports elements of both, and supports neither as a complete account.

The council’s handling of its own authority shows the same doubleness. Elections were real, contested and frequent; mandates were challenged; delegates were replaced; the by elections of April filled the seats of members who had resigned, died or refused the result, and brought the chamber’s composition closer to the districts’ temper. Attendance thinned as the crisis deepened, and decisions of the gravest kind, the hostage decree among them, were taken by a fraction of the whole body in sessions whose quorums later readers have cause to regret. The Committee of Public Safety, created to concentrate direction in May, concentrated instead the council’s quarrel with itself, and the minority’s boycott of the chamber over its creation stands as the episode’s neatest constitutional paradox: the federalists defended the Commune’s founding principle by absenting themselves from its government in the hour of its mortal danger, while the majority defended the Commune’s survival by violating the principle on which its government rested. Both halves of the paradox were argued in good faith from premises the Commune itself had supplied. That is worth dwelling on, because it locates the episode’s deepest lesson where Marx, perhaps, did not put it: not in any single institution the Commune devised, but in the demonstration that a government of revocable delegates remains a government, subject to the pressures that bend all governments toward secrecy, concentration and coercion when survival is at stake, and distinguishable from its rivals chiefly by the honesty with which the bending can be seen and named.

Marx’s letters while Paris burned

The public Marx of the address had a private counterpart whose letters, written in the same weeks, complicate the defence without contradicting it, and no serious account of the three positions can leave them in the footnotes. On 12 April, with the Commune not yet a month old and the first sorties already failed, Marx wrote to Kugelmann in terms that have become the most quoted private sentences of the episode, praising the Parisians for a historical initiative and an elasticity entirely unlike the dogmatic servility he attributed to German movements, and reaching for the image of a people storming heaven. The same letter, as noted earlier, lists the rising’s strategic errors with a staff officer’s coldness. Admirers quote the ardour and critics the errors, and the letter’s real significance is that it refuses to separate them. Six days later he wrote again, expanding the theoretical point that the pamphlet would carry: the fight in Paris mattered, whatever its outcome, because it was directed not at this or that abuse but at the state machine itself, and its makers understood, as no previous revolution had, that the transfer of the machine from hand to hand was not emancipation. A reader who follows the letters across April and May watches the theory of The Civil War in France being assembled in real time, under the pressure of dispatches, by a man who believed the experiment would probably be destroyed and believed simultaneously that it had already changed what could be thought about the state. That simultaneity is the emotional and intellectual core of the second position. The defence was never a prediction of victory. It was a judgement that a doomed experiment had disclosed something permanent, and the judgement’s sobriety is what separates it from the commemorative rhetoric that later wrapped the episode in untouchable glory.

Marx’s practical conduct matched the letters. He advised correspondents inside Paris, Frankel among them, on questions within his competence, and his counsel, where it can be traced, ran toward the unglamorous: attention to finance, to the provinces, to the presentation of the Commune’s case abroad. He organised the International’s response to the repression, collecting funds and testimony, sheltering refugees, and combating in the press the inventions of the Versailles publicity machine. He also watched the legend of his own direction of the Commune grow in the European newspapers with an irritation his letters do not hide, aware that the myth served his enemies’ prosecution of the International and would serve, in time, the hagiography of movements he could not control. It is one of the episode’s neater ironies that the supposed director of the Paris revolution spent the spring of 1871 in the British Museum’s reading room and in committee rooms in London, working from newspapers like any other analyst, and that his power over the event was exercised entirely after the fact, through the interpretation that outlived it. Interpretation, in the Commune’s case, proved to be the most durable form of power available. The generals who took Paris are footnotes in its story. The exile who explained it set the terms in which the story is still argued, which is precisely why the explanation must be held to the evidentiary standard this article has tried to enforce.

The provinces, the peasants and the foreign ring

Paris fell inward as much as it was stormed, because the France beyond the walls withheld the support without which a capital’s insurrection is a gesture. The provincial communes of March and April deserve more than the sentence they usually receive, since their fate explains the capital’s. Lyon rose and fell within days, its commune divided between radicals of the International and republicans who wanted a municipality rather than a revolution. Marseille’s commune, led by Gaston Crémieux, held the prefecture briefly and was retaken, its leader later executed. Saint Étienne, Narbonne under Émile Digeon, Le Creusot, Toulouse and Limoges each produced movements that the army and the local authorities extinguished in days, sometimes in hours. The pattern was uniform: a spark from Paris, a seizure of the town hall, a proclamation, and then isolation, as the surrounding countryside, the garrisons and the prefectures closed in. No provincial movement survived long enough to divert a single Versailles division from the march on Paris. The reasons reach deep into the structure of French society. The provinces had voted in February for peace and order after a catastrophic war. The word commune itself frightened property holders of every scale. The peasantry, enfranchised and land hungry in an earlier generation, had received its land from the great Revolution and saw in Paris not its liberator but a threat of new taxes, new requisitions and the old arrogance of the capital. The Commune’s address to the rural labourers, dropped from balloons when the wind allowed, promised that the land should belong to those who worked it, and was read, where it was read at all, through the mediation of curés and mayors who described its authors as brigands dividing other men’s property in Paris already.

Over the whole contest stood the foreign power that had made it possible. Bismarck’s interest in the civil war was precise: he required a French government strong enough to sign and execute the peace, collect the indemnity and hold France steady, and weak enough to pose no threat while doing so. The preliminaries of peace signed in February, and the definitive Treaty of Frankfurt in May, set the terms: Alsace and part of Lorraine annexed, an indemnity of five billion francs, occupation until payment. Within those terms the Prussian command held the ring. Its armies retained the northern and eastern forts around Paris, controlled the corridors through which men and supplies moved, released French prisoners in numbers calibrated to Thiers’s needs against the Commune, and permitted Versailles columns to pass through its lines to attack the city from the north east in the final week. German headquarters watched the Bloody Week through glasses from the heights they occupied. The Commune’s fate was therefore decided within a triangle: a government in Versailles determined on exemplary punishment, a countryside determined on quiet, and a conqueror determined that the first should finish the second’s war for it. Marx’s pamphlet grasped the triangle’s shape, in its passages on the collusion of the French possessing classes with the national enemy, and later historiography has filled in the documentary detail. Recognising the triangle disciplines the argument about the Commune’s errors. A rising can make every mistake in the catalogue and still deserve a structural explanation of its fall, and this one, whose mistakes have been catalogued most lovingly by its own tradition, had the misfortune to face a configuration of enemies in which correct play would very likely only have purchased a longer siege and a larger graveyard.

The long aftermath: trials, deportation and amnesty

Defeat had a bureaucracy, and its operations extended the Commune’s history across the decade that followed. The courts martial sat in Versailles and across France processing the arrested in batches, working from police dossiers, denunciations and the evidence of the parliamentary inquiry, and their judgements distributed the defeated among execution, imprisonment, penal servitude and deportation. The firing squads of the courts, as distinct from the summary killings of May, executed a smaller number, in the low hundreds by most counts, but the machinery’s reach was vast: tens of thousands of dossiers, sentences in absentia for the escaped, and the transportation of roughly four thousand five hundred men, women and children to New Caledonia in the Pacific, a voyage of months in the holds of naval transports to a penal colony at the other end of the earth. The deportees included figures the Commune’s memory would canonise. Louise Michel taught school in the colony, took an interest in the indigenous Kanak population that distinguished her from nearly all her fellow exiles, and supported their revolt of 1878 in defiance of the deportee community’s own instincts. Others sickened, broke or built lives in exile that the amnesty, when it came, could not restore. In London, Geneva, Brussels and New York the escaped Communards formed colonies of waiters, craftsmen, journalists and professional revolutionaries, sustaining newspapers, anniversaries and feuds, and feeding into the labour movements of their host countries cadres formed in the hardest school the century offered.

The politics of return consumed French public life through the 1870s. Partial pardons came first, grudging and conditional. The elections shifted. Victor Hugo, who had offered his Brussels house to the refugees and been expelled for it, lent his immense prestige to the amnesty campaign. The question split the young Third Republic between those who held that the republic could not be founded over unburied enemies and those who held that the Commune’s survivors were a standing conspiracy against order. A partial amnesty passed in 1879 and a full amnesty in July 1880, and the surviving deportees came home to crowds at the Paris stations that were themselves political demonstrations. The men who returned included veterans who would staff the socialist movements of the following decades, carrying the Commune’s memory as personal history rather than as doctrine. Among them the memory divided as the movement divided: the heirs of the Blanquists, the heirs of the International’s collectivists, the possibilists and the Marxists of Jules Guesde’s new party each claimed the succession, and each claim required its own Commune. The decade of the aftermath thus performed, in slow motion, the operation this article has tracked at every stage: the conversion of a ten week municipal revolution, made by men of several doctrines and none, into the property of doctrines that had in several cases not yet been founded. Understanding the conversion is not a debunking exercise. The movements that claimed the Commune drew real strength from it, and the France that amnestied its survivors was changed by the argument. But the historian’s duty to the original event remains what it was: to keep the record of March to May 1871 distinguishable from the uses made of it, and to notice that the men and women who made the Commune had no opportunity to vote on the meanings later assigned to their names.

Commemoration fixed the episode’s geography. The wall in Père Lachaise where the one hundred and forty seven were shot became the Mur des Fédérés, the shrine of the French left, drawing annual processions whose composition charted the movements’ fortunes for a century and a half. The basilica of Sacré Cœur rose on Montmartre, on the hill where the cannon had stood and the generals had died, voted by the national assembly as an act of expiation for the crimes of the Commune, a counter monument in white stone visible across the city the Commune had governed. Paris still carries the argument in its street plan: the martyrs’ wall in the east, the basilica of expiation in the north, the rebuilt Hôtel de Ville and the restored colonnades masking the burn lines of May. A visitor reading the city with the ledger in hand can walk the whole dispute in an afternoon, which is perhaps the most Parisian of the Commune’s afterlives and the most instructive. Monuments, like casualty figures, are arguments that have learned to stand still, and they should be read with the same suspicion and the same respect for what they cost.

What the case study can carry

The findings of this article can now be gathered, and they are deliberately narrower than the traditions built on the episode. The Commune was not a Marxist government: its council was dominated by Blanquists and Proudhonists, its International members were a minority divided among themselves, and its measures, assembled in the ledger, divide into siege relief, constitutional redesign and a modest cooperative gesture toward property, in descending order of weight. Marx’s relation to it moved through three documented positions: a warning against the rising issued in September 1870 on grounds of timing and forces; a complete defence written during and immediately after its fall, from which he extracted the conclusion that the working class cannot simply take over the ready made state machinery, a conclusion he and Engels wrote into the Manifesto’s 1872 preface and never withdrew; and a sober reassessment in 1881 that denied the Commune’s majority had been socialist and denied that its forms could be copied out of their conditions. The violence of the episode belongs to both sides and must be counted for both: around one hundred hostages killed by the Commune, including an archbishop shot in a prison yard, and a repression by Versailles whose death toll, traditionally given as twenty to thirty thousand, has been revised by burial record scholarship to several thousand, with the dispute unresolved and the character of the repression unaltered by the revision. The Commune fell because one city faced a nation, a countryside and a foreign conqueror aligned against it, and because its own choices, the delayed march, the untouched Bank, the divided command, spent what little margin the structure allowed. Its canonisation by later movements exceeds its evidence, as the strongest objection insists, and the founders’ own writings, read whole rather than selected, contain the evidence for that judgement alongside the defence.

What remains, when the legends on both sides are returned to their owners, is the reason the episode will not release its hold. In seventy two days a great city attempted to govern itself on principles that contradicted the centralised state at every joint: delegation instead of representation, recall instead of tenure, the workman’s wage instead of the official’s fortune, the armed citizen instead of the standing army. The attempt was made by artisans and clerks, Jacobins and federalists, theorists of mutually hostile schools, under bombardment and in the shadow of annihilation, and it functioned, imperfectly and contentiously, for as long as it was permitted to exist. History rarely supplies so clean an experiment in political form, and it supplied this one with the control conditions cruelly visible: the same city, the same population, governed before by an empire and after by a republic of the men who destroyed the experiment. The Commune’s value to the student of politics is not that it succeeded, nor that it failed, but that it isolated a variable. For ten weeks, the form of the state was different, and the difference can be studied in decrees, minutes, registers and graves. Marx saw the variable and named it. His three positions are the record of a mind adjusting to what the experiment had and had not shown, and they remain the best guide to reading it: warn honestly, defend loyally, reassess coldly, and never let the needs of a later argument rewrite what the registers say.

That discipline, applied consistently, also settles the question with which most readers arrive: whether the Commune should be celebrated or condemned. The question is badly formed, and the record shows why. A government that remitted the rents of a starved city, returned the workman’s tools from the pawnshop and capped its own members’ salaries also held an archbishop hostage and shot him in a prison yard. A repression that restored legal order to France also filled the Lobau barracks and the Luxembourg gardens with prisoners shot after capture, and shipped thousands to a Pacific penal colony for the crime of having been governed by the losing side. Judgement that cannot hold both columns of that account at once is not judgement but allegiance. The Commune’s historians have supplied, across a century and a half, the materials for a fuller reckoning than any earlier generation could make, and the materials point in one direction: the episode deserves neither the altar nor the cautionary tale, but the harder honour of being understood in full, as the brief experiment in which a great city tried to abolish the distance between rulers and ruled, failed against the arithmetic of nations, and left behind the questions that every subsequent attempt to close that distance has had to answer.

The other government: Versailles

Every civil war has two capitals, and the Commune cannot be understood without its counter image across the lines. Versailles in the spring of 1871 held the elected assembly of France, and that fact, which the Commune’s rhetoric treated as a technicality, was the foundation of its enemy’s power. The assembly had been chosen in February by universal male suffrage in elections held to settle the war, and whatever its monarchist complexion and rural bias, it embodied a legitimacy the Commune, elected by one city in defiance of that assembly, could not match in the eyes of most Frenchmen. Thiers understood the arithmetic perfectly. His policy through March and April was a performance of patience: withdraw, concentrate, negotiate the peace with Prussia, rebuild the army from the released prisoners, and let Paris stew in its own isolation until the reconquest could be presented not as a coup against the capital but as the republic restoring order to a metropolis in the grip of a faction. The assembly voted the funds, the provincial press amplified the atrocity stories, and the government allowed the Commune’s overtures toward negotiation, carried by leagues of republican conciliators and by the freemasons who processed to the walls in April, to fail without conceding an inch. Thiers had announced the principle early and never moved from it: the government would treat with individuals who laid down their arms, and with the Commune as a body, never.

The character of the Versailles regime complicates the morality play in both directions. It was not a monarchy and not, in its own self understanding, a counter revolution of the old style. It was a republic of notables, many of them converts to republicanism as the regime of order, and its leading figures believed sincerely that the Commune represented the dissolution of France into armed municipalities, the return of 1793 with dynamite. Its repression was correspondingly ideological: the army’s officers, many of them monarchists smarting from the humiliations of the war, were given to understand that Paris was to be made an example, and the summary executions of May were the harvest of that understanding as much as of indiscipline. Yet the same regime, within the decade, presided over the consolidation of the Third Republic, the amnesty debates and the slow enactment of reforms the Commune had demanded, and its founder Thiers ended as the republic’s first president in all but the title’s timing. Historians have long noted the paradox that the Commune was crushed by the government that would eventually govern on something closer to the Commune’s municipal programme than to the empire’s. The paradox dissolves on inspection. What Versailles defended in 1871 was not a programme but a monopoly: the right of the national state to decide which reforms France would have and at what pace. Against that monopoly the Commune’s mere existence was the offence, and its measures were secondary, which is why the punishment fell with equal weight on moderates and on incendiaries, and why the repression’s scale was set by politics rather than by the fighting.

The labour commission and the working day

If the constitutional measures were the Commune’s theory, the labour commission was its practice at the level of the working week, and its record rewards attention because it shows what social reform looked like when attempted by the men the reforms concerned. The commission for labour and exchange, presided over by Frankel with Malon among its driving members, worked through delegations from the trade unions and the International’s sections, receiving deputations from bakers, bronze workers, mechanics, laundresses, seamstresses and a procession of trades whose grievances had accumulated through the empire’s years of repression and the siege’s months of idleness. Its concrete achievements fit on a short page, and the shortness is itself informative. Night baking was abolished at the bakers’ demand. The private employment agencies, the bureaux de placement whose fees and favouritism the Paris trades had hated for a generation, were suppressed and their function transferred to municipal offices in the mairies. Fines and deductions imposed by employers on wages, the petty confiscations by which a workshop’s discipline was financed from the worker’s pocket, were prohibited in the enterprises within reach. Contracts for military supply were let with attention to the workshops’ organisation, and the great workshops decree of 16 April sent the commission’s inspectors through the abandoned ateliers of the faubourgs to catalogue machinery, stock and the men who knew how to run them. Beyond these acts stood a project rather than a record: the commission’s plan for a census of trades and a reorganisation of production through the chambers of the unions themselves, drafted in the Commune’s final fortnight and stillborn with it.

Two judgements about this record are possible, and both should be heard. The dismissive judgement notes how small the harvest is beside the rhetoric of social revolution, and how entirely it consists of the regulation of abuses rather than the transformation of ownership. The generous judgement replies that the commission’s significance was procedural before it was substantive: for the first time in the history of a great European state, the men and women who did the work sat, through their own delegates and organisations, in the office where the conditions of work were decided, and the trades responded with a flood of programmes, statistics and proposals that showed a governing competence the possessing classes had always denied them. Marx, reading the reports in London, valued the commission’s work in the second sense, and the workshops decree’s place in his citations, respectful but brief, reflects exactly that valuation: a tendency honoured for its direction rather than a transformation celebrated as achieved. The sequel gave the argument a strange coda. The Third Republic legalised the trade unions in 1884, built its labour law piecemeal across the following decades, and arrived by ordinary politics at much of the commission’s regulatory agenda, while the cooperative commonwealth the commission dreamed of remained where the Commune left it, in the project file. Which half of that sequel vindicates the commission depends on the theory of change a reader brings to it, and the article leaves the reader in possession of both.

The Commune inside the International’s quarrel

The episode’s consequences for Marx’s own organisation were swift and, in the end, fatal to it. The International emerged from the spring of 1871 publicly branded as the Commune’s author, a brand its enemies in every European chancellery found invaluable and its members in several countries found ruinous, as governments moved against sections whose actual offence had been to exist while Paris burned. Inside the organisation the catastrophe accelerated the struggle between the General Council’s supporters and the federalist opposition gathering around Bakunin, for whom the Commune proved the opposite lesson to the one Marx drew: that spontaneous federal revolt, not centralised political organisation, was the revolution’s true form, and that the Council’s attempt to discipline the movement in the Commune’s name repeated the authoritarian error. The Hague Congress of 1872 expelled Bakunin and transferred the Council to New York, a victory for Marx that amounted to the organisation’s honourable burial. Within a few years the International had ceased to exist in either form, and the Commune stood as its monument and its death warrant in one. The quarrel the episode bequeathed, between the reading of the Commune as a state to be generalised and the reading of it as a federation to be emulated, structured the socialist movement’s arguments for the next half century and supplies the historical root of the later dispute examined elsewhere in this series.

Engels’s introduction of 1891, written for a new edition of The Civil War in France on the rising’s twentieth anniversary, intervened in that quarrel with deliberate force. European socialism in 1891 was parliamentary, growing and increasingly comfortable, and Engels aimed the Commune at his own side: at the German party’s fear of its own shadow, at the superstition that the state was a neutral instrument or a necessary evil to be administered kindly, at the philistine dread of the very phrase dictatorship of the proletariat, which he invited his readers to contemplate in the Commune’s concrete form. The introduction is a polemic of high quality and should be read as one, with its selections and silences noted. It presents the Commune’s safeguards against officialdom as a complete answer to the problem of the state, passes lightly over the council’s divisions and the majority’s Jacobinism, and assigns the Commune’s economic errors to the two schools, Blanquist and Proudhonist, whose rivalry it blames for the whole failure, a distribution of blame that conveniently acquits the tradition Engels and Marx had founded. Read beside the 1881 letter, the introduction marks the distance between the founders’ private calibration of the episode and their public deployment of it, a distance that is not hypocrisy but the ordinary space between analysis and mobilisation. Later readers who took the introduction as the whole truth about the Commune, as generations of party education encouraged them to do, inherited the deployment without the calibration. Recovering the calibration is a large part of what the three-position sequence is for.

The hostage decree and the road to the prison yard

The killing of the hostages was the Commune’s darkest act, and its prehistory deserves the same exactness as its condemnation, because the sequence shows a policy of deterrence decaying, step by step, into the crime it was meant to prevent. The decree of 5 April was adopted in direct response to news from the front: Versailles troops were shooting captured National Guards out of hand, among them the wounded and prisoners taken in the failed sortie of 3 April, and the council, helpless to protect its men by force, reached for the ancient instrument of reprisal. Persons accused of complicity with Versailles would be held as hostages of the people of Paris, and the execution of a Guard or a Commune supporter by the enemy would be answered by the execution of three hostages chosen by lot. The archbishop was arrested the same week, with priests, gendarmes and officials gathered in a sweep directed by Rigault’s police apparatus, and the Commune offered, repeatedly and through intermediaries, the exchange it wanted above all others: Darboy and the clergy for Blanqui, whose election to the council in absentia had made him its symbolic president and whose presence it believed would transform its counsels. Thiers refused the exchange at every approach, calculating, as his correspondence makes plain, that the archbishop was worth more to Versailles as a prisoner the Commune dared not harm than as a free man, and that the Commune’s possession of him discredited it daily. The calculation held for seven weeks. The decree’s machinery, designed as a threat, accumulated prisoners it could neither exchange nor release, and when the city fell street by street in May, the threat’s logic executed itself in the prison yards and in the rue Haxo, carried out by men acting on the decree’s warrant, on local orders and on the momentum of revenge, in proportions historians still disentangle.

Placing the hostage killings beside the repression’s scale is an exercise the sources force on any honest reader, and it must be done without allowing either column to excuse the other. The Commune killed around one hundred prisoners across its life, most of them in its final agony, in acts its own best traditions condemn and its defenders have never succeeded in explaining away. Versailles killed, in one week, a number that even the lowest archival revision places in the several thousands, many of them prisoners shot after capture in organised batches, and followed the killing with a judicial machine that ground on for years. The asymmetry does not acquit the prison yard. It locates the prison yard inside a war whose other side practised reprisal not as a desperate deterrent but as a settled policy of exemplary terror, announced in advance by its commanders and ratified by its assembly’s applause. Moral accounting that suppresses either entry is propaganda. The Commune’s hostages belong in its ledger alongside its decrees, entered in the same ink, and the most respectful treatment their victims can receive from a historian of the episode is the treatment this article has attempted: their names remembered, their killers’ chain of decision traced, and their deaths neither inflated into the whole story nor dissolved into its context.

Errors in the standard accounts

Certain mistakes recur in coverage of the Commune with such regularity that cataloguing them is a service in itself, and each can be corrected from the record assembled in this article. The first error makes the Commune a Marxist government directed from London. The composition of the council, the minority position of the International’s members and the absence of any channel by which Marx could have directed events dispose of it, and the error survives chiefly because Marx’s defence was mistaken, by enemies and admirers alike, for evidence of command. The second error makes Marx the rising’s instigator. The September 1870 address and the winter correspondence show the opposite, and the error requires ignoring documents that have been in print for well over a century. The third error states the Bloody Week’s death toll as a settled figure, whether the traditional twenty or thirty thousand or, in newer writing that has absorbed the revision uncritically, a precise low number offered with false confidence. The evidence supports a dispute, not a figure, and the dispute’s terms, traditional estimates from nineteenth century accounts against several thousand from the burial registers, are the citable content. The fourth error stretches the Commune’s duration, whether by counting from the empire’s fall or by dissolving the episode into the siege, when the government under discussion lasted seventy two days and the elected council ten weeks at most. The fifth error describes the Commune’s programme as socialist in the ownership sense, a description the ledger refutes row by row: the property of Paris stood essentially untouched when the Commune fell, and its radicalism was constitutional, aimed at the form of government rather than at the title deeds of the city. A sixth error, subtler than the rest, quotes The Civil War in France as a report. The pamphlet is a defence and a theory built from a defence, magnificent in both capacities, and its chronicle requires the archives’ correction at the points where speed and partisanship governed its information. Readers who hold the six corrections together possess something rarer than a new interpretation of the Commune. They possess an account proof against the commemorations of both camps, which is the only kind of account the episode, at this distance, still needs.

The vote of 26 March and the two Parises

The elections that created the Commune repay the same register based scrutiny as the burial records, because they establish what kind of mandate the council carried and whose city it was. Paris voted by arrondissement on 26 March, eight days after the rising, in an atmosphere that mixed festival, intimidation and exhaustion in proportions that varied by district. Turnout across the city ran well below the levels of the February national elections, and the geography of participation was sharply uneven. The eastern and northern working class districts, Belleville, Ménilmontant, Montmartre, the faubourgs along the canal, voted heavily and returned the revolution’s candidates by large margins. The wealthy western districts abstained in mass, their electors fled, indifferent or waiting for Versailles, and several of the men elected there, moderate republicans and notables returned by small and unrepresentative polls, refused their seats or resigned within days, unwilling to lend their names to an insurrectionary municipality. The council that resulted was thus doubly unrepresentative in opposite directions: it spoke for an insurgent Paris that had voted with enthusiasm, and it claimed to govern a whole city a large part of which had declined to vote at all. Both facts belong in the record, and they complicate every simple statement about the Commune’s legitimacy. Against Versailles the Commune could plead that Paris had chosen its council in a free vote while the assembly rested on elections held under the shadow of invasion and defeat. Against the Commune, Versailles could plead that a minority of Parisians, however fervent, had erected a government over districts that rejected it, and that the nation as a whole had voted the other way six weeks earlier. The argument was never resolved because it could not be: two legitimacies, municipal and national, each with ballots behind them, faced one another across the fortifications, and the question was settled by the artillery of the side that had more of it.

The supplementary elections of 16 April, held to fill the seats emptied by resignation, death and refusal, shifted the chamber toward its final character and register a second finding. Turnout fell again, in some districts to levels that embarrassed the council’s own journals, as war weariness, the first defeats and the tightening blockade of supplies did their work. The men returned were on the whole more radical than those they replaced, the chamber’s centre of gravity moving toward the Blanquist and Jacobin majority that would vote the Committee of Public Safety in May. A government elected in enthusiasm was thus renewed in attrition, by a shrinking and hardening electorate, and the pattern is worth naming because it recurs in revolutionary situations with mechanical regularity: as the circle of participation narrows, the politics of those who remain intensify, and the distance between the government and the governed population it claims to embody grows in exact proportion to the sincerity of both. The Commune’s defenders have sometimes treated the April elections as a detail. They are better treated as the episode’s internal measure of its isolation, recorded in the driest of sources, the returning officers’ counts, and all the more trustworthy for it.

Bread, wages and the economy of a besieged city

Underneath the constitutional argument ran a material emergency that consumed more of the Commune’s hours than any question of doctrine, and a history of the episode that neglects it mistakes a government for a debating society. Paris entered the Commune already half starved by the Prussian siege, during which the population had eaten the animals of the zoos, the horses of the cab companies and, in the siege’s folklore that contemporaries repeated with documentary support in individual cases, the rats of the sewers and the cats of the roofs. The armistice had reopened supply lines briefly, and the civil war closed or taxed them again. The Commune’s provisioning commissions requisitioned stocks, fixed the supply of bread and meat to the districts, fought the speculators whose trade the blockade enriched, and attempted to keep the markets of the Halles functioning while the National Guard’s pay, thirty sous a day for the rank and file, served as the welfare system of entire quarters, the one reliable income in a city where the workshops stood idle and the wealthy employers had gone. The economics of the Commune cannot be separated from that wage. Hundreds of thousands of Parisians lived, directly or through a guardsman in the family, on the Commune’s military payroll, which is one reason the Guard’s battalions resisted every scheme of discipline that threatened the payment, and one reason the council’s financial dependence on the Bank’s reluctant releases was a strategic fact and not a bookkeeping embarrassment. A government paying an army it does not fully command, from a treasury it does not fully control, to defend a city it cannot feed from its own resources, is conducting three sieges at once, and the Commune conducted them for ten weeks.

The working population’s experience of the episode was therefore double in a way the political record flattens. For the districts that formed the Commune’s base, the spring of 1871 brought a real if fragile improvement in the terms of life: rents suspended, debts frozen, pawned tools redeemed, night work abolished in the bakeries, the employment agencies suppressed, children in secular schools, and the daily sense, recorded in letters and memoirs across every later collection, of a city whose public spaces and public offices had ceased to belong to somebody else. For the same districts it brought requisition, inflation in the markets, the departure of work and employers, sons and fathers at the ramparts, and at the end a repression that fell with deliberate concentration on those very streets. Both experiences are the Commune, and the historiography divides partly by which of them a writer’s sources preserve. The administrative record, payrolls, provisioning returns, pawnshop ledgers, school registers, keeps the first. The judicial record keeps the second. A complete history has to read the two archives against one another, and the effort of doing so is the largest single reason the episode continues to generate scholarship rather than merely commemoration: the sources do not agree about what the Commune was like to live under, because it was not one thing to live under, and the disagreement of the archives is the most faithful portrait of the event that remains.

One further material fact deserves recording, because it corrects the picture of a city living on air and enthusiasm. The Commune kept the ordinary machinery of urban life running to a degree its enemies found inconvenient to acknowledge. Water, markets, hospitals, the post within the walls, the registry of births and deaths, the pawnshops and the bakeries all continued under its administration, staffed largely by the same clerks and tradesmen who had served the empire, now taking their instructions from elected delegates. Marriages were celebrated at the mairies in large numbers, some couples choosing the Commune’s civil ceremony with a deliberateness that suggests the episode’s legitimacy, in its own districts, ran deeper than the abstention figures imply. The mundane continuity matters to the argument of this article. The Commune’s claim to have discovered a form of government rests not on the barricades but on this unglamorous evidence that a great city’s services could be administered for ten weeks by revocable delegates without the sky falling, and that the population of the eastern districts experienced the arrangement not as chaos but as the most intelligible government they had known. Versailles propaganda required a Paris of anarchy. The provisioning returns and the registry books describe a Paris of queues, ledgers and improvised order, and between the two descriptions the second has the better evidence.

The shadow of the first revolution

No government in French history has governed without ancestors looking over its shoulder, and the Commune’s chamber was crowded with the dead. The revolution of 1789 and the terror of 1793 and 1794 supplied the Commune’s vocabulary, its institutional temptations and its deepest divisions, and the uses the councillors made of that inheritance explain several decisions that baffle readers who meet the episode without its prehistory. The Jacobins on the council, Delescluze above all, understood themselves as the heirs of the Mountain, and for them the Commune’s emergency pointed toward the remedies of the Year Two: a Committee of Public Safety, energetic central direction, the revolutionary tribunal’s spirit if not its instrument. The Blanquists arrived at similar conclusions from a different genealogy, their master’s whole doctrine having been formed in meditation on why the revolutions of 1830 and 1848 had been stolen for want of a resolute minority. Against both stood the Proudhonists and most of the International members, for whom 1793 was not a model but a warning: the revolution that had centralised itself into the Terror and then into the Empire was the standing proof that revolutionary government devours its own purposes, and the federal commune, built upward from free municipalities, was designed as 1793’s antidote. The Commune thus contained, as a live quarrel among its members, the argument between the two great inheritances of the French revolutionary tradition, and its votes can be read as that argument’s minutes. The hostage decree, the Committee of Public Safety and the suppression of opposition newspapers belong to one inheritance. The mandate system, the recall and the federalist proclamation of the communes of France belong to the other. That the second set defines the Commune’s contribution to political thought while the first set defined much of its final conduct is the episode’s central irony, and Marx, whose own formation lay in argument with both inheritances, read the irony correctly: his pamphlet celebrates the antidote and passes over the relapse in the silence of an advocate who knows exactly which of the two his client’s reputation can afford.

The memory worked on the other side of the lines as well, and with heavier consequences. To the assembly at Versailles and to the France that read its newspapers, the Commune was 1793 returned with petroleum, and the fear of the Terror authorised in advance a repression of a scale that the military situation alone cannot explain. The generals and ministers who directed the reconquest spoke the language of exorcism. Paris was to be purged of the revolutionary tradition itself, and the executions of May, the deportations and the basilica of expiation voted for the hill of Montmartre were all, in their different registers, attempts to settle accounts not only with the men of 1871 but with the ghosts of 1793. The Commune’s own members, many of whom courted the comparison in their rhetoric, little understood how completely the comparison served their enemies. A government that names its emergency committee after the Committee of Public Safety should not be surprised when its opponents govern their repression by the memory of the guillotine, and the disproportion between the Commune’s actual terror, a hundred hostages, and the terror it was imagined to intend did real work in producing the disproportion of the punishment. Historians who treat the episode’s symbolism as decoration on its substance miss the mechanism: in 1871 the symbols were weapons, loaded by both sides, and the side with the artillery profited most from the exchange.

Church, schoolroom and the separation decree in practice

The decree of 2 April separating church and state was among the Commune’s earliest acts of principle, and its practical career inside the ten weeks illustrates the distance between the decree book and the street better than any other measure. On paper the separation was complete at a stroke: the budget of religions suppressed, the property of the religious congregations declared national, the clergy reduced to private citizens sustained by the voluntary contributions of the faithful. In practice the Commune moved through a city where the Church ran schools, hospitals and charities woven into the fabric of every poor district, and where religious practice among the population it governed ranged from fervent devotion to fervent hatred of the clergy as the empire’s accomplices. Churches were closed in some districts and left open in others; some were converted into clubs and meeting halls, their pulpits occupied by orators of the revolution in an inversion that delighted the Commune’s journals and appalled the provinces; arrests of priests gathered pace as the hostage policy took hold; and the education delegation, working with the municipalities, expelled the teaching congregations from the municipal schools and installed lay teachers, among them many women, in what amounted to a rehearsal for the school laws the Republic would pass a decade later. The measure’s authorship was broadly republican rather than narrowly socialist, the separation having been a demand of the democratic programme since 1848, and its fate after the Commune is the standing example of the episode’s indirect legislative legacy: annulled in the repression, argued for a generation, and enacted at last in 1905 by a parliament of the republic the Commune had defied, in a settlement more durable than the decree precisely because it arrived by the slow road the Commune had no time to travel. For the ledger’s purposes the row teaches a general lesson about reading the Commune’s acts: a decree’s significance may lie less in its operation during the ten weeks, which was partial and uneven, than in its demonstration that the policy was available, popular in the capital and administratively possible, a demonstration later legislators quietly used while publicly condemning the demonstrators.

For students and exam candidates

The examinable core of this episode is the three-position sequence, and it should be learned as an answer structure rather than as a story. Position one: in the Second Address of September 1870 Marx warned that a rising in besieged Paris would be desperate folly, a judgement about timing and forces that events vindicated. Position two: once the Commune existed, Marx defended it in The Civil War in France, adopted on 30 May 1871, and drew from its constitution, elected and revocable officials paid at workmen’s wages, a fused working body in place of parliamentarism, the armed people in place of a standing army, the conclusion he wrote into the 1872 preface to the Manifesto: the working class cannot simply take over the ready made state machinery. Position three: in the letter to Domela Nieuwenhuis of 1881 he called the Commune the rising of a city under exceptional conditions whose majority was in no sense socialist, and refused it as a model for imitation. Around that spine, secure the supporting facts that markers use to distinguish command of the topic: the Commune was governed mainly by Blanquists and Proudhonists, with International members a minority; it lasted seventy two days; its most cited measures were constitutional rather than expropriating; and the Bloody Week’s death toll is a historiographical dispute between traditional estimates of twenty to thirty thousand and a burial record revision arguing for several thousand. An answer that deploys the sequence, names the composition, and states the casualty question as a dispute will outperform any quantity of narrative, because it demonstrates the skill the topic is set to test: the separation of an event from the legends deposited on it.

For teachers and seminar leaders

The Commune teaches best when the class is made to work with the incompatible documents rather than with a summary. A productive sequence opens with the September 1870 address and asks students to reconstruct the strategic situation that made the warning reasonable, then sets The Civil War in France beside the April letters to Kugelmann, where praise and criticism occupy the same page, and closes with the 1881 letter and the question of what, exactly, changed between the three texts. The honest handling of Marx’s private doubts is a pedagogical asset rather than an embarrassment: students who meet the doubts inside the primary material, framed by the questions each document answers, learn to read a thinker’s development instead of memorising a posture, and they are inoculated against the debunking use of the same letters, which depends on presenting as a scandal what the full sequence shows to be a method. The ledger lends itself to seminar division of labour, each group auditing one measure against the criteria of enactment, authorship and afterlife, and the casualty dispute supplies a self contained exercise in source criticism: give the class the traditional figure and the revision with their respective bases, and require a citable sentence as the outcome. The seminar’s governing rule throughout is the one the material itself teaches, that advocacy and evidence can inhabit the same text and must be separated by the reader, since no editor will separate them on the reader’s behalf.

For researchers, historians, and journalists

The live research questions on the Commune are archival, and two of them dominate. The first is composition in the strict sense: the council’s membership can be reconstructed from electoral returns, section records of the International and police dossiers, yet the figures in circulation for International membership still vary between roughly seventeen and twenty five because affiliation itself was layered, with men formed by Proudhon sitting as International members and Blanquists overlapping with both. Prosopographical work that fixes the chamber’s sociology would settle arguments that still run on inherited estimates. The second is the casualty revision. Tombs’s burial record method has moved the Bloody Week’s toll from the traditional twenty to thirty thousand toward several thousand, and the debate now turns on the registers’ coverage: bodies in suburban and military burial grounds, deaths in the camps, the convoys and the hospitals after the week itself, and the boundary between killed in the suppression and killed by it. Researchers should also note how much of the financial record remains underused; the Bank of France’s archives permit the flows to Versailles and to the Commune to be stated with a precision that moralised debate rarely achieves. Journalists face the simpler discipline of citation. The figure to print is a range with its sources attached: traditional estimates of twenty to thirty thousand from nineteenth century accounts, an archival revision arguing for several thousand, the dispute open. The composition to print is Blanquist and Proudhonist predominance with an International minority. The Marx to print is three positions, not one. Each of these forms is longer than the shorthand it replaces, and on this episode the shorthand has a hundred and fifty year record of being wrong.

Frequently Asked Questions

Q: What was the Paris Commune?

The Paris Commune was the revolutionary municipal government that ruled Paris from March to May 1871, after the government of Adolphe Thiers withdrew to Versailles and the city elected its own council. Born from the siege of Paris in the Franco Prussian war and from the attempt to seize the National Guard’s cannon on 18 March, it combined a claim to municipal self government with the programmes of the Blanquists, Proudhonists, Jacobins and International members who sat in its chamber. It legislated on rents, labour, education and the constitution of office itself, fought a civil war against the national government, and was destroyed in the Bloody Week of 21 to 28 May, after which its survivors faced execution, imprisonment and deportation.

Q: How long did the Paris Commune last?

Seventy two days, from the rising of 18 March 1871 to the fall of the last barricade on 28 May. The elected Commune itself was proclaimed on 28 March, so the council’s legislative life spans roughly ten weeks, and the Commune’s entire existence fits inside a single spring. The brevity is central to every judgement about it. Its institutions had no time to settle, many of its decrees never operated, and its experiment in government was terminated by military force rather than by its own internal development. Later movements nonetheless treated the seventy two days as a complete demonstration, and the tension between the episode’s brevity and the weight placed on it is one of the standing problems in its historiography.

Q: Was the Paris Commune a socialist government?

Not in the ownership sense, and not a Marxist one. The council was dominated by Blanquists and Proudhonists, members of Marx’s International formed a small minority, and the Commune undertook no general expropriation of property: the factories, banks and great fortunes of Paris stood untouched when it fell. Its measures were chiefly constitutional, remaking how officials were elected, paid and recalled, together with siege relief for tenants and debtors and a modest transfer of abandoned workshops to workers’ cooperatives. Marx defended it completely once it existed and drew theoretical conclusions from its form of government, yet a decade later he described its majority as in no sense socialist, a judgement the composition of the council fully supports.

Q: What measures did the Commune actually pass?

The Commune’s decrees fall into three families. For relief, it remitted rents accrued through the siege, postponed commercial debts, returned pledged tools and clothing from the pawnshops, and fed the city through canteens and requisition. For the constitution of government, it made officials elective and revocable, capped their pay at workmen’s wages, abolished the standing army in favour of the National Guard, separated church and state, and opened the way to free secular education. For labour and property, it abolished night work in the bakeries, suppressed the private employment agencies, banned employers’ fines on wages, and ordered abandoned workshops to be inventoried and handed to workers’ cooperative associations. The ledger in this article sets each measure beside its authors and its afterlife.

Q: How many people died in the Bloody Week?

The figure is genuinely disputed, and the dispute is the answer. Nineteenth century accounts on both sides, including the Commune’s first historian Lissagaray and its Versailles enemies, produced estimates of twenty to thirty thousand killed during the week of 21 to 28 May 1871, and those numbers passed into a century of textbooks. Archival research published in the early twenty first century, above all Robert Tombs’s work on Paris burial and cemetery records, argues that the registers support a far lower toll, in the region of several thousand killed in the week itself. Critics reply that the registers may miss bodies buried beyond the municipal system and deaths in camps and convoys. Neither figure should be quoted without its source and its basis.

Q: What did Marx conclude from the Commune?

Marx concluded that the working class cannot simply take over the existing state machinery and use it for new purposes, because the centralised bureaucracy, army and police of the modern state had been built, revolution after revolution, into an apparatus standing over society. The Commune showed him the alternative form: officials elected and revocable at short terms, paid at workmen’s wages, a council that worked and legislated at once rather than debating while ministers governed, and the standing army replaced by the armed population. He stated the conclusion in The Civil War in France and then, with Engels, wrote it into the 1872 preface to the Communist Manifesto as the one correction the Commune had proved to their earlier programme. He never withdrew it.

Q: Did Marx support the rising at the time?

His support changed across the episode, and the change is documented. Before the rising, no: the International’s Second Address of September 1870, which Marx drafted, warned that an insurrection in besieged Paris would be desperate folly, and his private letters through the winter repeated the warning. Once the Commune existed, his support was complete in public and qualified in private: The Civil War in France defended it before the world, while his letters to Kugelmann in April 1871 praised the Parisians’ initiative and simultaneously listed their errors, the premature surrender of power by the Central Committee and the failure to march on Versailles. A decade later he judged its majority to have been in no sense socialist. Warning, defence and reassessment are the three positions, and all three are genuinely his.

Q: Why did the Paris Commune fail?

It failed chiefly because it was one city against a nation. The provinces had voted for peace and a monarchist assembly, the provincial communes that rose in sympathy were suppressed within days, the peasantry distrusted Paris, and Bismarck released French prisoners so that Versailles could rebuild an army exceeding one hundred thousand men, while the Prussian ring around the capital prevented escape or relief. Inside the walls, choices consumed what margin remained: the Central Committee held elections instead of marching on Versailles while it was weak, the Commune left the Bank of France and its reserves untouched while the Bank financed Versailles, and military command stayed divided among successive delegates and district battalions that resisted coordination. Historians still argue over whether better choices could have changed the result or merely its date.

Q: What is The Civil War in France about?

The Civil War in France is the address Karl Marx drafted for the General Council of the International, adopted on 30 May 1871, two days after the Commune fell, and published in London in June. Its first parts indict Thiers and the Versailles government and trace the growth of the French state machine through the century’s revolutions. Its central part answers the atrocity charges against the Commune and then analyses the Commune itself as a new form of government: elective, revocable, paid at workmen’s wages, executive and legislative together. Its reporting was written at speed from exile and requires checking against the archives, but its constitutional analysis became the foundation text for every later Marxist account of the Commune, greatly extended by Engels’s introduction of 1891.

Q: Why does the Commune matter so much to Marxists?

Because it is the one episode in which Marx revised his theory of the state in response to a workers’ government that actually existed, however briefly and however little it resembled his own following. From the Commune came the conclusion written into the Manifesto’s 1872 preface, that the state machine must be broken up rather than captured, and the institutional model of revocable delegates on workmen’s wages that later theory, above all Lenin’s State and Revolution, rebuilt into a programme. The Commune also supplies the tradition with its founding martyrdom and its ritual calendar of commemoration. Its hold is therefore double, theoretical and emotional, and the two holds are often confused: the episode’s evidentiary value is narrower than its symbolic value, and keeping the two apart is the discipline this article has tried to practise.

Q: Which Commune measures did Marx treat as the real discovery?

The constitutional measures, not the economic ones. In The Civil War in France the theoretical weight falls on election with recall for all officials, the payment of officials at workmen’s wages, the council as a working body uniting legislation and execution, the replacement of the standing army by the armed people, and the conversion of the police into revocable agents. These were the measures from which Marx drew his conclusion about the state machine. The economic decrees receive brief and respectful treatment: the workshops decree is praised as a tendency pointing toward cooperative production rather than analysed as an achieved transformation, and the relief measures, rents, pawnshops and moratoria, pass almost without theoretical comment. Engels’s introduction of 1891 kept the same emphasis, which is why the ledger in this article records Marx’s citations measure by measure.

Q: How should a teacher present Marx’s private doubts about the rising?

As primary evidence inside the sequence, not as a revelation against it. Students should meet the September 1870 warning, the April 1871 letters holding praise and criticism together, the public defence of May, and the cool letter of 1881 as answers to four different questions: whether to rise, how to judge a rising that exists, what the rising revealed, and whether to copy it. Presented that way, the doubts teach Marx’s method, the subordination of judgement to conditions, and they lose their power as a debunking device, which depends on quoting one document against the others. The teacher’s framing matters: a private letter is evidence of what Marx thought, not a verdict that cancels a public text, and the class should be asked to specify what each document would have to show for the familiar charge of inconsistency to hold.

Q: What does the Commune ledger separate between legend and record?

It separates what the Commune enacted from what later commemoration assigned to it, measure by measure. On the record side stand the dated decrees: rents remitted, officials made revocable and capped in pay, church and state separated, night baking ended, abandoned workshops ordered to cooperative associations, and the modest scale and late timing of that last decree, which transferred only a handful of enterprises. On the legend side stand the claims the record does not carry: that production was socialised, that Marx directed the government, that the death toll of the Bloody Week is a settled figure, and that the council spoke with one doctrinal voice. The ledger also separates authorship, assigning each measure to the tendency that proposed it, and citation, recording which measures Marx actually treated as models. The result is a Commune smaller than its legend and more interesting than it.

Q: Why did the Commune’s treatment of the Bank of France matter?

Because the Bank financed the Commune’s destruction while the Commune protected the Bank. Its delegate at the Bank, the Proudhonist Charles Beslay, drew out only the sums belonging to the city and the Commune’s accounts, somewhere around sixteen million francs across the episode, and left the reserves of French finance untouched on the principle that they belonged to the commerce of the whole country and that seizure would be theft and ruin. Through its provincial branches the Bank meanwhile supplied Versailles with funds reckoned in the hundreds of millions, paying for the army that retook Paris. Engels in 1891 counted the restraint among the Commune’s cardinal errors, while Beslay’s reasoning, legality, credit and the opinion of the provinces, was coherent within Proudhonist doctrine. The episode matters because it shows the Commune’s economic convictions, not merely its circumstances, shaping its fate.

Q: How did later movements use the Commune as a template?

Selectively, and the selections reveal the movements. Engels’s edition of 1891 presented the Commune to European socialism as the concrete form of working class rule and a warning against reverence for the state. Lenin in 1917 rebuilt its measures, revocable officials, workmen’s wages, the armed people, into the programme of The State and Revolution, fusing them with the Russian soviets, which had themselves grown up under the Commune’s inspiration. The French left made the Mur des Fédérés in Père Lachaise a shrine of annual commemoration shared by rival parties. Council movements after the First World War claimed it as the ancestor of workers’ councils, Chinese revolutionaries named the Shanghai Commune of 1967 after it, and the Paris of 1968 quoted it on its walls. Each use kept part of the record and discarded the rest, usually the warning of 1870 and the letter of 1881.