Few names in political writing attract quotation the way the name of Karl Marx does, and Marx quotes consequently lead a double life. They serve as evidence in arguments about what a nineteenth century thinker actually held, and they serve as ammunition manufactured long after his death for arguments he never joined. A line credited to Marx arrives carrying borrowed authority. It settles disputes in seminar rooms, decorates placards and book jackets, opens essays whose authors never opened the book the line supposedly came from, and survives every correction because a correction is always less portable than the line it corrects. The sentence fits on a poster. The footnote explaining why the sentence is false does not.

The damage done by a false attribution is not symmetrical, and it is not trivial. When a fabricated line is placed in the mouth of a dead thinker, the argument stops being about the line and becomes about the man, and the man cannot answer. Supporters quote inventions that make their tradition sound prophetic or compassionate. Opponents quote inventions that make the same tradition sound sinister or absurd. Both sides are spared the labour of reading, and both sides acquire a weapon that the documentary record never issued. A reader who has watched this happen a few times develops a suspicion that hardens into a different and lazier error, the assumption that every famous line must be fake. That assumption is also wrong, and correcting it matters just as much, because some of the most circulated sentences are genuine, some are genuine in substance while altered in wording, and some are genuine sentences that Marx himself took from writers who came before him. Lumping all of these together under the single label of fake quotes produces fresh errors while pretending to repair old ones.

What follows is an audit rather than a debunking list. An audit begins from a classification, applies a stated method to named cases one by one, records a verdict for each case with the evidence that supports the verdict, and leaves the reader equipped to repeat the exercise on lines the audit never examined. The classification used here has four categories, and the method has a handful of steps that any reader can perform with access to a decent library or to the collected works in a scholarly edition. The cases are the lines that circulate most widely under the names of Marx and, in two instances, Lenin, since Lenin attributions travel through the same channels and fail in the same ways. Some verdicts will please readers who admire Marx and some will discomfort them. Some will please his critics and some will take a favourite weapon out of their hands. That distribution is the point. A quotation is not rescued because it helps a side, and it is not condemned because it harms one. The test is documentary or it is nothing.

Marx quotes and the trade in borrowed authority

Quotation is the cheapest form of authority available to a writer. A statistic can be challenged on its method, an argument on its logic, a historical claim on its evidence, but a sentence attributed to a canonical figure arrives wrapped in the prestige of the figure and is usually received before it is examined. The mechanism deserves a precise description, because the whole misattribution economy runs on it. When a speaker attributes a sentence to Marx, the audience is invited to perform two acts of trust at once: trust that the sentence expresses something worth believing, because Marx is treated as a mind worth believing, and trust that the speaker has done the work of finding the sentence where Marx wrote it. Almost nobody in the audience performs the second check, because the first act of trust has already done the persuasive work. Attribution is therefore a credit system, and like every credit system it invites forgery by anyone who finds the genuine currency expensive to obtain.

Reading Marx is genuinely expensive. The corpus is large, much of it was published posthumously, the major economic work exists in multiple drafts, and the prose is dense with the vocabulary of German philosophy and British political economy. A person who needs a Marx sentence for a speech on Friday has strong incentives to accept the first plausible sentence offered by a search engine, a quotation anthology, or a colleague’s slide deck. Each act of acceptance multiplies the sentence. Search engines rank pages partly by how often their contents are repeated elsewhere, so a fabricated line that has been copied ten thousand times acquires the appearance of massive corroboration, when in fact there is only one original act of invention and ten thousand acts of copying. Repetition masquerades as verification. This is the single most important mechanical fact about the fake quotation trade, and it explains why the most frequently encountered lines are not necessarily the most authentic ones but simply the most repeatable ones: short, balanced, transferable to almost any argument, and phrased in a register that sounds like translated wisdom.

The trade also benefits from a peculiarity of Marx’s position in public memory. He is simultaneously one of the most read and one of the least read of major thinkers, a combination that creates an enormous market for summaries of his views and very little market discipline on their accuracy. The Communist Manifesto is short and is genuinely read. Capital is long and is genuinely cited, which is a different thing. Between the pamphlet everybody finishes and the treatise everybody invokes lies a vast territory of journalism, letters, speeches, drafts and notebooks that almost nobody outside scholarship has traversed. A fabricated line can hide in that territory indefinitely, because challenging it seems to require a search through fifty volumes, and few challengers or defenders have the shelf space or the patience. The audit below is designed to lower that cost. Most of the famous cases can be settled, or honestly reported as unsettled, with a small number of targeted checks rather than an exhaustive search, and the checks are described in full so that the reader can watch them being applied before being asked to apply them alone.

Symmetry is the discipline that separates an audit from a pamphlet. Fabricated quotations circulate on all sides of politics, and any honest account has to include examples used by supporters as well as by opponents, and has to say so plainly. The line about democracy being a road, examined below, flatters a democratic reading of the tradition and is quoted approvingly by people who hold that reading. The line about rope, also examined below, is quoted almost exclusively by opponents who want the tradition to sound cynically conspiratorial. Both lines fail the same documentary test in the same way, and an audit that kept the second while sparing the first would be propaganda wearing the costume of scholarship. The same symmetry governs the treatment of lines that turn out to be genuine. A genuine line is recorded as genuine even when its content is uncomfortable for admirers, and a borrowed line is recorded as borrowed even when the borrowing is entirely honourable, as borrowings usually are. Writers take phrases from their predecessors constantly, and Marx, who read voraciously and argued through other men’s formulations, was no exception.

A further distinction needs stating before the categories are introduced, because confusion about it generates a large share of the quarrels in this area. To say that a quotation is fabricated is to make a claim about a text: no such sentence has been located in the writings of the person named. To say that a quotation is distorted is to claim that a real sentence exists and that the circulated version has altered its wording or amputated its context until the sense shifted. To say that a quotation is borrowed is to claim that the sentence is real and correctly credited to the person who used it, but that the person did not originate it, so any argument that treats the sentence as his invention is mistaken. To say that a quotation is genuine is to claim that the words, in an identified translation, stand in an identified work at an identified place. These are four different findings, they require four different corrections, and they carry four different implications for the arguments built on the lines in question. Collapsing them into the single verdict of fake is the characteristic error of the debunking genre, and it is an error this article refuses at the level of method, not merely of tone.

The four-category test

The four-category test is the instrument this article proposes, teaches and then uses. Before repeating any attributed line, establish whether it is fabricated, distorted, borrowed or genuine, because the three failure modes require different corrections and lumping them together as fake quotes gets the borrowed and distorted cases wrong. The test sounds simple, and at the level of classification it is simple. Its value lies in forcing the person who applies it to perform distinct operations for each category instead of reaching for the nearest verdict. A fabricated line demands a search report: where was the corpus searched, in which editions, with what result. A distorted line demands collation: place the circulated wording beside the wording in the work, and beside the passage that surrounds it, and exhibit the differences. A borrowed line demands a genealogy: trace the formulation backwards through earlier writers until its circulation before the famous user is established. A genuine line demands a citation: work, part or chapter, and the translation relied on, since an exact English wording is always a translator’s wording before it is anyone else’s.

Each category also has a characteristic temptation attached to it, and naming the temptations is part of teaching the test. Fabrication tempts the auditor toward overclaiming, toward announcing that a line was invented by a named enemy at a named moment, when the evidence usually supports only the finding that no source has been located. Distortion tempts the auditor toward pedantry, toward treating every small variation between translations as a scandal, when some variations are innocent and only those that move the sense belong in the category. Borrowing tempts the auditor toward a false exposure, toward presenting an ordinary act of intellectual inheritance as a kind of theft, when the interesting finding is usually that a formula had a career before its most famous user gave it a permanent home. Genuineness tempts the auditor toward triumphalism, toward treating a verified line as a verified argument, when verification settles only who wrote the words and nothing about whether the words are true. The test disciplines all four temptations by assigning each case its own evidentiary standard and by requiring the auditor to state, for every verdict, what evidence would overturn it.

Applied to the caseload of famous lines, the test produces a distribution that surprises readers who arrive expecting a massacre. Of the lines examined in detail below, a minority are fabrications with no located source. The single most famous sentence of all, the one about religion and opium, is a distortion case, real in substance and altered in one word and in its framing. The single most famous distributive slogan is a borrowing case, real in the work where everyone says it appears, and older than that work by more than a century in its ancestry. Several lines that circulate with weak or no sourcing turn out, on inspection, to be genuine and correctly attributed, including the closing summons of the Manifesto and the most quoted sentence of the Theses on Feuerbach. An audit that reported only the fabrications would misrepresent the tradition’s textual situation almost as badly as the quotation sites do. The record is mixed, and the mixture is the finding.

There is a fifth possible outcome that the four categories must be honest enough to accommodate, which is the verdict of unresolved. Sometimes a line is attributed in print early enough and plausibly enough that fabrication cannot be asserted, yet no text has been located, and the correct entry in the audit is that the attribution stands unproven rather than condemned. The distinction between no source located and source disproven matters to the ethics of the whole exercise. Where this article reports a line as fabricated, it means that searches of the collected works and of the standard scholarly resources have located nothing, that the attributions offered in circulation carry no work, chapter or letter that survives checking, and that the pattern of circulation matches invention rather than transmission. Where the evidence supports only the weaker finding, the weaker finding is what gets printed. An audit that cannot say unproven when unproven is the truth will eventually say fabricated when it is not, and will deserve the same distrust as the aggregators it criticises.

The test also scales. It was designed for Marx because the Marx caseload is large and consequential, but nothing in its structure is specific to one author. Any figure whose name carries authority and whose corpus is large enough to hide in will accumulate the same four kinds of error, and the Lenin cases included below demonstrate the portability within this article itself. Teachers who adopt the test for classroom use generally find that students transfer it to other names within a single assignment, which is the behaviour the instrument was built to produce. What the test cannot do is supply the labour. Classification is cheap once the searching, collating and tracing have been done, and the searching is the substance. The sections that follow therefore linger on method as much as on verdicts, because a reader who leaves with verdicts alone has gained a list, while a reader who leaves with the method has gained the means to audit the next line that arrives, from whatever direction, wearing whatever name.

The quotation audit table

The quotation audit table gathers the verdicts in one place so that a reader in a hurry can find a line, read its category, and follow the trail to the section where the evidence is laid out. A table of verdicts is a dangerous object, because verdicts detached from evidence are exactly how the misattribution economy reproduces itself, so the table is built to resist its own misuse. Every row names the actual source where one exists, and where none exists the row says so in words that report a search rather than pronounce an invention. The accurate wording column gives the documented wording for the distorted and genuine cases, the earlier lineage for the borrowed case, and the reason no source can be quoted for the fabricated cases. Readers who cite the table should cite the reasoning beneath it as well.

Quotation as circulated	Category	Actual source	Accurate wording or finding
Religion is the opium of the masses	Distorted	Introduction to the Critique of Hegel’s Philosophy of Right	The wording is the opium of the people, inside a passage on suffering and protest
Democracy is the road to socialism	Fabricated	No source located	Attributed to Marx in speeches and print, with no work or letter ever produced
The capitalists will sell us the rope	Fabricated	No source located	Attributed variously to Marx and to Lenin, with no text behind either attribution
A lie told often enough becomes the truth	Fabricated	No source located	Attributed to Lenin, absent from his collected writings and speeches
Give me four years to teach the children	Fabricated	No source located	Attributed to Lenin, circulating in several wordings, none traceable to a text
From each according to his ability, to each according to his needs	Borrowed	Critique of the Gotha Programme, with a longer French lineage	Genuine in the work of 1875, and formulated by earlier French socialist writers
The philosophers have only interpreted the world	Genuine	Theses on Feuerbach, thesis eleven	Correctly attributed, in the translation of the text Engels published
Working men of all countries, unite	Genuine	Closing lines of the Communist Manifesto	Correctly attributed, with wording set by the authorised translation
The first time as tragedy, the second time as farce	Genuine	Opening of The Eighteenth Brumaire of Louis Bonaparte	Correctly attributed, completing a remark credited inside the text to Hegel

Reading the table vertically teaches the lesson faster than any single row can. The fabricated group clusters around lines that function as confessions: the enemy admitting his method, the revolutionary admitting his cynicism, the tradition admitting that its road runs through institutions it publicly criticises. Confession lines should always attract suspicion, because real writers rarely supply their opponents with such perfectly shaped admissions, whereas inventors supply them constantly, since the confession is the point of the invention. The distorted and borrowed rows cluster around lines that function as doctrines, and there the failure is subtler: a word exchanged, a context removed, a pedigree forgotten. The genuine rows, finally, are not the dull residue of the exercise. They demonstrate that verification can end in confirmation, and that an audit’s willingness to confirm is what purchases its right to correct.

Category one: lines with no located source

Fabrication is the gravest verdict in the audit and the one that requires the most careful statement. A text either exists or it does not, but a searcher can only report on searches actually performed, so every fabrication verdict in this section carries an implicit rider: no source has been located in the collected works, in the standard scholarly editions, or in the reference literature that tracks these attributions, and the attributions circulating in print and online offer no work, chapter, speech or letter that a checker can open. That rider is not weakness. It is the difference between reporting and inventing, and it leaves the door open, as honest scholarship must, to the production of a genuine text by anyone who believes one exists. In case after case, that door has stood open for decades and nobody has walked through it with a page reference, while the lines themselves have been reprinted thousands of times. The pattern of circulation, confident attribution paired with permanent absence of citation, is itself evidence about how the lines were born.

Consider first the sentence about democracy as a road. It circulates in the form democracy is the road to socialism, attributed to Marx, and it does heavy work wherever it appears, because it appears to settle from inside the tradition a question that the tradition argued about for a century: whether the road to a socialist society runs through democratic institutions or around them. The sentence is quoted approvingly by democratic socialists who want Marx’s blessing for a parliamentary strategy, and quoted alarmedly by critics who want Marx’s confession that democratic forms are instrumental. Both uses require the sentence to be Marx’s. No edition of his writings contains it. No letter, speech, article or notebook entry has been produced that contains it, in any language, and the attributions that accompany it in circulation never progress beyond the bare name. What Marx demonstrably did say on the adjacent question is more conditional and more interesting, and it survives in a reported speech delivered at Amsterdam in 1872 after the Hague Congress, where he allowed that in countries such as England, America and possibly Holland the workers might reach their goal by peaceful means, while insisting that in most continental countries force would be the lever. A reported concession about particular countries, hedged by its speaker and tied to institutions he named, is a different object from a universal aphorism about democracy as such. The aphorism is the invention, and its career shows how invention works: a complicated documented position is compressed, the compression is sharpened into a slogan, and the slogan is then credited to the most authoritative name available.

The rope line performs the same operation for the opposing camp. In its fullest circulated form it promises that the capitalists will sell the revolutionaries the rope with which they will be hanged, a sentence attributed most often to Lenin and sometimes to Marx, and quoted to establish that the tradition understood its enemies as greedy to the point of self destruction and planned to exploit that greed. Attribution that migrates between two authors is a diagnostic sign in itself, since genuine quotations have one home and stay in it, while invented ones drift toward whichever name the quoter’s audience will find most damning or most inspiring. No text of Lenin’s contains the sentence. No text of Marx’s contains it. The image of the rope and the sale does not appear in the collected works of either man in any form that a searcher has located, and the printed history of the attribution is a chain of writers citing other writers, each link confident, the chain anchored to nothing. The line survives because it dramatises a real argument, that commercial appetite can override political caution, in the form of a villain’s boast. The argument exists in the literature and can be debated on evidence. The boast does not exist, and quoting it substitutes theatre for the debate.

Who first put the rope line into print?

No first printing has been established, and the audit records the case as an attribution chain rather than an origin story. Writers cite earlier writers, translators and memoirists are invoked without page references, and each retelling moves the wording slightly while keeping the image intact, which is the signature of oral transmission dressed as citation.

That answer deserves expansion, because the missing origin is instructive. Genuine quotations from this tradition have traceable births: a pamphlet dated and placed, a newspaper issue, a letter whose recipient kept it, a manuscript an editor later printed. The rope line has none of these, across more than a century of alleged existence and despite the intense scholarly scrutiny that Lenin’s corpus in particular has received from editors with every incentive to catalogue his sayings. Scholars of Leniniana maintain registers of apocrypha precisely because the attributed sayings are so numerous, and the rope line sits in those registers among the unattributed. The two Lenin lines that follow belong to the same register, and they fail in the same way, which is why the audit treats the three as a family.

The sentence about the lie is the most successful of the family. A lie told often enough becomes the truth is attributed to Lenin in countless books, speeches and articles, usually to establish that the tradition theorised deception as a technique of rule. The sentence has the ring of a maxim from a practitioner, worldly and compressed, and that ring is precisely what should make a checker pause, because Lenin’s documented prose style, argumentative, polemical, thick with qualifications and references to named opponents, does not produce free floating maxims of this shape. Searches of his collected works locate no such sentence, and no speech, article or letter has been produced containing it. The thought has been attributed to others as well, including figures of the following generation on the opposing side, and those attributions are equally undocumented, which suggests that the sentence belongs to a floating stock of wisdom about propaganda that attaches itself to whichever notorious name is nearest. A maxim in search of an author will always find one, and the author it finds will always be the one the quoter already distrusts.

The sentence about the children completes the family. In its commonest form it promises that a few years of teaching the children will plant a seed that can never be uprooted, and it is attributed to Lenin to establish that the tradition regarded schooling as indoctrination by design. The wording varies from retelling to retelling, sometimes four years, sometimes a generation, sometimes the classroom, sometimes the youth organisation, and that variation across the whole sentence, not merely in a word or two, marks the line as a paraphrase in search of a text rather than a text undergoing translation. No writing of Lenin’s containing the sentence or its substance in this form has been located. Lenin wrote extensively about education, and his documented positions, including his insistence that schooling could not be neutral in a class society, are available to be quoted accurately and argued with honestly. The invented sentence survives because it is stronger and simpler than the documented positions, and because a quoter who reaches for it is spared the reading that the documented positions would require.

Four fabrication cases, then, and four different political uses, two flattering to the tradition’s self image and two damning, failing an identical documentary test. The symmetry is worth pausing over because it disciplines the reader’s response. It is tempting, having watched an opponent’s favourite line dissolve, to lower the guard when a congenial line appears, and the audit’s usefulness depends on refusing that temptation in both directions. A separate study in this series of marxism in political discourse traces how such lines move through speeches and commentary once they enter circulation, and the movement it documents is indifferent to the politics of the carrier. The lines travel because they are useful, and they are useful because they are unattributed in every way that matters while wearing attribution as a badge.

Category two: the opium passage and the cost of a stripped context

Distortion is the most instructive category in the audit because it shows error operating on a genuine text, where the correction is not deletion but restoration. The outstanding case is the sentence about religion, which more people can quote, in its circulated form, than any other sentence Marx wrote, and which almost nobody quotes correctly or in the setting where he placed it. The circulated version does two kinds of work on the original. It changes one word, substituting masses for people, and it amputates the sentence from the paragraph that gives the sentence its tone, so that a passage about suffering and consolation is made to sound like a sneer at the credulous. Neither change is large. Together they reverse the emotional register of the text while preserving just enough of its wording to pass as quotation, which is why the case repays slow treatment. A reader who understands exactly what happened to this sentence possesses the pattern by which thousands of lesser distortions operate.

Did Marx say religion is the opium of the masses?

Not in those words. He wrote that religion is the opium of the people, in a passage that also calls it the sigh of the oppressed creature and the heart of a heartless world. The masses version and the stripped context both change the sense.

The passage stands near the opening of the introduction Marx wrote for his Critique of Hegel’s Philosophy of Right, composed in the winter of 1843 and 1844 and published in the Deutsch-Französische Jahrbücher, the short lived journal he edited with Arnold Ruge in Paris. The essay is a young man’s reckoning with the German philosophical tradition and with the Germany of his day, and the discussion of religion occupies its first paragraphs because, as the essay frames the matter, criticism of religion is the premise of all criticism, the first object that German thought had to clear away before it could address the state, law and political economy. Anyone who wants the subject treated at full length, as an account of what the tradition made of religion rather than as a question of attribution, will find it in the companion study of marx on religion explained, which follows the theme across the whole career. The concern here is narrower: what the sentences say, word by word, and what has been done to them since.

Restored to its setting, the famous sentence is the last of a sequence. Religious suffering, the passage argues, is at once an expression of real suffering and a protest against real suffering, a formulation that gives religion a double character before any drug is mentioned: it registers pain truthfully, and it objects to the conditions causing the pain, even though the objection is expressed in a form that cannot remove them. The passage then calls religion the sigh of the oppressed creature, the heart of a heartless world and the soul of soulless conditions, three metaphors of consolation and animation in a world described as lacking heart and soul, and only after those three does it arrive at the sentence the world remembers: religion is the opium of the people. The sequence matters. By the time the drug appears, the text has established that religion is what suffering people produce to make an unendurable world endurable, an expression of their condition and a protest against it, and the drug metaphor enters as a description of consolation under conditions that need consoling, not as an accusation of stupidity against the consoled.

Opium itself carried different associations for the first readers of the passage than it carries for audiences now. In the 1840s opium was the standard analgesic of European medicine, sold openly, prescribed widely, taken by labourers for pain and by mothers to quiet children, as well as feared as a poison and a habit. A metaphor drawn from that substance therefore held relief and danger in a single image: the thing that dulls pain also dulls the faculty that might address the pain’s cause, and it does both whether the sufferer is deceived or clear sighted about the transaction. Commentators dispute how much of that ambivalence Marx intended, and the dispute is a genuine one that belongs to interpretation. What belongs to attribution is narrower and settled: the passage describes religion as something the oppressed produce and use, in a world characterised as heartless, and it directs its criticism at the conditions that make the consolation necessary. The sentences that follow the famous one confirm the aim, since the essay argues that to call on people to surrender their illusions about their condition is to call on them to surrender a condition that requires illusions. The target is the condition. The drug is its symptom and its palliative at once.

What surrounds the opium sentence in the passage?

The sentence closes a sequence in which religious suffering is described as both expression of real suffering and protest against it, and religion is named the sigh of the oppressed creature and the heart of a heartless world. Restoration of that sequence turns an apparent sneer into a diagnosis of consolation.

Against that restored text, measure the circulated version and its two alterations. The substitution of masses for people looks negligible and is not. People, in the passage, names the sufferers in their humanity, the same collective subject whose heartlessness of circumstance the metaphors lament. Masses imports a later political vocabulary, the vocabulary of mass parties and mass movements and eventually of mass deception, and it tilts the sentence toward contempt for a crowd imagined as passive and manipulable. That tilt is then amplified by the amputation of context, because a reader who meets the drug sentence alone, without the sigh and the heart and the protest, naturally hears a theorist of manipulation describing a trick played on the simple. Marx’s paragraph describes nearly the opposite relationship: the sufferers are the authors of the consolation, and criticism is directed at the world requiring it. Distortion of this kind cannot be corrected by swapping one word in a quotation database. It is corrected by restoring the paragraph, which is why the audit prints the passage’s movement in full rather than offering a rival slogan.

Translation complicates the case in a way that generalises across the whole article, and the opium sentence is the right place to introduce the complication. Marx wrote in German, and the German word in the sentence is Volk, a word whose range covers people, nation, folk and populace, with connotations no single English word reproduces. Every English rendering is therefore a decision, and people is the decision taken by the standard scholarly translation printed in the collected works. A checker who compares a circulated English line against a German original must pass through the translator’s choices, and a quotation offered without a named translation conceals those choices entirely. The masses variant is not a rival translation of Volk that any scholarly edition prints; it is a corruption that entered circulation and outcompeted the documented wording because it sounded more like what the quoters expected Marx to say. Expectation is the great corruptor of texts, and it operates on translators, editors and copyists alike, always pulling the wording toward what the audience already believes the author thought.

The opium case also demonstrates why distortion is the hardest category to police socially. A fabrication can be killed, in principle, by the production of the collected works and a failed search. A distortion survives every crude defence because its defenders can truthfully say that Marx wrote about religion and opium in a famous passage, and the correction, one word and one paragraph, sounds pedantic beside the confidence of the original error. Yet the difference between the versions is the difference between two theories of religion: one in which belief is a consolation produced by sufferers under heartless conditions and criticism aims at the conditions, and one in which belief is a drug administered to a passive crowd and criticism aims at the crowd’s credulity. Arguments built on the second theory are not supported by the text routinely cited for them. Writers who find the second theory persuasive remain free to defend it, on evidence drawn from elsewhere, under their own names. What the audit removes is only the borrowed signature at the bottom of it.

Category three: the formula Marx borrowed

Borrowing is the category that debunking lists handle worst, because the borrowed case looks like a triumph for the debunker and is actually a lesson in how ideas travel. The leading case is the distributive formula that stands, in the popular imagination, as Marx’s personal motto and summary of his entire doctrine: from each according to his ability, to each according to his needs. The formula is genuinely his in the only sense that citation requires, since it stands in his Critique of the Gotha Programme, the private critical commentary he wrote in 1875 on the draft programme of the united German workers’ parties and which Engels published after his death. Anyone quoting the sentence and citing that work is quoting accurately. The error enters with the assumption, encouraged by the quotation trade, that Marx invented the formula, coined it as his slogan, and placed it at the centre of his system. All three parts of that assumption fail the record, and tracing how they fail teaches more about the tradition than the formula itself does.

Did the ability and needs formula begin with Marx?

No. The pairing of contribution according to capacity with distribution according to need circulated in French socialist writing decades before Marx used it, passing through the Saint-Simonian milieu, the writings associated with Louis Blanc and the utopian literature of the period, with roots reaching still further back.

The genealogy can be followed in outline without pretending it is a single straight line. In 1755 Étienne-Gabriel Morelly published his Code de la Nature, a design for a society of common property in which citizens would contribute according to their strength and receive according to their needs, and the distributive pairing is recognisably present there in substance a century and more before the Gotha critique. The formula in something close to its canonical wording was a commonplace of French socialism in the 1830s and 1840s: it moved through the Saint-Simonian school, whose disciples paired faculties and needs as the organising terms of their doctrine; it appeared in the orbit of Louis Blanc, whose writings on the organisation of labour made the pairing famous enough that later writers credited it to him; and it surfaced in Étienne Cabet’s description of his Icarian community. By the time Marx took it up in 1875, the sentence had been a slogan of the French movement for a generation, in a political culture he knew intimately from his Paris years and his polemics against its leading figures. He was quoting the movement to itself.

The context of the use confirms the genealogy, because in the Critique of the Gotha Programme the formula is not a banner Marx raises over his own doctrine but a standard he holds against the programme he is criticising. The passage in which it appears describes a higher phase of communist society, after the division of labour and the distinction between mental and physical work have been overcome and labour has become a primary want of life rather than merely a means to it, and only under those transformed conditions, Marx argues, can society inscribe the old slogan on its banners. The surrounding pages are austere about the society that actually emerges from capitalism, insisting that its distribution will still be governed by contribution, still marked by equal right and therefore by inequality of outcome between differently endowed individuals, defects he calls inevitable in the first phase. A writer who quotes the formula as Marx’s simple promise of the future has therefore committed a double error: crediting him with another generation’s coinage, and detaching the coinage from an argument that surrounds it with conditions. The full text repays reading, and the series guide to the critique of the gotha programme guide follows that argument page by page for readers who want the austere version rather than the banner.

None of this diminishes the formula or its famous user, and the audit is emphatic on the point because the borrowing category is so often misread as an exposure. Thinkers of the first rank routinely give permanent form to formulations that their milieu had been polishing for decades, and the polishing is itself a collective intellectual achievement. What the genealogy corrects is a specific false belief, that the sentence sprang from one mind as a definition of a doctrine, a belief that distorts both the doctrine and the mind. Marx’s distinctive contributions lie elsewhere, in the analysis of how the society around him actually worked, and treating a borrowed banner as his signature misdirects readers away from the work he actually did. The correction also cuts against a mirror error among critics, who quote the formula as evidence of a personal utopianism its user supposedly indulged, when the text shows him borrowing a utopian slogan in order to surround it with conditions its originators had not imposed. Genealogy is not a weapon. It is a map, and maps serve whoever reads them.

Category four: lines that survive the audit

An audit that only destroys would leave its readers with a lopsided picture and a corrosive habit, the habit of assuming that prominence implies falsehood. The genuine category exists to correct that picture, and the cases in it are chosen for their prominence rather than their convenience. These are lines as famous as any fabrication, quoted as often and in the same venues, which pass every stage of the verification procedure: a named work, a locatable passage, a documented wording in an identified translation. Their survival matters methodologically, because it proves the test is a test and not a machine for debunking, and it matters substantively, because the genuine lines are the ones arguments about the tradition should actually be built on.

The closing summons of the Communist Manifesto is the most reproduced sentence in the entire literature of the movement, and it is genuine. The pamphlet of 1848 ends its argument with an appeal that the standard English rendering gives as: “Working men of all countries, unite!” The wording deserves the same care the audit has applied elsewhere, because the German text addresses the proletarians of all countries, and the English form familiar worldwide descends from the authorised translation of 1888, prepared under Engels’s eye from Samuel Moore’s rendering, which chose working men where a later age might choose workers. The sentence is Marx and Engels’s, its place in the work is exact, and its career needs no invention to explain it. A companion guide to the communist manifesto readers guide in this series reads the pamphlet whole, including the ten measures that quotation traders often detach from the text and circulate as a free standing programme, a detachment the guide examines as a textual matter in its own right. For the audit the point is simple: the most famous summons in the tradition requires no doctoring, and generations of printers who reproduced it faithfully demonstrate that accuracy and circulation are compatible when institutions care to make them so.

The eleventh of the Theses on Feuerbach supplies the second genuine case, and the more philosophically loaded one. Marx wrote the theses in a notebook in the spring of 1845 as a private reckoning with the philosophy he was leaving, and Engels published them in 1888 as an appendix to his own book on Feuerbach, describing them as the first document in which the germ of the new world outlook was deposited. The final thesis, in the standard translation, opens: “The philosophers have only interpreted the world, in various ways”. The sentence continues by opposing interpretation to change, and the whole is quoted accurately so often that its accuracy can be overlooked. Its survival under audit carries a lesson about the kinds of text that resist corruption. The theses are short, numbered, canonical in form, printed in a fixed place in every edition, and attached to a commentary tradition that quotes them entire. A line with a stable address is hard to forge and easy to check, and the contrast with the floating maxims of category one, which have no address at all, states the audit’s practical moral in a single comparison.

The opening of The Eighteenth Brumaire of Louis Bonaparte, Marx’s analysis of the French coup of 1851 published in 1852, supplies the third case and demonstrates a refinement the audit has not yet needed: a genuine quotation that contains an attribution inside itself. The essay begins by crediting Hegel with the remark that all great world historical facts and personages appear twice, and then adds the completion for which the passage is remembered: “the first time as tragedy, the second time as farce”. Checker and quoter alike should notice what the text models. Marx attributes the ground remark to Hegel without a page reference, much as his careless quoters attribute remarks to him, and scholars have searched Hegel for the sentence in that form without locating it exactly, so the passage is at once a genuine Marx sentence and a small lesson in how even careful writers transmit the gist of a predecessor as a remark. The audit records the Marx sentence as genuine because Marx’s text is locatable and the wording documented, and records the embedded Hegel attribution as the kind of loose credit the verification procedure exists to catch, whoever practices it. Symmetry, once again, is not a courtesy extended to one side. It is the method, applied to the tradition’s own founder with the same instruments applied to his forgers.

Other genuine lines could be added, and the FAQ below takes up the wider question of which famous sentences pass. The three cases here establish the pattern: the lines that survive are anchored in short canonical texts or famous openings, they are quoted from editions rather than from other quotations, and their wording, where it varies, varies between identified translations rather than between a text and a rumour. That pattern is the optimistic finding of the whole audit. Misattribution is not a law of literary nature. It flourishes where texts are long, unlocated and quoted at second hand, and it retreats wherever a culture of citation gives a line an address and teaches its readers to demand one.

How to check a quotation: the verification procedure

Everything so far has exhibited verdicts. This section teaches the procedure that produces them, stated as a sequence of operations any reader can perform, because the procedure is the article’s most valuable content and the part that remains useful when the particular verdicts are forgotten. The procedure has five operations. Require a named work and a chapter or section rather than a bare attribution. Search the text, in a scholarly edition or a reliable digital surrogate, for the sentence and for its key terms. Examine the wording for compatibility with the period, the language and the author’s documented habits. Establish the translation, where the original is not English, and compare renderings. Finally, weigh the source that offered the quotation, treating aggregation sites and unsourced anthologies as no evidence at all. Each operation is cheap compared with the search through fifty volumes that sceptics imagine, and the five together settle the great majority of cases that a working writer, teacher or student will ever meet.

The demand for a named work is the operation with the highest yield per unit of effort, because it converts an unfalsifiable atmosphere of authority into a checkable claim. Marx wrote a finite number of books, pamphlets, articles and letters, and they are catalogued. A quotation that belongs to him belongs to one of them, or to a speech reported by identifiable witnesses, or to a notebook editors have printed. When a quoter cannot say which, the quotation has failed its first examination, and the burden of proof sits with the quoter rather than with the sceptic. Notice how rarely circulated lines survive this demand. The rope line, the lie line, the democracy line are never accompanied by a work and a chapter, and their carriers, when pressed, retreat to the claim that the sentiment is documented even if the sentence is not, which concedes the audit’s point while changing the subject. Sentiment is not quotation. A paraphrase honestly labelled as paraphrase is a legitimate instrument of argument, and the procedure does not forbid paraphrase; it forbids dressing paraphrase in quotation marks and a famous name.

Searching the text is the second operation, and it is easier than the size of the corpus suggests, because famous lines, if genuine, are rarely buried. A sentence that has been quoted ten thousand times in the secondary literature will be findable in the work it inhabits through indexes, through digital full text of the collected works, and through the scholarly literature that discusses the passage. The search should proceed by key terms rather than by the whole sentence, since translation may have varied the wording: search for rope, for lie, for the distinctive noun pairing, across the corpus or the relevant portion of it. A failed search is a report, not a metaphysical certainty, and the honest form of words records what was searched. In practice, the fabrication verdicts in this article rest on layered failure: no hit in the corpus, no hit in the registers of apocrypha maintained by scholars, no citation in the serious literature, and no page reference produced by any of the line’s thousands of carriers. Failure at one layer is inconclusive. Failure at all four is as close to settlement as attribution questions come.

What makes a line too perfect to trust?

A line earns suspicion when it answers a present argument too neatly, confesses exactly what the quoter’s opponents are accused of, uses the vocabulary of the quoter’s own period, and arrives without a work, a chapter or a translator attached. Perfection of fit is evidence about the quoter, not about the author.

The third operation, the examination of wording and fit, is the one that turns checking from a clerical task into a critical one. Nineteenth century German prose, rendered into English by the translators of the period and of the scholarly editions, has a grain: long periodic sentences, a philosophical vocabulary built from a small set of loaded terms, qualifications stacked inside the clause rather than appended after it. A maxim that scans like a poster, balanced and punchy and free of qualification, has usually acquired that shape in transmission or in invention, because the grain of the original resists it. The test is probabilistic, not decisive, since Marx wrote journalism and pamphlets as well as treatises and could be as punchy as any pamphleteer when the genre invited it, as the Manifesto shows on every page. Genre awareness therefore refines the suspicion: a poster shaped sentence attributed to the Manifesto deserves less suspicion than the identical shape attributed to the notebooks or to Capital, where the sentence structures are forbidding and the quotable lines are correspondingly rare and correspondingly likely to be manufactured from paraphrase.

The fourth operation, establishing the translation, is treated at length in the next section, and the fifth, weighing the offering source, deserves plain words here. Quotation aggregation sites compile without checking, copy one another, and rank by popularity, so their testimony about attribution is worth exactly nothing, neither for a line nor against it. Anthologies of quotations compiled by editors who name their sources in an apparatus are a different class of object and can be used as a starting point, provided the apparatus is followed to the work itself. Newspaper archives and scholarly journals occupy the middle ground: a quotation in a serious newspaper may rest on a subeditor’s memory, while the same quotation in a journal with footnotes can be followed home. The rule that organises all of this is the rule of the trail. A source that offers a trail, work, edition, page, translator, can be audited and therefore trusted provisionally. A source that offers only a name offers nothing, however impressive the name or however many times the offering is repeated. Readers who want a fuller account of citation practice in this field, including the conventions for citing the collected works and the manuscripts, will find one in the series guide to citing marx and engels, which treats the scholarly apparatus as a working tool rather than a ritual.

One more discipline completes the procedure, and it concerns what the checker does with a negative or a positive result after reaching it. A line that fails is not to be silently dropped while the argument it supported continues unchanged, because the argument borrowed force from the attribution and owes that force back. The honest correction states that the attribution failed, restates the argument on whatever evidence genuinely supports it, and lets the conclusion weaken if the evidence is weaker. A line that passes, equally, is not to be treated as an argument won, since verification establishes authorship and nothing else. Between those two disciplines lies the whole ethics of quotation, and it is teachable in an afternoon. The tragedy of the subject is that the afternoon is so rarely spent, in newsrooms and classrooms and debating societies where the lines do their busiest work.

Three audits performed in full

Procedures are believed when they are watched in operation, so this section performs the audit three times, slowly, on lines chosen to land in three different categories. The performances are condensed from the case discussions earlier in the article and add nothing to the verdicts. Their purpose is to exhibit the sequence of decisions, the points at which each case could have turned out otherwise, and the exact form of words in which an honest checker reports each stage.

Take first the sentence about democracy as a road. The opening demand asks for a work. The carriers of the line, in print and online, supply none, attributing the sentence to Marx without title, date or occasion, so the audit proceeds to search on the carriers’ behalf. The key terms are unusually searchable, since the sentence depends on a metaphor of roads and a pairing of democracy with socialism that Marx’s vocabulary makes distinctive. Searches of the collected works in English and of the standard German edition locate no such sentence and no close relative. The adjacent documented material, the reported remarks on peaceful and forcible roads in the Amsterdam speech and the discussions of universal suffrage scattered through the writings on France and England, differs from the aphorism in exactly the respects the audit has been trained to notice: conditionality, named countries, named institutions, hesitation where the aphorism has polish. Registers of spurious attributions consulted at the next stage list the sentence among the unattributed. At that point the verdict of fabricated is recorded, with its rider, and the audit notes what would overturn it: production of a text, in any edition, containing the sentence or a true ancestor of it. No such production has occurred. The case also yields the audit’s standard advice for writers attracted to the line: the documented Amsterdam concession, cited properly as a reported speech, gives an honest writer most of what the aphorism promised, at the price of the conditions attached, and the conditions are the history.

Take next the ability and needs formula, where the audit’s middle operations do the work. The opening demand is met instantly, since the sentence stands in the Critique of the Gotha Programme, and a searcher who opens that short text finds it in the first part, inside the discussion of distribution in a higher phase of communist society. A fabrication verdict is therefore excluded within minutes, and a distortion verdict fails as well, because the circulated wording matches the standard translation of the passage, allowance made for the rendering of needs and wants between editions. The live question is the third category, and the audit turns to genealogy. The pairing of capacity and need as distributive terms is traced backwards: present in the French socialist literature of the 1830s and 1840s, attached to the names of Blanc and the Saint-Simonians, rooted further back in Morelly’s design of the previous century. Each step of that tracing rests on printed texts with dates, so the borrowing verdict rests on documents rather than on atmosphere, and the audit records the verdict with its precise scope: the sentence is Marx’s by use and citation, the movement’s by inheritance, and nobody’s by immaculate conception. Writers citing the formula correctly cite the Gotha critique. Writers describing the formula as Marx’s invention make a claim the genealogy refutes, and the refutation, characteristically, makes the story better rather than worse, since a formula with a century of life explains its own durability in a way a coinage cannot.

Take finally the closing summons of the Manifesto, where the audit exists to be survived. The demand for a work is met by the pamphlet itself, the search finds the sentence at the end of the fourth section in every edition, and the wording operation turns up the one wrinkle the honest quoter must handle: the English form descends from the authorised translation, while the German addresses proletarians, so a writer discussing the wording should name the translation relied on, as the earlier section of this article did. Genealogy finds no earlier owner of the summons in this form, though the pamphlet’s rhetoric of union drew on a movement already rich in such appeals, and no category but genuine fits the evidence. The verdict is recorded with the same formality as the condemnations, because an audit that relaxes its paperwork for congenial results will eventually relax its standards for them too. Three cases, three categories, one procedure. A reader who has followed the performances owns the instrument now, whether or not the reader remembers a single verdict.

Translation and the manufacture of wording

Every exact English quotation from Marx is a quotation from a translator, and this elementary fact, once absorbed, reorganises the whole subject. Marx wrote his mature work in German, his journalism in German and English, his letters in whichever language his correspondent used, and the English sentences that circulate under his name descend from a layered history of rendering: the translations he supervised or authorised in his lifetime, the posthumous translations prepared under Engels’s direction, the standardised renderings of the twentieth century collected works, and the independent translations that continue to appear as publishers commission new versions of the major texts. Each layer made decisions, and the decisions differ, so the same German sentence exists in English in several legitimate forms, distinguishable by a checker who knows to ask which form is being quoted and in several illegitimate forms, distinguishable by their lack of any edition that prints them.

The authorised layer carries a special interest for the audit because Marx himself intervened in it. The French translation of Capital, produced in instalments in the 1870s, was revised by Marx so extensively that he declared the result to possess a scientific value independent of the original, a remark that tells a checker two things at once: that the author treated translation as interpretation substantial enough to alter the work, and that no translation, however authorised, can be treated as a transparent window onto the German. The first English translation of Capital, produced by Samuel Moore and Edward Aveling and edited by Engels, set the English terminology for generations, and the Manifesto’s authorised English version of 1888, from the same hand, fixed the wording of the pamphlet that the Anglophone world still quotes. When a line from those translations is quoted in a modern essay, the words are a joint production, and honest citation acknowledges the partnership in the way the genuine cases above were cited, by naming the translation where wording itself is under discussion.

Drift enters through several distinct mechanisms, and the audit distinguishes them because they carry different blame. There is, first, legitimate variation, where two competent translators render one German sentence differently and both printings are documented; a checker who finds a line in an unfamiliar wording should search for the sentence’s other renderings before pronouncing judgement, since the unfamiliar form may be a genuine translation he has not met. There is, second, the drift of quotation at second hand, where a genuine rendering is copied with small errors, the errors are copied in turn, and after enough generations the sentence on the poster differs from every edition while resembling all of them; the masses variant of the opium sentence is the type specimen of this mechanism, a single word replaced somewhere in the chain and then replicated by the authority of print itself. There is, third, the drift of improvement, where a quoter or editor sharpens a genuine sentence for effect, trimming a qualification or tightening a balance, and the sharpened form outcompetes the original because it is, by design, more quotable. And there is, fourth, back translation, where a paraphrase composed in English is rendered into the literature as a quotation and acquires a spurious German ancestry nobody ever inspects.

Why does the same sentence differ between editions?

Because translation resolves what the original leaves open. A German term with a range of meanings must become one English word, translators of different generations choose differently under different conventions, and authorised versions add the author’s own revisions, so documented variation is the normal condition of a much translated text.

The checker’s defence against all four mechanisms is the same habit of concreteness: quote from an edition, name the edition, and treat any wording that cannot be found in an edition as unquoted until it is found. That habit exposes a pleasant irony in the subject. Marx’s texts are among the most carefully edited of any modern thinker’s, thanks to the historical critical edition that prints the manuscripts, the variants and the notebooks, and thanks to the registers of his reading that let scholars watch him absorb the writers he quoted. The materials for accurate quotation exist in unusual abundance. The inaccurate tradition flourishes not because the record is obscure but because the inaccurate tradition never consults it. A related case in this series, examined in the study of the thesis antithesis synthesis myth, shows the same pattern in the realm of doctrine: a formula attributed to a thinker, taught for generations, traceable on inspection to commentators rather than to the thinker, and surviving every correction because the formula is easier to teach than the texts are to read.

Why fabrications flourish

The audit has concentrated on texts. A complete account must also explain the demand side, the reasons fabricated and distorted lines find such a ready market that supply never fails. Four pressures do most of the work, and they operate on supporters and opponents alike, which is why the phenomenon cannot be assigned to any camp’s dishonesty.

The first pressure is compression. Arguments in public life are conducted under savage constraints of space and attention, and a quotation compresses an authority’s whole corpus into a sentence that fits a poster, a broadcast segment or a social feed. Compression rewards the lines that compress best, and invented lines, designed by their circulation for compressibility, compress better than genuine ones, which drag their contexts behind them. The second pressure is closure. A genuine text invites further reading and further dispute, whereas a sharp attributed maxim appears to end discussion, and speakers under pressure reach for endings. The third pressure is tribal marking. Quoting the founder’s supposed confession or the founder’s supposed prophecy signals membership and alignment before any argument begins, and the signal works whether or not the words are his, which removes the quoter’s incentive to check and adds an incentive not to, since checking risks the signal. The fourth pressure is the asymmetry of correction. A fabrication costs nothing to launch and a great deal to refute, because the refutation must prove a negative across a large corpus, and by the time the refutation is printed the fabrication has been copied into fresh venues where the refutation will never be read. Anyone who has corrected the same line in the same publication twice understands why editors quietly despair of the genre.

Underneath the four pressures sits a subtler one that honest auditors should admit, because it operates on scholars too: the desire for the founder to have said the thing the moment needs. Every living tradition reads its founder as a contemporary in disguise, and quotation is where that reading leaves its fingerprints. Democratic socialists needed a Marx who blessed the parliamentary road, and a sentence obligingly exists. Cold warriors needed a Lenin who confessed the mechanics of deception, and a sentence obligingly exists. Devotees of the religion debates needed a Marx who sneered at believers, and the opium sentence, lightly altered and stripped, obligingly sneers. In each case the need preceded the evidence and manufactured it. Recognising the pattern in one’s own camp is the entrance examination for the audit, and it is deliberately set at the start of the method rather than the end, because a checker who cannot suspect congenial lines will produce propaganda with footnotes and call the product rigour.

The institutions that might be expected to resist these pressures have mixed records, and candour requires the audit to say so. Quotation anthologies range from scholarly instruments with full apparatus to decorative gift books assembled from other gift books, and the decorative end of the trade feeds the aggregation sites that feed the speeches. Publishers of political commentary operate under deadlines that make a fifty volume corpus search unthinkable and a search engine irresistible. Universities, for their part, punish fabricated quotation in student work with real severity while the surrounding culture rewards it in public work with applause, a contradiction students notice and remember. None of this excuses the individual quoter, since the verification procedure described in this article was designed to fit inside real deadlines and real libraries. But it explains why the procedure must be taught as a practice rather than assumed as a norm, and why the persona sections that close this article address students, teachers and working journalists as three points of intervention in one economy of error.

The newsroom hazard

The brief for this article directs that the journalist’s exposure receive extended treatment, and the direction is sound, because journalism is where fabricated quotation does its most consequential work and where the professional cost of an error is most visible. A politician’s speechwriter who launches a fabricated line risks a fact checking column. A student who copies one risks a grade. A journalist who prints one risks the standing of the publication and hands every interested party a documented error with which to discount the surrounding reporting, including the parts that were true. Quotation is, for the working reporter, a professional hazard of a particular kind: the material is seductive precisely because it arrives pre packaged, it is checked less rigorously than claims about the living because the dead cannot complain or litigate, and errors in it are preserved forever in archives that other journalists then mine as sources, so that one newsroom’s lapse becomes the next decade’s corroboration.

The archive loop deserves emphasis, because it converts a private failure into a public ratchet. A fabricated line printed in a reputable newspaper acquires the newspaper’s authority, is harvested by the aggregation sites the audit has already dismissed, is quoted from those sites into books, and is then available to be quoted back into newspapers with the books as apparent confirmation. Each circuit adds a layer of respectability while adding no evidence whatsoever. Breaking the loop requires only that one participant in it perform the verification procedure before rather than after publication, and the procedure, in its newsroom form, is scaled to the deadline: demand the work and chapter from whoever supplied the line, search the digital corpus by key terms, check the wording against the standard translation, and, where any stage fails, either drop the quotation or print the attribution honestly hedged as attributed and unverified, a formulation that costs a sentence its punch and saves the publication its credibility. Editors who institutionalise the sequence, in a style guide or a checking desk’s routine, will find that the supply of fabricated lines from contributors dries up within a year, since contributors test what a desk will wave through and adjust their habits to its standards.

What should a verified quotation card record?

A card that can be trusted later records the sentence exactly as printed in a named edition, the work and the part or chapter where it stands, the translator whose wording is being used, the surrounding sentence where context qualifies the sense, and the date on which the checker verified the entry against the edition itself.

The card is not a metaphor. Working journalists, speechwriters and lecturers all keep, or should keep, a store of verified material separated from the stream of unsorted clippings, and the discipline of the store is the discipline of the audit in miniature. The series companion tool exists for precisely this kind of keeping: verified quotation cards, with the work and chapter attached to every line and the translator named, can be kept in VaultBook, where an entry preserved with its evidence can be rechecked in seconds when a deadline asks whether a favourite line may be used again. The point of the tool is the separation it enforces. A line in the verified store has passed the procedure. A line outside it is a rumour wearing a famous name, and the professional habit the audit exists to build is the habit of treating those two objects as differently as a ledger treats cash and promises.

Two recurring newsroom situations deserve direct address, because they generate a disproportionate share of the printed errors. The first is the obituary and anniversary package, assembled under time pressure from clippings, in which a subject’s views or an opponent’s views are summarised through famous lines nobody on the desk has verified since the clippings share the same errors the package is about to repeat. The second is the quotation supplied by an interested source, a campaign, a pressure group, a partisan institute, whose research is competent exactly up to the point where competence would weaken the quotation. In both situations the supplying document will present the line with maximum confidence and minimum apparatus, and in both the desk’s defence is the same unimpressed request for a work and a chapter. Sources that possess a genuine text produce the reference with pleasure. Sources that are retailing an invention produce indignation, anecdotes about the line’s fame, and further clippings, and the response is itself the finding a checker needs.

The speech, the placard and the platform

A further constituency runs risks that the newsroom section does not cover, the politicians, staffers and campaign writers who place attributed lines before the largest audiences of all. The dynamics differ from journalism in one respect that matters: a speech is an act of advocacy, and its audience expects selection and emphasis, so a fabricated line in a speech is rarely a lapse of research alone. It is usually a choice, half conscious, to prefer the line that lands over the line that can be defended, made inside an institution that rewards landing and punishes hesitation. The correction such institutions fear is also distinctive. A speaker corrected on a quotation loses more than a point of scholarship, because the correction invites the audience to reclassify the entire speech as performance, and opponents keep a corrected quotation in circulation for years precisely because it performs that reclassification so efficiently.

The lines examined in this article map onto platform use with uncomfortable precision. The democracy aphorism is a speechwriter’s gift to a reformist platform, which is why it was invented or at least why it was preserved, and its collapse under audit leaves the platform needing the longer documented argument from the Amsterdam speech, with its conditions and its named countries, which is a better argument and a worse slogan. The rope line is the corresponding gift to the opposing platform, a confession placed in the enemy’s mouth, and its collapse leaves critics needing the documented record of revolutionary strategy, which contains material enough for honest controversy without theatrical assistance. The children line serves platforms concerned with education, where the documented writings of Lenin on schooling, with their explicit refusal of neutral education, offer quotable substance that no invention was needed to supply. In case after case the pattern repeats: the genuine record is richer than the fabrication and harder to use, demanding context the platform begrudges, and the fabrication is the price a movement pays, in credibility, for refusing the context. Staffers who internalise that arithmetic become valuable to their principals in a way that mere fluency cannot make them, because they protect the speech from the one class of attack that cannot be answered on the night.

Teachers of rhetoric might usefully set the two kinds of line side by side in a classroom and ask which the students would rather deliver. The fabricated lines win the first vote almost every time, being shorter, sharper and more certain. The lesson lands when the class is then shown the corrections, dated and sourced, and invited to imagine delivering the speech in the week after the correction appears. Eloquence that cannot survive a footnote is a liability wearing the costume of an asset, and the platform tradition that learns to prefer survivable lines will find, as the Manifesto’s genuine summons proves across more than a century and a half, that accuracy is not the enemy of memorability but its only durable foundation.

The objection that wording does not matter

Every audit of this kind meets a strong objection, and fairness requires stating it at full strength before answering it. The objection runs as follows. Quotation policing is a distraction, a pedant’s game played at the expense of substance. A fabricated line may still summarise, fairly and compactly, a position its supposed author demonstrably held, and in such a case correcting the wording changes nothing about the truth under discussion while consuming time and goodwill that serious argument needs. Worse, the debunking format flatters the debunker’s side, allowing partisans to win points on citation technicalities while avoiding the real contest over whether the tradition’s documented positions are defensible. A politics conducted by footnote, the objection concludes, is a politics that has forgotten what the argument was about, and the energy spent hunting false attributions would be better spent reading the genuine works and fighting over what they actually say.

Much of that deserves concession, and the audit concedes it in specific terms rather than in a gesture. There exist fabricated lines that function as fair paraphrase, compressing documented positions into sentences their author never wrote without distorting the positions themselves, and an honest audit identifies such cases as paraphrase rather than denouncing them as lies. The Amsterdam material illustrates the boundary: a speaker who says that Marx allowed a peaceful road in some countries is paraphrasing a documented speech fairly, and the audit’s quarrel begins only when the paraphrase is fitted with quotation marks and promoted into an aphorism about democracy as such. It is also true that attribution disputes can be used as evasions, and the audit’s symmetry rules exist precisely to deny partisans that use: a correction that applies the same test to congenial and uncongenial lines cannot serve as one side’s technicality. And it is true, finally, that verified wording settles nothing about truth. The eleventh thesis is genuinely Marx’s, and its claim about philosophy remains exactly as contestable as it was before the audit confirmed the attribution.

The answer to the objection rests on the cases where fabrication is not paraphrase but contradiction, where the invented line asserts something the documented texts oppose, so that the wording is the substance and correcting it changes what the evidence supports. The opium case is the standing example: the altered and amputated sentence attributes to Marx a theory of religion as a drug administered to a passive crowd, while the passage itself theorises religion as a consolation produced by sufferers and directs criticism at the conditions requiring it, and an argument about Marx on religion built on the circulated sentence is an argument about a position he did not take. The democracy aphorism behaves similarly in the opposite political direction, crediting its subject with a settled doctrine of the parliamentary road that his conditional, country by country discussion does not contain, so that supporters who build on it are building on air. The rope line attributes a conspiratorial boast that no text supports and that colours every subsequent reading of the tradition’s ethics. In each case the objection’s premise fails: the fabrication does not summarise the position, it replaces the position, and replacing is a substantive act with substantive victims, among them the quoter’s own audience, who are being persuaded by evidence that does not exist. Where the premise holds, where a line is honest paraphrase, the audit’s own categories protect it. The objection therefore refines the enterprise rather than refuting it, and the refined enterprise, suspicious of invention, tolerant of labelled paraphrase, and indifferent to which side a correction inconveniences, is the one this article has tried to practise throughout.

Can a genuine quotation still mislead?

Yes, and the mechanism is amputation rather than alteration. A sentence verified word for word can still mislead when it is severed from a passage that qualifies it, from the conditions attached to it, or from the argumentative setting that shows whether its author was endorsing a view or stating an opponent’s case for examination.

That possibility is the reason the audit treats context restoration as a correction equal in rank to wording correction, and why the genuine cases above were presented with their settings attached. The Gotha formula is genuine and misleads whenever its conditions are dropped. The eleventh thesis is genuine and misleads whenever it is quoted as a dismissal of thought rather than as one moment in an argument about philosophy’s tasks. Verification, in other words, has two storeys: the words must be the author’s, and the use must be the passage’s. A checker who performs only the first storey has done half the work, and quotation traders know it, which is why so much of the trade, once crude fabrication is challenged, retreats into genuine sentences deployed against their contexts. The four categories classify attributions. The two storeys govern uses, and a complete quotation culture needs both, taught together, as this article has tried to teach them.

The anthology trade and the manufacture of pedigree

Between the corpus and the quoter stands an industry, and the audit would be incomplete without describing it, because most false attributions reach their audiences through the industry’s products rather than through anyone’s deliberate fraud. Quotation anthologies are old and honourable in their best form: the great dictionaries of quotations were compiled by editors who read the works they excerpted, printed references with their extracts, and corrected their references across editions as readers reported errors. In their lesser form, the same books are assembled by copying predecessors, and a copying chain behaves exactly like the oral tradition examined above, preserving errors with the same fidelity as truths and adding fresh errors at every transcription. The commercial logic is remorseless: a quotation book sells on coverage and charm, not on apparatus, and an editor deciding whether to spend the afternoon verifying one entry against a fifty volume corpus is spending money the book’s buyers will never know was spent. The predictable outcome is two literatures wearing one costume, the scholarly anthology with its trails of references and the decorative anthology with none, and readers who cannot tell them apart because the costumes are identical.

Digitisation changed the scale of the industry without changing its logic, and in one respect worsened the outcome. The online archive made genuine texts radically more available, so that a checker in a small town can search more of Marx in an evening than a professor could reach in a month a generation ago, and the same networks made unattributed lines radically more repeatable, so that the checker is searching against a tide the same technology raised. A specific misunderstanding about the new landscape deserves correction because it appears wherever this subject is discussed: the presence of a text on a reputable archive site verifies that text, and nothing else. An archive that hosts the Critique of the Gotha Programme guarantees a reader access to what Marx wrote in that work. It confers no authority on a sentence attributed to Marx by a forum post, an image macro or an anthology merely because the sentence sounds like something in the archive, and the habit of gesturing at the archive’s existence as though it underwrote the whole cloud of attributed sayings confuses a library with a warranty.

Does a line quoted in many books count as verified?

No. Multiplicity of printing measures circulation, not verification, because books copy books in this trade and a single unexamined entry in one anthology can seed hundreds of later printings. Verification requires a chain that ends in an edition of the author’s work, opened at the passage, not a census of repetitions.

Publishers and editors who want to break the copying economy have practical instruments available, and the audit lists them as recommendations rather than complaints. An anthology that prints its source for every entry, in whatever compressed form, submits itself to correction and will be corrected, to its own improvement, by the scholarly public that uses it. A publisher’s style rule that attributed lines in manuscripts must arrive with a work and chapter reference shifts the cost of checking from the publisher, who would pay it thousands of times, to the author, who pays it once and who is usually the person who chose the line. A corrections culture that treats a false attribution as an error of fact, printed in the corrections column with the same prominence as a wrong date, changes the incentives of everyone upstream. None of these instruments requires new technology or new law. All of them exist somewhere in the trade already, and the distance between the best and worst practice in quotation publishing is a distance of will rather than of means.

Anatomy of an attribution chain

Fabrications have biographies, and reconstructing the standard biography teaches checkers where in a line’s life the decisive failure occurred. The typical chain begins with a paraphrase, offered honestly or carelessly, in a memoir, a polemic or a work of popularisation, where the writer renders a position in his own words and a later reader, or the writer himself in a later edition, hardens the rendering into direct speech. The hardened sentence then enters an anthology or a speech, acquiring the author’s name in its most authoritative form, and from there passes into textbooks, journalism and quotation sites, each venue borrowing credibility from the last. At every link the sentence is slightly adjusted to its new surroundings, a word modernised, a clause trimmed, a context shed, until the line in circulation is a collective composition authored by its transmitters, wearing the name of the one participant who wrote none of it. The rope line and the lie line both exhibit the finished state of this biography: no birthplace, a hundred godparents, wording that wanders, and a confidence that grows with each retelling in exact proportion as the evidence thins.

Two features of the chain repay a checker’s attention because they offer points of intervention. The first is the hardening moment, where paraphrase becomes quotation, since it is the only moment at which a decision was taken that all later transmitters merely inherit, and since writers can be trained, by editors and teachers, to feel the decision as it happens. The discipline is grammatical: quotation marks are a claim about documents, and a writer who reaches for them should be reaching in the same motion for the document. The second feature is the anthology moment, where a line acquires the appearance of scholarly ratification, since anthologies are where checkers can do the most good per hour spent. An error corrected in a widely used anthology dies in a hundred future books at once, which is why the compilers of the scholarly dictionaries of quotations maintain correspondence with their readers and why their volumes, alone in the trade, grow more accurate as they age. The aggregation sites invert that happy property, growing less accurate as they age because they harvest the uncorrected literature faster than corrections can propagate. A checker deciding where to spend effort should spend it where correction propagates: on the anthology, the textbook, the much quoted speech, and on the reference works of record, rather than on chasing individual repetitions across the network, which multiply faster than any corrector can follow.

The chain also explains the audit’s insistence on reporting fabrication verdicts as search results rather than as origin stories. The origin of a collective composition is genuinely lost in most cases, distributed among paraphrasers, hardeners and trimmers whose individual contributions cannot now be separated, and an auditor who invents a villain for the first link commits the same offence as the chain itself, manufacturing certainty where the record holds none. What can be stated is the condition of the chain: that it is anchored to no text, that its links cite one another in a circle, and that a circle of citations is not a source but a symptom. Readers trained to ask of any famous line not merely who said it first but what the chain of custody looks like will find the question answers most cases before the corpus search begins, because genuine lines have chains that terminate in editions, and fabricated ones have chains that terminate in other chains.

Borderline cases and the ethics of the unresolved verdict

An audit that presented only clean cases would mislead about the terrain, because the literature of attribution is full of lines that sit between the categories and test the checker’s honesty. Three exhibits from the Marx corpus display the main kinds of borderline, and the audit’s handling of them demonstrates why the unresolved verdict, clearly labelled, is a scholarly result rather than a failure of nerve.

The first exhibit is reported speech. Marx is widely quoted as disowning some of his French admirers with the remark that he, at any rate, was not a Marxist. No such sentence stands in his published writings, and a crude audit might file the line among the fabrications. The record is better than that. Engels, writing to Eduard Bernstein in 1882 and returning to the matter in later correspondence, reported the remark as Marx’s own, in French, and attributed it to Marx’s exasperation at the doctrinaire Marxism of certain French disciples. The line therefore rests on the testimony of the closest possible witness, recorded in documents that survive and can be inspected, while lacking a text in the subject’s hand. The honest classification is a fifth practical category the checker will need: reported, attributed by a named contemporary in a locatable document, to be quoted with its reporting frame attached rather than as a sentence from a work. Stripping the frame, as circulation routinely does, manufactures a quotation out of a report, and restoring the frame costs the quoter one clause and buys complete accuracy.

The second exhibit is the authorised text that its own editors later qualified. The Manifesto opens its first section by declaring the history of all hitherto existing society to be the history of class struggles, a sentence quoted everywhere and genuinely present in the pamphlet. When Engels prepared the English edition of 1888, he added a footnote restricting the claim to written history, on the ground that the communal organisation of early societies had become known only after the pamphlet was written. A quoter of the sentence in any modern context therefore handles a text whose surviving author himself narrowed its scope, and while the sentence remains genuine beyond argument, accuracy about the tradition’s position requires the footnote to travel with it in any serious use. The case refines the audit’s concept of genuineness: a line can be verbatim and correctly attributed and still require an editorial health warning, not because anyone falsified it but because the record around it kept moving.

The third exhibit is the phrase that condenses a genuine argument so successfully that the argument is forgotten. The description of the Manifesto’s programme as a ten point plan or a list of planks belongs here: the pamphlet’s second section does enumerate measures, and the enumeration is genuinely present, but the circulating detached list, shorn of the text that presents the measures as an initial programme for the most advanced countries and subordinates them to the conquest of political power and the development of production, behaves in argument like a free standing constitution its authors never drafted. The audit does not reclassify the measures, which are text, but it flags the detachment, which is a use, under the second storey of verification described earlier. Readers who want the pamphlet’s own framing of its measures should go to the text, and the series readers guide named earlier follows that framing closely. The borderline exhibits converge on a single lesson: the space between genuine and fabricated is populated, and a checker equipped only with a binary stamp will falsify the population in one direction or the other. The four categories, the reported speech rider and the unresolved verdict are the audit’s attempt to give that population its proper census.

What evidence would overturn a fabrication verdict?

Production of the text. A page in a scholarly edition, a letter in an inspected archive, a newspaper issue of the period carrying the speech, any single document that a checker can open and read would overturn the verdict for that line, and the audit records each fabrication finding as a standing invitation to produce exactly that.

The invitation is not rhetorical, because attribution scholarship does occasionally receive such productions, and lines have been promoted out of the apocrypha when manuscripts surfaced or when a speech report in a provincial newspaper was finally indexed. The history of the corpus itself encourages humility on this point: works of Marx continued to surface and to be printed for a century after his death, and the notebooks are still altering the scholarly picture. A verdict in this field is therefore a report on the present state of the evidence, dated and revisable, and the audit’s verdicts are offered in that spirit throughout. What humility does not license is the pretence that all attributions stand equal until the last archive is catalogued. The evidence against the rope line and the lie line is not a gap that one document is likely to fill but a layered pattern of absence across editions, registers and a century of motivated searching by scholars who would have prized the texts. Confidence calibrated to evidence, neither inflated into certainty nor dissolved into agnosticism, is the professional virtue the whole exercise exists to teach.

Further cases from the wider caseload

The headline cases in the audit table were chosen because they dominate circulation, but the caseload extends well beyond them, and a reader who has absorbed the method deserves to watch it work across a broader and more varied field. The cases in this section are selected to exercise corners of the test the famous lines do not reach: attribution shared between two collaborators, a phrase credited to a thinker who credited it onward to somebody else, a slogan whose commonest form is a variant rather than the documented wording, and opening sentences whose fame has made their translation history invisible.

Begin with the sentence about ruling ideas. The German Ideology, the manuscript Marx and Engels drafted in 1845 and 1846 and left unpublished in their lifetimes, contains the formulation that the ideas of the ruling class are in every epoch the ruling ideas, rendered in the standard translation as a sentence of twelve words that lecturers quote constantly. The attribution here is genuine and the passage is locatable, yet the case is more delicate than it looks, because the manuscript is a joint production with a complicated editorial history, printed in full only in the twentieth century from papers whose arrangement editors had to reconstruct. A quoter who assigns the sentence to Marx alone follows convention, since the work travels under both names and the convention of the literature assigns its formulations to the partnership, but the case teaches a checker to notice how much attribution in this field is partnership attribution, tidied by later habit into a single famous surname. Whole shelves of sentences credited to Marx are, at the level of the documents, sentences of the firm, and the tidying matters whenever an argument turns on separating the two men’s views, as arguments about the state and about nature sometimes do.

The withering away of the state supplies the sharpest instance of that tidying. The phrase in its famous form, that the state is not abolished but withers away, is Engels’s, printed in Anti-Dühring in 1878, and it summarises a line of thought the two men shared while giving it a wording that is one man’s alone. Circulation credits the wording to Marx with complete confidence and complete inaccuracy, and the error is consequential rather than clerical, because scholars who dispute the doctrine often dispute precisely the metaphor of withering, with its suggestion of a natural and automatic process, and a dispute aimed at the metaphor should be aimed at its author. The case generalises into a rule checkers can apply across the whole corpus: where a famous formulation concerns the state, nature or the dialectic in its systematic form, the probability that the wording is Engels’s rises sharply, and the audit of a Marx attribution should always include the Engels shelf before any verdict is returned. Partnership cuts both ways, and sentences of Marx’s journalism have also travelled under Engels’s name, but the traffic in the direction examined here is heavier, because Engels outlived his friend, systematised the shared doctrine in popular works, and wrote the formulations a movement could memorise.

The alternative of socialism or barbarism exhibits a different structure again, a chain in which every link credits somebody else. The phrase is associated above all with Rosa Luxemburg, who in her wartime pamphlet on the crisis of German social democracy presented the alternative as the tradition’s own and credited the thought to Engels, writing as though the stark choice between socialist transformation and relapse into barbarism had been posed by the elder partner generations earlier. Searchers who follow her credit into Engels’s writings find the components, discussions of society’s possible regression and of the alternative futures facing it, without finding the phrase itself in her form. The saying therefore stands as a formulation by Luxemburg about an idea she located in Engels, subsequently credited in circulation to Marx, to Engels and to Luxemburg in rotation, an attribution chain with three famous names and one missing text. Nothing in the case is disreputable: Luxemburg’s pamphlet is a work of argument, not a quotation dictionary, and her crediting of Engels reads as an honest gesture toward a lineage. The falsification happens downstream, where the gesture is forgotten and the phrase acquires whichever single author the quoter requires. Checkers meeting a phrase that seems to belong to an entire tradition should suspect this structure at once, because traditions coin phrases collectively far more often than founders coin them singly.

The slogan about workers and unity shows the test operating at the smallest possible scale, the scale of a single variant. The documented closing of the Manifesto, in the authorised English, summons the working men of all countries to unite. The form in which the summons actually lives on banners, badges and the walls of meeting halls addresses the workers of the world instead, a rendering that condenses the geography, exchanges the period’s working men for the more durable workers, and owes its existence to no edition at all, being the movement’s own folk translation of its founding text. Classify the variant honestly and it falls under distortion of the mildest grade: the sense survives almost intact, the source is unmistakable, no argument is falsified by the change, and yet the words on the banner are not the words in the book, and a scholar quoting the banner form as the text would be corrected by any examiner. The case earns its place in the audit because it marks the threshold of the category. Below this threshold of consequence, variation is the ordinary weather of transmission and no audit should pretend otherwise. Above it, where wording or framing begins to move arguments, the category bites. The opium case sits far above the threshold. The banner sits almost on it, and knowing where the line runs is most of what expertise in this area consists in.

The spectre that opens the Manifesto deserves a sentence in the same register. A spectre is haunting Europe is a genuine opening in the standard rendering of a German sentence whose verb suggests a going about, a walking abroad, that English translators have variously caught, and the sentence’s career illustrates how a translation can become so completely canonical that the original disappears behind it. No checker corrects the spectre. The audit simply notes that even here, at the most verified point in the entire literature, the English words belong to a translator’s decision, and that the habit of naming the translation, learned on this harmless case, is the habit that protects the checker on the dangerous ones.

The defences scholarship has built

Against the copying economy described earlier, scholarship has slowly constructed a set of defences, and a reader who knows they exist can use them instead of beginning every check from the bare corpus. The oldest defence is the quotation dictionary with an apparatus, a genre whose serious practitioners treat an attribution as a claim requiring the same support as any other factual claim. The tradition runs from the nineteenth century compilers who read their authors in full to modern works such as the Yale Book of Quotations, edited by Fred Shapiro and published in 2006, which made a method of hunting sayings back to their earliest located printings and corrected hundreds of inherited attributions in the process, many of them far from politics. The lesson such works teach a checker is procedural humility: if professional quotation hunters, working full time with great libraries, regularly discover that a line’s earliest located appearance postdates its supposed author by a century, then an amateur’s confident attribution, resting on having seen the line somewhere, deserves no confidence at all.

A second defence is the scholarly edition with its indexes and registers, the collected works and the historical critical edition behind them, whose apparatus lets a checker establish not merely whether a sentence exists but how the text that contains it was composed, revised and published. A third defence, specific to this subject, is the register of apocrypha: scholars of Lenin’s reception, confronted with an attributed output far exceeding anything a mortal could have said, have catalogued the spurious sayings with their earliest traced appearances, and similar lists exist for Marx in the scholarly literature and in the better reference works. A checker who consults such a register before searching the corpus saves hours, since the famous fabrications have usually been hunted already by somebody better equipped, and the register converts a lonely negative into a documented one. The existence of these defences reframes the persistence of error. The lines survive not because nobody can check them but because checking is a habit rather than a resource, practised inside scholarship and abandoned at its borders, in the speech, the classroom handout and the leader column where the lines do their work.

A fourth defence deserves naming because it is the one this article belongs to: the specialist literature of reception studies, which treats misattribution itself as evidence. A fabricated line, once identified, does not become worthless to the historian. It becomes a document of the period that fabricated it, registering what that period needed its enemies or its founders to have said, and the scholar of reception reads the rope line and the democracy aphorism the way an archaeologist reads a forged charter, as a false witness to the transaction it claims and a true witness to the desires of the forger’s age. This is one more reason for the audit’s care about symmetry. A collection of fabrications drawn from only one camp would not merely be unfair as polemic. It would be useless as history, since the phenomenon under study is the universal appetite of movements for founders who speak in slogans, and the appetite can only be measured across the whole field.

Searching a corpus without fooling yourself

Since the audit’s verdicts rest so often on searches, the searches themselves need a discipline, and this section sets out the failure modes of corpus searching so that a reader performing the procedure alone can recognise them. The first failure mode is the single language search. Marx’s corpus exists in German, in the English of the collected works, in the French of the period and in dozens of other renderings, and a sentence absent from the English volumes may be present in a letter printed only in German, while a sentence present in English circulation may be a back translation of a German paraphrase. A serious negative search therefore names the editions searched and their languages, and treats an English only search as a preliminary, not a finding. The second failure mode is the whole sentence search. Circulated wording differs from documented wording precisely in the cases under investigation, so a checker who searches only for the circulated string will miss the genuine ancestor whose existence changes the verdict from fabricated to distorted. Key term searching, and searching for the sentence’s conceptual skeleton, protects against that miss.

The third failure mode is overtrust in digitised text. Optical character recognition garbles exactly the features a search depends on, older typefaces defeat it in patterned ways, and a corpus search that returns nothing may have interrogated a corrupted surrogate rather than the work. Where a verdict matters, the prudent checker confirms a negative against a printed index or a second digital source. The fourth failure mode is the unreported search, in which a checker announces that a line is not in the works without saying which works, which edition, which language and which terms were tried. Such announcements are themselves attributions of a kind, claims about a corpus that later checkers cannot audit, and they reproduce, inside the correction literature, the vice the correction literature exists to fight. The form of words the audit has used throughout, no source located, followed where space allows by the scope of the search, is the repair. It converts a verdict into a result, and a result can be checked, extended or overturned, which is all that scholarship ever asks of a claim.

There is also a psychology of searching to guard against, because the searcher is rarely neutral about the outcome. A checker who hopes a line is genuine searches until it is found and stops at the first plausible ancestor. A checker who hopes it is fake searches the obvious places, finds nothing and stops with satisfaction. Both stopped too early, and the procedure’s insistence on fixed stages exists to overrule hope in both directions: the demand for a work, the key term search, the genealogy for anything found, the translation check for anything quoted. Performed in order, the stages take a famous line perhaps an hour in a good library. Skipped in the order hope prefers, they produce the literature the audit has been correcting, including, on occasion, corrections that themselves require correction, the saddest genre in the whole field and the easiest to avoid by reporting searches honestly.

What falls when a line falls

It is reasonable to ask what, beyond embarrassment, a correction actually accomplishes, since fabricated lines are rarely the load bearing members of serious arguments and their removal leaves most books standing. The answer begins with the observation that the lines are load bearing exactly where it matters most: in the compressed public forms, the speech, the editorial, the examination answer and the textbook paragraph, where a single attributed sentence often carries the entire representation of a thinker’s position for an audience that will meet no other. When the democracy aphorism falls, a generation of textbook readers loses its one sentence about Marx and political strategy, and the replacement of that sentence by the conditional Amsterdam material changes, in a small way, what the tradition is understood to have taught. Corrections propagate through the same channels as errors, only more slowly, and each corrected textbook and revised reference entry withdraws the line from future circulation at its most authoritative point of reproduction.

The deeper effect of a correction is disciplinary, and it operates on writers rather than on texts. A public culture in which famous attributions are routinely checked is a culture in which writers compose with the checker looking over the shoulder, and the anticipation of checking raises the standard of the first draft more effectively than any amount of subsequent correction. This is the sense in which the audit’s real subject has never been the ten lines in its table. The lines are specimens. The subject is the standard of evidence a literate culture applies to the cheapest and most abused of its instruments, the borrowed authority of the dead, and the standard rises when corrections are seen to happen, to be survivable, and to be applied without respect to the convenience of the corrector’s own camp. A correction culture that only ever corrected opponents would lower the standard while appearing to raise it, teaching writers that attribution is a weapon rather than a discipline. The symmetry maintained throughout this article is therefore not ornament. It is the mechanism by which corrections persuade.

Something should also be said about the writers whose lines fall, because the audit’s targets include the long dead as well as the living quoter. Anthologists who copied in good faith, speechwriters who trusted a standard reference, teachers who taught what their own teachers taught: the chain of transmission is populated at every link by people acting reasonably inside a system that did not reward checking. Assigning the system’s failure to its most visible link is satisfying and unjust, and it also misdirects reform, since the links change constantly while the system persists. The audit has accordingly named no villains. It has described mechanisms, exhibited cases, taught a procedure and recommended institutional repairs, in the belief that a reader who understands how a fabrication is manufactured will manufacture fewer of them, knowingly or not, than a reader who has merely been told which contemporaries to despise for repeating one.

There is also a quieter gain that corrections produce inside the tradition itself, easy to miss because it shows up as an absence. Every fabricated line in circulation occupies a place that a genuine passage might have occupied in the same speech, syllabus or anthology, and the genuine passages are, almost without exception, stranger, richer and more resistant to summary than their replacements. A movement educated on the Amsterdam speech rather than the aphorism knows something real about the complexity of its own inheritance. A seminar that reads the whole opium paragraph argues about a subtler theory than the one the poster sells. A reader who meets the Gotha formula with its genealogy attached has learned, in passing, that ideas have collective authorship and long apprenticeships, which is a more Marxist lesson than the myth of the solitary coinage it replaces. Correction, on this accounting, is not the subtraction of colour from a tradition’s memory but its restoration, and the audit’s tables of verdicts, dry as they look, are finally in the service of better reading: fewer signatures, more texts, and the arguments of the dead allowed to be as difficult as their authors left them.

The founder as quoter

There is an illuminating way to test whether the standards of this audit are anachronistic, imposed on the past by a fussier present, and that is to examine how the founder himself handled the words of others. Marx was one of the great quoters of his century. His pages are dense with Shakespeare, Goethe, Dante, Heine, the classical economists quoted in order to be dismantled, factory inspectors quoted in order to be believed, and parliamentary reports quoted at lengths that tried his publishers’ patience. Quotation in his hands was not decoration but method: the adversary’s own words, accurately rendered, were the strongest evidence against the adversary, and the whole architecture of Capital depends on the reader trusting that the words inside the quotation marks are the words in the documents. A writer who built his masterpiece out of other men’s sentences, cited to their reports and editions, understood the ethics of the instrument from the inside.

His practice repays attention in its details. The preface to the first edition of Capital closes its account of the reader’s difficulties by adapting a line of Dante, letting people talk while the work follows its own course, and the adaptation is offered as an adaptation, a motto rather than a citation, which is exactly the labelling discipline the audit recommends. The early manuscripts reach for Shakespeare’s Timon of Athens to characterise money as the universal agent of confusion and exchange, and the borrowing is credited to its play in the text, so that a reader can go to the source and weigh the use. Where Marx quotes the economists he names them, edition by edition in the long critical notebooks, with a polemicist’s fury and a scholar’s apparatus running side by side. None of this made him immune to the ordinary failures of a working writer: quotations from memory drift in his pages as in everyone’s, reported remarks of his contemporaries pass through his letters with the looseness of all correspondence, and scholars of the texts have spent careers regularising his references. The point is the direction of the discipline. In the corpus itself, quotation aspires to be checkable, and the aspiration is visible on the page. A tradition that quotes its founder carelessly is therefore not merely inaccurate in the abstract. It is careless by the founder’s own stated and practised standard, which is the one judgement in this whole subject that no camp can dismiss as partisan.

The case also refines the audit’s sympathy for the quoter. Marx worked before indexes, before searchable text, before the collected works existed to be consulted, quoting from a memory stocked by a prodigious reading life and from notebooks he kept for the purpose, the excerpt notebooks in which he copied passages from his reading with their sources attached. Those notebooks are the ancestor of the verified quotation card recommended earlier in this article: the founder’s own store, in which a passage and its provenance were preserved together because he knew that a sentence detached from its source decays into whatever the next argument needs it to be. The instrument has been rebuilt in every century since, in the commonplace book, the card index and the digital vault, and the technology matters less than the invariant principle the notebooks embody. Keep the words and the address together, or lose both.

The poster economy

A word is owed to the humblest and most powerful channel through which attributed lines circulate, the world of printed ephemera: posters, calendars, mugs, shirts, bookmarks, classroom wall charts and the illustrated gift book. It would be easy to treat such objects as beneath an audit’s notice, and the ease would be a mistake, because for a large share of the population these objects are the literature of quotation, the form in which famous sentences are first met and most often reencountered. Nobody cites a mug in a footnote, but the mug’s sentence enters the memory with the authority of print and the warmth of the familiar, and from memory it passes into speeches and essays whose authors would be astonished to learn where their learning came from. The ephemera trade has no checking desk, no corrections column and no scholarly readership to answer to, and its commercial test for a line is purely aesthetic: does it look well in a strong typeface over a portrait. Under that test, invented lines enjoy a structural advantage no argument can match, since they were shaped by the same test during their long selection in circulation, while genuine lines arrive with their qualifications attached and their syntax unposterlike.

The portrait deserves its own sentence in this analysis, because the pairing of face and sentence is the ephemera trade’s central device and the misattribution economy’s emblem. A sentence printed alone invites the question of its source. A sentence printed beneath a famous face answers the question visually before it can be asked, and the answer feels like evidence while being only design. The bearded face over the opium sentence, over the rope line, over sentences their subject would have disowned in three languages, performs the whole fraud in a single glance: presence impersonating provenance. Printers of ephemera are rarely villains in this story, since they print what the anthologies certify and the anthologies print what the culture repeats, but the device they use is the precise point at which the audit’s procedure can be taught to the widest audience at the lowest cost. A classroom that has learned to see the portrait as a claim rather than as a decoration has learned the first operation of the verification method without opening a single volume, and teachers report, when they use the device, that students begin asking for sources unprompted, which is the behaviour the entire apparatus of this article exists to produce.

There is a reciprocal lesson for the producers of serious books. The ephemera trade flourishes in the space left by the genuine article’s refusal to compete for memorability, and scholars who lament the poster while declining to notice that the poster answers a real hunger, for portable sentences that carry a tradition in miniature, cede the territory by default. The tradition possesses portable sentences of the first quality, verified and addressed, some of them exhibited in this audit. Putting those sentences into circulation with their addresses attached, in forms as well made as the fabrications they compete with, is a legitimate part of the correction enterprise, and the publishers of the scholarly editions, whose selected volumes already do this quietly, deserve more imitators than they have found.

The limits of the audit

Honesty about an instrument includes an account of what the instrument cannot do, and the four-category test has limits that its advocates should state as plainly as its powers. The test classifies attributions against a documentary record, and the record is neither complete nor closed. Marx’s manuscripts continued to be published for more than a century after his death, the excerpt notebooks are still being absorbed by scholarship, and correspondence surfaces in archives and auction rooms often enough that no generation of checkers is entitled to call its corpus final. A fabrication verdict is therefore always, in strictness, a verdict on the record as it stands, and the audit has written its verdicts in that form throughout. The limit does not weaken the verdicts that rest on layered absence, since a line missing from the published works, the registers, the correspondence and a century of motivated commentary is missing from everything a quoter could possibly be relying on, but it does fix their logical status as reports rather than revelations, and reports can be updated by evidence in a way revelations cannot.

The test also reaches its limit at the boundary of the oral. Founders of movements speak at meetings, dictate to secretaries, argue across dinner tables and are reported by hearers of varying gifts, and the documentary record of such speech is a gradient running from stenographic reports through letters written the same evening to memoirs composed decades later by interested parties. The reported remark about not being a Marxist was placed on that gradient earlier, with its frame restored, and the gradient itself is the finding: oral attributions are not a single class of thing, and the checker who demands of them the standards of a printed text will discard genuine testimony while the checker who accepts them at face value will canonise tavern talk. The procedure’s answer is the reporting frame, kept permanently attached, so that the reader receives the evidence at its true strength and no stronger. Where the frame has been lost in transmission and cannot be recovered, the unresolved verdict is the audit’s honest resting place, and a literature that contained more unresolved verdicts and fewer certain ones, on both sides of every attribution dispute, would be a more truthful literature than the one the quotation trade has built.

A final limit concerns the audit’s own temptation, which mirrors the temptation it corrects. The debunker’s art is pleasurable, and pleasure is a solvent of rigour in correctors as surely as affection is in quoters. A checker who has demolished three famous lines in an afternoon begins to expect the fourth to fall, and expectation, as the opium case showed, is the great corruptor of texts, pulling verdicts this time rather than wordings toward what the audience already believes. The procedures of this article, the fixed stages, the reported searches, the standing invitation to produce the text, exist as much to protect the audit from its own momentum as to protect the corpus from its quoters. Any reader who applies the test should therefore apply it, once at least, to a line the reader hopes is genuine and to a line the reader hopes is fake, in the same sitting, with the same care. The instrument that survives that exercise can be trusted with the rest of the caseload, and the checker who performs it has understood the article better than a reader who memorised every verdict in the table and skipped the method that produced them.

How the lines entered the textbooks

Between the scholarly edition and the poster stands a middle kingdom of transmission that neither the corpus search nor the ephemera analysis fully covers: the textbook, the encyclopaedia entry, the party education pamphlet, the examination syllabus and the popular history. These are the texts through which most literate people actually receive their Marx, and they have a distinctive relationship to quotation, because they must compress a corpus into chapters and a chapter into memorable sentences, and because their authors work from secondary literature under contract deadlines rather than from the archive as a vocation. A textbook writer who needs the tradition’s view of religion in one sentence reaches for the sentence the culture hands over, and the culture hands over the masses variant with the context already amputated, since that is the form in which the sentence has been optimised by fifty years of identical reaching. The writer is not careless by the standards of the trade. The trade’s standards are the problem, and they are self sealing, because the textbook that prints the corrupted sentence becomes the authority from which the next textbook copies it, and a checker challenging the sentence will be answered, in perfect sincerity, with a shelf of textbooks.

Encyclopaedias amplify the same mechanism with greater force, since their entries are treated as terminal authorities by readers who would never treat a newspaper that way, and an attributed line inside an encyclopaedia article acquires a half life measured in decades. The better encyclopaedias have responded with the same apparatus the scholarly anthologies use, references attached to claims and revision across editions, and their entries on this subject have grown correspondingly careful. The middling reference work, compiled quickly from other reference works, remains the single most efficient machine for converting a rumour into a fact that the culture possesses. Party and movement education materials form a special case within the middle kingdom, because they quote the founders devotionally as well as informationally, and devotional quotation has its own standards, in which a line’s function of binding the group can quietly outrank its documentary status in the minds of compilers who would be horrified to hear the ranking stated. The audit states it without horror, since the phenomenon appears in the educational materials of every movement that possesses founders, and since the cure is the same everywhere: the pamphlet that prints its references shames the pamphlet that does not, and movements, like publishers, respond to shame administered by their own standards.

The examination syllabus deserves separate mention because it closes the loop between the middle kingdom and the student persona addressed later in this article. Where a syllabus or a mark scheme canonises a line, the line becomes practically true for educational purposes regardless of its documentary status, since candidates are rewarded for reproducing it and penalised, in effect, for the time verification would cost them. Teachers caught in that machinery are not its authors, and the audit’s sympathy extends to them, but the machinery explains why the classroom can be the last place a correction reaches even though it is the first place accuracy is preached. Reform at this level is slow and specific: a syllabus board persuaded to attach references to its canonical quotations, a mark scheme that credits the candidate who flags an attribution problem, an examiner’s report that names the year’s most repeated fabrication. Such reforms occur, and where they occur the lines in the audit’s first category begin, measurably, to lose their educational market, which is the only market whose loss a fabrication cannot survive for long.

The audit’s sources and standards

A work that criticises other people’s citation practice owes its readers an account of its own, and this section supplies it in the same spirit of concreteness the article has demanded throughout. The verdicts rest on the standard scholarly resources for this corpus: the collected works in English for the published writings, journalism and correspondence; the historical critical edition for the manuscripts, drafts and notebooks, including material the collected works print in different form; the registers and reference literature on spurious attributions maintained by specialists in both Marx and Lenin reception; and the quotation dictionaries with apparatus, whose earliest located printings anchor the history of the circulated lines. Where a verdict depends on the contents of a particular work, the work is named in the discussion of the case, so that a reader can repeat the decisive step of the check without repeating the whole audit. Where a verdict depends on absence, the scope of the search is described in the terms set out in the section on searching, and the standing invitation to produce a text applies to this article exactly as it applies to the lines the article corrects.

The standards of grading follow from the method. Genuine requires a locatable passage in a named work and a documented wording in an identified translation, and nothing in the article is graded genuine on weaker terms. Distorted requires the genuine passage to be exhibited beside the circulated form, with the alteration specified, which is why the opium discussion lingers over a single word and a single paragraph rather than asserting a difference in tone. Borrowed requires printed antecedents with dates earlier than the famous use, produced for the formula case from the French literature, and the audit has declined to grade any other line borrowed on the strength of atmosphere or resemblance alone. Fabricated requires the layered failure described in the searching section: no location in the corpus, no entry in the serious registers other than among the apocrypha, no citation in the scholarly literature, and no reference produced by the line’s carriers. The asymmetry between the standards is deliberate. Condemnation must climb a higher stair than confirmation, because a false condemnation destroys something real while a false confirmation merely preserves something questionable, and a checker who internalises the asymmetry will find that it governs doubtful cases without further instruction: when the evidence balances, the verdict falls toward the milder finding, and the milder finding is printed with its reasoning so that stronger evidence, in either direction, can move it.

Two declarations complete the account, and both are made so that readers can discount the audit appropriately if they judge the discounts necessary. The first is that no verdict in the article was reached by consulting the political convenience of the verdict, and the internal evidence for the declaration is the distribution of the findings themselves, which take favourite lines away from admirers and critics in roughly equal number and confirm favourite lines for both with equal indifference. A reader who suspects the distribution has been arranged to look balanced is invited to apply the procedure to the cases personally, since the article’s method, unlike its conclusions, cannot be faked by arrangement. The second declaration concerns the limits of one article’s competence. Attribution scholarship is a living field whose registers grow, and the findings printed here are the findings the record supported at the time of checking, offered with the search reports that make their revision possible. An audit that claimed more than that for itself would fail its own first operation, the demand that claims arrive with their evidence attached and their scope honestly stated.

The second name: Lenin and the industrial scale of attribution

Two of the audit’s fabrication cases wear Lenin’s name rather than Marx’s, and the substitution is more than a convenience of the caseload, because the Lenin attribution industry operated at a scale and with an institutional backing that the Marx industry never matched, and its products illuminate the mechanisms of the whole trade. Lenin left a large corpus, carefully preserved, and his successors built a state pedagogy on his authority, in which the leader’s sayings were collected, printed on the walls of institutions, set for memorisation and quoted as warrants for policy across seventy years. A system that requires a founder’s sentence for every occasion will manufacture sentences for the occasions the founder neglected, and the registers of Lenin apocrypha, compiled by scholars working inside and outside that system, run to hundreds of entries: maxims on education, on deception, on organisation, on art, each polished to the same worldly compression and each missing from the fifty five volumes of the collected works that the same state published with immense scholarly care. The paradox is instructive. No corpus in the century was better edited, and no name attracted more invention, because the demand for the founder’s voice was institutional rather than literary and therefore unlimited.

The lie maxim and the children saying belong to that industry’s output, and their migration into Western circulation followed the channels the anatomy section described, with an additional twist: lines manufactured or polished inside one camp’s pedagogy were harvested by the opposing camp as confessions, so that a sentence invented to inspire loyalty in one decade could be quoted to prove cynicism in the next, its function reversing while its documentary status, nonexistent, remained the same. Attribution fog of this density defeats the naive checker who assumes that a line quoted by both sides must have some foundation, since quotation by both sides is exactly what a well travelled fabrication achieves. It also defeats the checker who assumes that a saying’s presence in an official looking collection settles the matter, because the official collections of sayings were compiled under the same demand pressures as the speeches that quoted them. Against fog of this kind the audit’s instruments are unglamorous and sufficient: the collected works are finite and indexed, the registers exist, and a sentence that can be found in neither has no claim on a reader’s belief however many institutions have repeated it. The Lenin cases thus perform a service beyond their own verdicts. They demonstrate that the verification procedure scales to the hardest environment attribution offers, a state sponsored quotation economy, and returns the same answer it returns in a schoolroom: show the text, or withdraw the quotation marks.

There is a human footnote to the industry that checkers should carry with them. Lenin himself, in his published polemics, quoted his opponents at punishing length and with documentary exactness, regarding accurate quotation of an adversary as a condition of honest controversy, and his collected works are among the most citation dense political writings of their period. The fog that gathered around his name after his death was therefore not a continuation of his practice but a replacement of it, authority quoted where he had quoted evidence. The pattern matches the Marx case so exactly that the parallel can be stated as a small law of the subject: the fog around a founder’s name thickens in proportion as the founder’s own standards are abandoned by the institutions claiming him, and the surest measure of an institution’s fidelity to a thinker is not the number of his sentences on its walls but the number it can produce, on demand, with the page attached.

Quotation in argument: what a verified line can and cannot do

The audit has policed the supply of quotations. A closing analytical section should police their use, because a verified line can still be made to perform work it cannot honestly do, and readers equipped only with verification will be defenceless against the misuse of genuine texts. Attributed sentences do three legitimate kinds of work in argument. They serve as evidence of what their author believed, which is the use the audit has assumed throughout. They serve as exhibits of a tradition’s self understanding, where the quoter’s claim concerns the tradition rather than the author, and where a line’s reception history, including its distortions, is itself the evidence. And they serve as compressed formulations, borrowed for their quality, where the quoter adopts the sentence as his own position and the famous name functions as a credit line rather than as a proof. Each use has its own standard. Evidence of belief requires the full rigour of this article: located text, documented wording, context restored. Exhibits of reception require the opposite handling in one respect, since the distorted or invented line is a genuine exhibit of the culture that circulated it, provided the quoter presents it as circulation rather than as authorship. Compressed formulations require only honesty of adoption, the willingness to defend the sentence as one’s own and to let the credit line go if the attribution fails.

The abuses map onto the uses one for one. Evidence of belief is abused by amputation, the genuine sentence severed from its qualifying passage, as the genuine cases section showed. Exhibits of reception are abused by laundering, the circulated line presented first as what everybody says and then, a paragraph later, as what the author said, the hedge dissolving between mentions. Compressed formulations are abused by authority borrowing, where the quoter adopts a sentence for its quality in public and retreats, when challenged on substance, to the position that he was only quoting, enjoying the sentence’s force and the name’s prestige while accepting responsibility for neither. A reader trained by this article can police all three abuses with a single question adapted to each use: for evidence, where is the passage; for exhibits, whose voice is this presented as being; for formulations, would the quoter print the sentence over his own name alone. The third question is the most searching, and quoters who feel its force will find their quotation practice changing upstream, at the moment of selection, which is where every reform in this article has finally been aimed.

It follows that the four-category test, for all the labour this article has spent on it, is the floor of quotation ethics and not its ceiling. The test sorts the words. Judgment must still sort the uses, and judgment is educated by exactly the material the audit has assembled: passages restored, genealogies traced, chains anatomised, corrections followed into textbooks and classrooms. A reader who has worked through the cases possesses not merely a list of verdicts but a calibrated sense of how authority is manufactured and how it can be honestly earned, and that sense will operate on the next attributed sentence the reader meets, from any quarter, wearing any name, long after the particular fate of the rope line has been forgotten. Supplying that sense, rather than winning any single attribution dispute, has been the article’s object from its first paragraph.

The reference shelf: a practical tour

Method sections in articles of this kind tend to describe resources in the abstract, and abstraction is where good intentions go to die, so this section walks the actual shelf a checker needs, naming what each item is for and what it cannot do. The shelf has four levels, and the levels correspond to the stages of the verification procedure, which means a checker can assemble the shelf gradually, beginning with the level that answers the commonest questions and adding the others as the caseload demands.

The first level is the corpus in a usable edition. For English readers the fifty volume collected works of Marx and Engels is the working instrument: it prints the published writings, the journalism, the letters and the major manuscripts in a uniform translation with indexes, and its digital availability in library subscriptions and archives puts a key term search within reach of any reader with a library card. Its limitations should be known rather than discovered. The translation is a translation, with the consequences this article has explored. The arrangement is chronological and generic rather than thematic, so a checker hunting a theme must search rather than browse. And the edition’s texts rest on editorial work that the historical critical edition has in places superseded, so a verdict of the highest consequence, a line on which a book or a lawsuit turns, should be confirmed against the critical text where one exists. For the famous caseload, the collected works suffice nine times in ten, and the tenth case announces itself by the disagreement of the secondary literature.

The second level is the critical edition and its spin offs. The Marx-Engels-Gesamtausgabe prints the texts in their original languages with the variants, the drafts and the excerpt notebooks, and no checker of English quotations will use it daily, but its existence disciplines the whole field, because claims about what Marx wrote can in principle be taken to the manuscripts’ printed record and settled there. Around it cluster the scholarly aids a serious checker keeps near: the registers of correspondence that identify letters and their dates, the bibliographies of the periodical writings that let a piece of journalism be found by its first printing, and the concordances and indexes that older scholarship built and digital search has supplemented rather than replaced. The third level is the attribution literature itself: the quotation dictionaries with apparatus, whose entries give earliest located printings; the registers of apocrypha for both names in this article’s title territory; and the reception studies that trace particular lines through the culture. A checker who consults this level first, before searching the corpus, works in the efficient order, since the famous cases have usually been hunted by specialists whose reports can be verified in minutes and whose negative results carry the weight of their method.

The fourth level is the checker’s own store, and it is the only level that grows more valuable the longer it is kept. Every verified line, entered with its edition, its translator, its section and its surrounding sentence, is a future deadline met and a future error refused, and the store compounds like capital in the one benign sense of that metaphor. The store should also hold the corpses: the lines checked and found false, entered with their verdicts and the searches that produced them, because a fabrication refused once will be offered again, by a different contributor, in a different season, wearing a different urgency, and the checker who must search from nothing each time will eventually wave one through out of fatigue. Fatigue, in the end, is the fabricator’s chief ally inside every institution, and the whole apparatus of levels and stores and registers is best understood as a set of defences against it: machinery for making the right thing the easy thing, built by people who learned that exhortation does not survive contact with a deadline.

A checker assembling the shelf on a budget should know the order of acquisition the audit recommends. The corpus comes first, since nothing else can substitute for the texts. The attribution literature comes second, since it multiplies the corpus by telling the checker where not to dig. The critical aids come third, for the minority of cases that reach them. The personal store begins on the first day and never finishes. Libraries hold the first three levels for any reader who asks, which returns the subject to its persistent moral: the resources for accuracy are public, old and unglamorous, and the scandal of the famous fabrications is not that they are hard to detect but that detection is a habit the culture teaches in its schools and forgets in its institutions, one unexamined quotation at a time.

The oral tradition of movements

One channel of transmission remains unexamined, and it is the oldest: the spoken life of movements themselves, in meetings, marches, strike committees, party schools and the long evenings of exile in which this tradition, like every tradition, actually reproduced itself. Spoken transmission has laws of its own that differ from print in ways a checker must understand, because lines that live mainly in speech acquire a characteristic pattern of damage. A sentence repeated aloud is shortened to what the breath can carry, sharpened at the point where the room responds, and regularised into rhythm, since rhythm is memory’s technology. Within a few years of retelling, a conditional sentence from a reported speech can become an unconditional maxim with a founder’s name nailed to it, and every participant in the process acted innocently, remembering the sense as the room received it and honestly believing the maxim to be the sense. The democracy aphorism bears the marks of that workshop: a reported, conditional, country specific concession, planed by retelling into a universal road. No forger is required. A movement’s memory, doing its ordinary work of compression, manufactures founders’ sayings the way a river manufactures pebbles, by friction, over time, without intent.

Recognising the oral workshop changes the checker’s posture toward the people who carry false lines, and the change matters for the correction enterprise. A speaker who retails the democracy aphorism is usually not propagating a lie but transmitting a tradition, and correction addressed to him as exposure will fail where correction addressed to him as information succeeds, since what he needs is the Amsterdam text and not a lecture on his integrity. The audit has been written in that posture throughout, describing mechanisms rather than villains, and the posture is strategic as well as charitable: traditions accept corrections from those who treat their memory as a serious thing that errs, and reject corrections from those who treat it as a fraud that lies. The scholarly literature of oral tradition, developed for epics and ballads long before it was applied to politics, supplies the deeper comfort that none of this is peculiar to one movement. Wherever a community’s identity is carried in sayings, the sayings migrate toward the forms the community needs, and founders grow more quotable as their texts grow less read. The printing press was supposed to end that migration by fixing the record. The record stands fixed, in editions of exemplary care, and the migration continues around it, which is the final proof that misattribution is not a failure of archives but a habit of cultures, corrigible only at the level of habit, one checker, one classroom and one quotation card at a time.

Questions the audit leaves open

A long audit should end its analytical work by inventorying its own ignorance, partly as a discipline and partly because the open questions in this field are genuinely interesting and some of them are answerable by readers equipped with the method. The first open question is distributional. This article examined the lines that circulate most widely, a sample selected by fame, and fame is not a random property: lines become famous by being useful, portable and dramatic, the very properties fabrication optimises, so the proportion of fabrications in the famous sample certainly overstates their proportion in the whole cloud of attributed sentences, most of which are genuine lines quoted from real passages by people doing honest work. Nobody knows the true distribution, because nobody has performed a census of attributed sentences across a defined literature, and the census is performable: take the quotation entries for this corpus in a defined set of anthologies and textbooks, classify each by the four categories with the searches reported, and publish the base rates. Until that work exists, every generalisation about how much of the tradition’s quotation is false, including the pessimistic generalisations this article has resisted and the optimistic ones it has also resisted, rests on a biased sample, and the honest sentence is that the famous end of the distribution is corrupt in known ways while the bulk beneath it is unmeasured.

The second open question concerns the hardening moment identified in the anatomy of the chain. Scholars can often show that a line was a paraphrase in one decade and a quotation in the next, but the precise printing at which quotation marks first closed around it is rarely found, and finding it matters, because the hardening is the one point in the chain where a single identifiable writer made the decision on which all later circulation depends. The digitisation of nineteenth and twentieth century periodicals is making such findings possible for the first time, as full text archives let a researcher watch a sentence’s punctuation change across decades of reprinting, and several of the lines in this article’s table are candidates for that treatment. A researcher who located the hardening printing of the rope line would not overturn the fabrication verdict, since the line would remain unanchored in its supposed author’s works, but would transform the case from a search report into a documented history, and documented histories are what finally kill fabrications in the reference literature, because an anthology editor can correct an entry whose true origin is known more confidently than one whose origin is merely absent.

The third open question is comparative, and it may be the most valuable of the three. Marx and Lenin are not the only founders whose names attract this fog, and the audit’s procedures, developed on their caseload, invite application to the other heavily quoted figures of the modern period: statesmen whose speeches were drafted by others, scientists whose popular image was built by journalists, religious and business figures whose sayings were compiled by disciples with interests of their own. Whether the four categories distribute themselves similarly in those caseloads, whether fabrication flourishes most where a state pedagogy demanded sayings or where a commercial market did, whether reported speech forms a larger share where the founder was primarily a speaker rather than a writer, are questions with real answers that nobody has yet assembled on a common method. The four-category test was built to be portable, and its author in this article offers it to that comparative work without proprietary feelings, since a test, like a quotation, earns its authority only by surviving the uses of people who did not invent it. Readers who apply it elsewhere and find its categories wanting are invited to say so with cases attached. That invitation, extended to critics of the instrument itself, is the same standing invitation the audit extends to defenders of the lines it condemns, and closing the article’s analysis on it is deliberate: one procedure, one standard of evidence, applied to founder and fabrication and auditor alike.

A short code for quoters

The argument of this article can be compressed, for the reader who intends to use it, into a small number of working rules, stated in prose because the rules govern prose and should be met in the form they govern. Quote from editions, never from other quotations, and where a line reaches you without an edition attached, treat the missing attachment as the first fact about the line. Name the work and the part of it whenever you print a sentence inside quotation marks, and name the translator whenever the wording itself carries weight in your argument. Label paraphrase as paraphrase without embarrassment, since a fairly labelled summary of a documented position is a stronger instrument than a borrowed sentence that cannot survive checking, and it has the additional virtue of being yours. Restore context where context qualifies, most of all where the qualification inconveniences your case, because that is the circumstance in which readers, and history, will judge whether your citations are evidence or decoration. Keep what you have verified in a store apart from what you have merely collected, and let the boundary between the two stores be the most jealously guarded line in your working life. Finally, apply every one of these rules with special rigour to the lines you love, because affection is the solvent in which checking dissolves, and the line that flatters your argument is the line your argument can least afford to lose in public.

None of this requires a reader to become a scholar of the corpus, and none of it is in tension with writing vividly, arguing forcefully or quoting the tradition with pleasure where the tradition can bear the weight. The genuine lines are magnificent. The eleventh thesis has survived every audit for well over a century and lost none of its force in the process, and the Manifesto’s summons has carried more meetings than any fabrication in the table. Accuracy, practised as a habit rather than performed as a scolding, turns out to be not the tax a writer pays on quotation but the condition of quotation’s power, since a sentence that can be traced to its page borrows the authority not merely of a famous name but of a verifiable world in which names stand behind their words. That world is the one the four-category test is built to keep open, one checked line at a time.

The sentence and the signature

Stand back from the caseload and a single shape recurs in every category, and naming it gives the reader a final instrument to carry away. In each failure, a sentence and a signature have been separated and rejoined wrongly: a genuine sentence given a grandeur its setting denies, a movement’s sentence given a single author, an invented sentence given the most valuable signature available. The whole misattribution economy is a trade in signatures, conducted because signatures are the currency in which intellectual authority is denominated, and the audit’s procedure is, at bottom, a set of assay techniques for that currency. The four categories are the four ways an assay can fail or pass. The verification steps are the assayer’s bench. The quotation card, the reference shelf, the reported search and the restored paragraph are the assayer’s tools, and the reader who has followed the argument this far owns the bench entire.

What the assay protects, finally, is not the reputation of one thinker, which is sturdy enough to survive both his forgers and his defenders, and not the pride of the scholarly guild, which can look after itself. It protects the reader’s one irreplaceable asset, the ability to know whose argument is being encountered at any moment: the founder’s, the movement’s, the translator’s, the anthologist’s or the forger’s. Arguments cannot be weighed until their owners are known, and a culture that lets signatures drift loses the ability to weigh anything, drifting itself into a condition where all sentences are equally authoritative and therefore none are. The lines in the audit table will be corrected, in time, by the slow machinery of editions and registers and revised textbooks. The habit of asking for the page will outlast every one of the verdicts, and it is the habit, rather than any verdict, that this long examination has been written to install.

For students and exam candidates

A fabricated quotation inside an assessed essay is a citation failure before it is anything else, and markers read it as evidence about the candidate’s whole method, since a writer who has not verified the most checkable sentence in the essay invites doubt about every reference that follows. The practical defence is cheap. Build your quotations from editions and from course texts that print their sources, keep the work and section beside every line you store, and treat any sentence encountered on a quotation site as unusable until you have watched it standing in the work itself. Where a syllabus writer or a lecturer supplies a famous line in class, asking after its home is not impertinence but the exact skill the discipline exists to teach, and the question, asked politely, often improves the teaching. The lines certified genuine in this audit, the eleventh thesis and the Manifesto summons among them, are safe foundations for essays precisely because their addresses are stable, while the opium sentence, handled with its paragraph and its documented wording, lets a strong candidate demonstrate method and knowledge in a single move. Examiners reward the candidate who shows the working. In this corner of the syllabus, showing the working means quoting less and citing better.

For teachers and seminar leaders

The classroom is where the attribution economy is either interrupted or reproduced, and teachers control the interruption point. A serviceable exercise runs across one seminar: distribute a dozen attributed lines without verdicts, require each student to return a classification with the evidence trail attached, and grade the trails rather than the verdicts, since the trail is the skill. Students discover within the hour that the lines they were most certain about are the ones with no homes, and that discovery, made by their own searching, outperforms any lecture on source criticism. Teachers should also audit their own materials with the same instruments, because inherited slide decks and set texts carry the accumulated errors of the copying trade, and a course that corrects its founder’s quotations in public teaches more about intellectual honesty than a course that merely preaches it. The symmetry rule belongs on the syllabus in so many words: lines flattering to every position represented in the room get checked by the same standard in the same hour, and the teacher’s own favourite sentence goes into the pile with the rest, unprotected.

For researchers, historians, and journalists

For working researchers and journalists this subject is not a curiosity but an occupational exposure, treated at length in the newsroom section earlier in the article, and the persona guidance can therefore be compressed into the standing orders that desks and research groups actually adopt. No attributed line enters print on the authority of another printing. Every line carries its work, its section and its translator, or it travels labelled as reported or attributed and unverified. Verified material is kept apart from clippings, in a store whose entries record the edition consulted and the date of checking, so that a line verified once need not be verified from nothing at every deadline, and so that a line never verified is never promoted by familiarity into the appearance of verification. Historians have the additional duty of the footnote that travels: where an argument depends on an attribution the scholarship disputes, the dispute belongs in the note, because later writers will quote the conclusion and shed the caution, and the only defence against that shedding is to print the caution where the conclusion cannot be detached from it. The registers of apocrypha kept by scholars of this corpus are working tools, consulted like dictionaries, and a newsroom or a department that keeps them within reach will find that the famous lines mostly check themselves.

Frequently Asked Questions

Q: Which famous quotes did Marx never say?

Searches of the collected writings locate no source for the sentence casting democracy as a road to socialism, none for the promise involving capitalists selling rope, and none for the saying about children and a few years of teaching, which circulates under Lenin’s name. Each attribution arrives without a work, a speech or a letter attached, and each has been reprinted for decades on the strength of repetition alone, which the audit treats as circulation rather than evidence.

Q: Did Marx say religion is the opium of the masses?

The documented sentence ends with the people rather than the masses, and it stands in the introduction to the Critique of Hegel’s Philosophy of Right, within a passage presenting religious suffering as simultaneously an expression of real suffering and a protest against it. Restoring the neighbouring metaphors, which figure religion as a sigh and as a heart supplied to a heartless world, shifts the register from contempt toward diagnosis, so the circulated variant misleads twice over, once in wording and once in tone.

Q: Did Marx say democracy is the road to socialism?

No text containing that aphorism has been produced from any period of his life, in any language, and reference works on spurious attributions list it among the unlocated. What the record does contain is a reported address delivered at Amsterdam in 1872, where a peaceful route for workers was conceded for a small number of named countries while force was expected elsewhere. The concession is conditional and reported; the aphorism is universal and invented, and the distance between them is the measure of the fabrication.

Q: Did Marx invent from each according to his ability?

The sentence stands in his Critique of the Gotha Programme and may be quoted from there with confidence, but the pairing of capacity and need as distributive principles had circulated through French socialist writing for a generation before 1875, travelling with the Saint-Simonians and with Louis Blanc, with an ancestor in Morelly’s design of the previous century. Marx adopted a slogan his audience already knew and then fenced it with conditions about the society in which alone, he argued, it could operate.

Q: How do you check whether a Marx quotation is real?

Begin by requiring the name of a work and the part of it where the sentence stands, since an attribution that cannot supply an address has failed its first test. Then search a scholarly edition or its digital equivalent by the sentence’s distinctive terms rather than by its full wording, allowing for variation between translations. Inspect the register of the sentence for period plausibility, identify the translator whose English is being quoted, and discount any source that offers only a famous name as its warrant.

Q: Why are there so many fake Marx quotations?

The corpus is vast and forbidding, the name carries exceptional authority, and public argument rewards short portable sentences far more than it rewards accuracy, so invention enjoys a permanent market among admirers seeking blessings and opponents seeking confessions alike. Repetition then impersonates corroboration, as copying chains fill search results and anthologies with what looks like independent confirmation. Correction labours under the opposite conditions, since disproving an attribution requires a search, a report and a reader willing to trade a sharp line for a careful one.

Q: Did Lenin say a lie told often enough becomes the truth?

Nothing matching that maxim has been found in Lenin’s published writings, speeches or correspondence, and scholars who catalogue his apocrypha treat the attribution as ungrounded. The sentiment migrates between notorious names, having been attached to figures of opposing camps with equal confidence and equal absence of documentation, which marks it as a floating piece of wisdom about propaganda that settles on whichever speaker an audience is prepared to believe practised deception as a doctrine.

Q: Did Marx say the capitalists will sell us the rope?

The sentence is credited in circulation to Lenin more often than to Marx, and shifting credit of that kind is itself a warning sign, because a genuine sentence has one home. Neither man’s collected works contain the image of a sale of rope for a hanging, and the printed trail consists of writers citing writers backwards into fog. An argument about commercial appetite overriding political caution exists in the literature and can be made from evidence; the boast itself cannot be quoted from anywhere.

Q: Which famous Marx quotations are genuine?

The closing appeal of the Communist Manifesto passes every stage of checking, standing at the end of the pamphlet’s fourth section in wording settled by the authorised translation. The opening of the eleventh thesis on Feuerbach passes as well, with its notebook origin and its publication by Engels fully documented. The completion Marx added to a remark he credited to Hegel, contrasting a first appearance as tragedy with a second as farce, opens The Eighteenth Brumaire and can be verified there in any edition.

Q: Does translation affect whether a quotation is accurate?

Profoundly, because an exact English sentence from a German author is always some translator’s construction, and generations of translators working under different conventions have produced several legitimate English lives for single German sentences. Accuracy therefore has to be claimed relative to a named rendering found in a named edition, not in the abstract. Drift enters when a rendering is copied with small errors, sharpened for effect by later hands, or replaced altogether by a paraphrase that hardens into quotation marks during transmission.

Q: How does the four-category test sort a quotation found in print?

Sorting begins with location. A sentence found in an identified work, matching a named translation, is genuine unless an earlier genealogy demotes it to borrowed, as happens when a formulation proves to have circulated before its famous user employed it. A sentence whose home exists but whose circulated wording or framing has shifted the sense is distorted. A sentence for which no home can be produced after searching editions and registers is recorded as fabricated, with the search reported so the verdict can be retested.

Q: What should a journalist do before printing a famous Marx line?

Treat the line as an allegation and make it prove itself on the deadline’s timescale. Ask whoever supplied it for the work and the chapter, run the distinctive terms through a digital corpus, compare the wording against the standard translation, and consult the registers of known apocrypha for the corpus. When any stage fails, the choices are to drop the sentence, to print it openly flagged as an unverified attribution, or to substitute the documented material on the same point, which is usually stronger and always defensible.

Q: Why does the opium passage read differently in full?

Because the sentence arrives in the original as the conclusion of a short argument rather than as a free standing gibe. The sentences before it describe religious suffering as a true register of pain and as an objection to the conditions producing that pain, and they cast religion as consolation fashioned by the sufferers themselves. Read after those sentences, the drug comparison describes what a heartless world drives its inhabitants to make for their own relief, and the criticism lands on the world before it touches the consoled.

Q: How do translation choices manufacture quotations Marx never wrote?

A translator must pick one English word for a German term carrying several, and a choice repeated through anthologies can harden into the supposedly original wording while rival editions print something else. Manufacture proper begins later in the chain, when copyists alter a rendering slightly, editors polish a sentence toward the quotable, or a summary written in English acquires quotation marks and a spurious pedigree. Each stage is small, defensible in isolation and invisible in the finished product, which is why the defence is to quote from editions rather than from other quotations.

Q: What evidence settles an attribution when two sources disagree?

The disagreement is settled by descending from printings about the text to the text itself, in the most scholarly edition available, and by weighing the chain each side offers. A source that terminates in a work, a section and a translator defeats a source that terminates in another secondary printing, however numerous those printings are. Where both sides terminate in documents, a contemporary report by a named witness outweighs a later attribution without provenance, and where neither terminates anywhere solid, the honest settlement is a verdict of unproven, recorded as such.