Du Bois's Later Turn and Legacy

Du Bois published his first book in 1896 and his last in 1962. He wrote across seven decades, two world wars, the Great Depression, decolonisation and the Cold War, and he changed his mind about several things he is most famous for.

Most accounts stop at 1935. This lesson takes the later Du Bois seriously — including where he went that his admirers find awkward — and then asks what the whole body of work leaves the discipline.

A man in his nineties, packing.

In 1961, W. E. B. Du Bois is ninety-three. He has been under sustained pressure from the American state for a decade. In 1951 he was indicted as an unregistered agent of a foreign principal for his work with the Peace Information Center; he was tried at eighty-three and acquitted on a directed verdict when the judge found the government had no case. His passport was confiscated for eight years. He was widely dropped from Black organisations that could not afford the association, including the NAACP he had co-founded.

In October 1961 he applies to join the Communist Party of the United States, in a letter that is either the clarifying statement of a long life or the misjudgement of a very old man, depending on who is reading it.

Weeks later he accepts an invitation from Kwame Nkrumah to move to Ghana and direct the Encyclopedia Africana — a project he had first proposed in 1909.

He takes Ghanaian citizenship in February 1963 after the US embassy declines to renew his passport.

He dies in Accra on 27 August 1963, aged ninety-five. The following day, at the March on Washington, Roy Wilkins tells a crowd of a quarter of a million that Du Bois has died, and notes that his was "the voice that was calling you" long before most of them were born.

A state funeral is held in Accra. He is buried near Christiansborg Castle.

Sixty-seven years earlier, a young man with a fresh Harvard doctorate had knocked on five thousand doors in Philadelphia, convinced that the cure for racism was knowledge, and that the world was thinking wrong about race only because it did not know.

The question this lesson has to face is one that biographies usually dodge.

Du Bois changed his mind about fundamental things. Not in the ordinary way of a scholar refining a position, but in ways that reversed earlier commitments.

He began convinced that science would settle the race question, and abandoned that view. He began advocating integration, and by the 1930s was arguing for voluntary Black economic separatism. He began a liberal reformer, and ended a communist. He formulated the Talented Tenth and later repudiated it. He was a passionate American patriot in 1918, urging Black Americans to "close ranks" behind the war effort, and by the 1950s regarded the United States as the principal obstacle to human freedom.

Two bad ways to handle this. Treat the late positions as the senility of a great man, and quietly teach only the young Du Bois. Or treat them as the culmination, and read the whole life as building towards Accra.

The better approach is to ask what each change was a response to — because in almost every case, Du Bois changed his mind after the world provided evidence against his previous position. The changes are data about the world, not only about the man.

The turns

From science to power

The first and most consequential change.

The young Du Bois's programme (4.5.1): "The world was thinking wrong about race, because it did not know. The ultimate evil was stupidity. The cure for it was knowledge based on scientific investigation."

What happened. He produced The Philadelphia Negro. He produced sixteen Atlanta monographs. The conditions he documented did not improve; disenfranchisement spread, segregation hardened, and lynching continued at a rate of dozens to over a hundred a year.

The moment he identifies as decisive: the lynching of Sam Hose in Georgia in 1899. Du Bois, then at Atlanta, prepared a careful statement of the facts and set out to deliver it to a newspaper editor. On the way he learned that Hose had been burned alive before a crowd of thousands and that his knuckles were on display in a shop window on Mitchell Street.

He turned back.

"One could not be a calm, cool, and detached scientist while Negroes were lynched, murdered and starved."

The conclusion he drew was not that facts are useless. It is that facts do not act on their own, and that a scholar who supposes they will has misunderstood how change occurs.

This is the hardest version of the question Weber addressed in "Science as a Vocation" (4.3.9), and it is worth setting them side by side.

Weber: the scholar clarifies means, exposes consequences, and confronts the audience with inconvenient facts — and must not use the authority of the chair to advance a political position, because the audience cannot answer back.

Du Bois: he did that, for fifteen years, with more rigour than almost anyone. It did not stop the lynching.

Neither position is straightforwardly right. Weber's warning about scholarly authority being spent on politics is sound, and Du Bois's own later work is more polemical and less carefully evidenced than his early work — which arguably cost it standing that his early work had earned.

And Du Bois's objection is also sound: Weber's model presumes an audience that will act on evidence, and an institution that will transmit it. Du Bois had neither. He was excluded from the funding, the appointments and the citations (4.5.5), and the facts he produced were not being read.

The honest formulation: Weber's ethic is a professional discipline available to those with a professional position. Du Bois's turn is what happens when the position is denied and the facts are ignored. Which of them a reader finds more persuasive probably depends on which situation they are in.

From integration to self-organisation

In 1934 Du Bois published a series of Crisis editorials arguing that Black Americans should build their own economic institutions — cooperatives, businesses, schools, a "group economy" — rather than wait for an integration that was not arriving.

He was forced to resign from the NAACP over it. The organisation's programme was integrationist and its leadership regarded his position as an endorsement of segregation.

His actual argument was more precise than that. He distinguished enforced segregation, which he opposed absolutely, from voluntary self-organisation, which he regarded as necessary given that segregation existed and was not ending. His point: an organisation cannot spend forty years telling people not to build their own institutions while the institutions they are excluded from remain closed.

This debate has never gone away. It runs through the Black Power movement, through arguments about historically Black colleges, through debates on ethnic economic enclaves, and through every discussion of whether solidarity within a subordinated group reinforces or resists its subordination. Du Bois's distinction — between segregation imposed and association chosen — remains the sharpest available tool for it.

From the Talented Tenth to the Guiding Hundredth

The Talented Tenth, formulated in 1903, held that "the Negro race, like all races, is going to be saved by its exceptional men" — that a college-educated leadership class was the necessary agent of Black advancement.

By 1948 he had repudiated it, in an address to the Sigma Pi Phi fraternity that is one of the most candid pieces of self-criticism any major scholar has produced.

His objection to his own doctrine, in substance: he had assumed that education produces social responsibility. It does not. The educated Black elite he had counted on had, in large numbers, used their education for personal advancement and had detached themselves from the mass of Black Americans. He had, he says, been thinking in terms of individual talent when the problem required organised commitment.

His replacement — the "Guiding Hundredth" — specifies selection by dedication and sacrifice rather than by attainment, organised into a disciplined group, working for economic reorganisation rather than individual mobility.

This matters for the discipline because the Talented Tenth is still frequently attributed to Du Bois as his settled view, and because the self-criticism contains a general finding: credentialed elites drawn from a subordinated group do not reliably act on its behalf, and any theory of change that assumes they will is defective. That is a testable claim about class formation within subordinated populations, and the evidence for it is substantial.

From reform to socialism to communism

Du Bois joined the Socialist Party briefly in 1911. His work from Black Reconstruction (1935) onwards is explicitly Marxist in framework, though never orthodox — the wages-of-whiteness argument is a critique of Marxism as much as an application of it (4.5.4).

And his admiration for the Soviet Union, formed on visits in 1926, 1936, 1949 and 1958, is the part that requires honesty rather than management.

The straightforward statement.

Du Bois praised the Soviet Union at points when its crimes were extensive and, in some cases, known. He wrote admiringly of Stalin after his death in 1953, in terms that are indefensible. He was not alone among Western intellectuals in this, and that is a context rather than an excuse.

What he was responding to, and it is worth understanding rather than dismissing: the Soviet Union was the only major power that formally repudiated racial hierarchy, supported anti-colonial movements, and treated him as a distinguished scholar rather than as a problem. For a man who had been indicted, stripped of his passport, and dropped by organisations he founded, the contrast was not abstract.

That explains the attraction. It does not make the judgement correct.

And the lesson for reading any thinker: a scholar can be right about one thing and badly wrong about another, and the two facts do not cancel. Du Bois's analysis of the colour line was among the best-supported predictions in social science. His assessment of Stalin was wrong. Both are true, and a course that suppresses either is not teaching.

This course has applied that standard to Comte's Religion of Humanity, to Spencer and eugenics, to Marx's writings on India, to Durkheim's colonial ethnography and to Weber's Orientalism. It applies here too, and it would be a form of condescension to make an exception.

The legacy

What the whole body of work leaves the discipline, stated as claims rather than as tribute.

A method. The Philadelphia Negro established the American community study: multi-method, historically grounded, spatially mapped, and built on controlled comparison. The Chicago School's monographs follow its template.

A theory of consciousness. Double consciousness derives an interior state from a structural position (4.5.2) — the founding move of standpoint theory, and the ancestor of a large literature on identity, code-switching and the labour of managing perception.

A global frame. The colour line as the organising principle of the modern world system (4.5.3), anticipating world-systems theory, dependency theory, postcolonial studies and racial capitalism.

A mechanism. The public and psychological wage (4.5.4) — a materialist account of why cross-racial class solidarity fails, which requires no appeal to deception or to inherent prejudice, and which founded whiteness studies.

A model of historical sociology. Black Reconstruction, and specifically its final chapter, which is a sociology of knowledge applied to a live scholarly consensus.

And a finding about disciplines. That a field's account of its own origins is an institutional product, and that exclusion carries an epistemic cost paid by the field (4.5.5).

Not a counterfactual this time — a straightforward comparison of dates.

Double consciousness: 1903. Standpoint theory: 1980s. Eighty years.

The colour line as a global system: 1900. World-systems theory: 1974. Postcolonial studies: 1978 onwards. Seventy-plus years.

Colonial competition as a cause of world war, with a racial mechanism binding metropolitan workers to their rulers: 1915. The dependency and unequal-exchange literature: 1960s–70s. Fifty years.

The public and psychological wage: 1935. Whiteness studies: 1991. Fifty-six years.

The community study with survey, mapping and controlled comparison: 1899. Recognised as a founding text of American urban sociology: after 2015. A hundred and sixteen years.

Race as a constitutive rather than incidental feature of modernity: stated from 1900. Widely accepted in mainstream sociological theory: within living memory.

These gaps are the measure of the exclusion, and they are not a matter of interpretation. The work was published, dated and available throughout. A discipline that had read Du Bois in 1935 would have arrived at several of its most important twentieth-century ideas decades earlier, and would have avoided at least one major policy-relevant error about Black family structure that it made in the 1960s for want of a study published in 1908.

Go deeper

Du Bois among the founders

Where does he sit? The comparison is worth making explicitly, because it is the point of putting him in Part 4.

With Marx: shares the materialist framework and the analysis of exploitation; adds that the working class is divided by a racial line that pays real wages in status, which is why the predicted solidarity did not arrive (4.1.8, 4.5.4).

With Durkheim: shares the commitment to empirical demonstration and the use of rates and comparison; differs entirely on consensus — where Durkheim asks what holds a society together, Du Bois asks what a society's cohesion is built on and who pays for it.

With Weber: shares the German historical training, the concern with meaning and consciousness, and the three-dimensional view of stratification — the public and psychological wage is Weber's status dimension (4.3.6) demonstrated historically, arrived at independently. Weber knew and admired his work and published him in Germany.

With Simmel: shares the derivation of consciousness from structural position. The stranger (4.4.3) and double consciousness (4.5.2) are close cousins, formulated five years apart, and the difference between them — that Simmel's stranger can leave — is the whole difference between a formal analysis and one that includes power.

And what he has that none of them has: the analysis of the modern world as a racial order, and the position from which to see it. That is not a supplement to classical theory. It is a missing structural feature of the object all four of them were studying.

The debates about him now

Is he a founder or a corrective? The strong claim — Morris's — is that he founded American empirical sociology and that the discipline's history should be rewritten accordingly. The moderate claim is that he was a major early figure unjustly excluded. The debate is live and it turns partly on how much independent development one credits to Chicago.

Is "Du Bois the sociologist" a construction? Some scholars argue that the current reclamation reads back a disciplinary identity onto someone who moved freely across history, sociology, political theory, journalism and fiction, and who would not have recognised the boundary. There is something in this, and it applies to Marx and Weber too.

And the political question. How much should the late positions weigh? The reasonable answer is: they are part of the record, they should be stated accurately, they do not retroactively invalidate the empirical work, and the empirical work does not excuse them.

The practical inheritance

Contemporary fields that are directly downstream: the sociology of race and ethnicity in its entirety; whiteness studies; comparative race scholarship; urban ethnography; the sociology of education on Black educational attainment; racial capitalism; standpoint epistemology; and the sociology of knowledge as applied to disciplines themselves.

And the practices: community-based research; the insistence that studied populations be treated as agents rather than as objects; the demand that a study's categories be examined for whose assumptions they carry; and the norm — recent, and not universal — that a discipline should periodically ask whose work it has failed to read.

"Du Bois abandoned scholarship for politics." He continued producing major scholarship — Black Reconstruction (1935), Black Folk Then and Now (1939), Dusk of Dawn (1940), The World and Africa (1946) — throughout his political career.

"The Talented Tenth was his view." He repudiated it explicitly in 1948, on the ground that education does not produce social responsibility.

"His 1934 position endorsed segregation." He distinguished enforced segregation, which he opposed absolutely, from voluntary self-organisation under conditions of exclusion.

"His communism discredits his sociology." His empirical work stands or falls on its evidence, which has been checked and largely confirmed. His judgement of Stalin was wrong and should be stated as such. Both.

"He was consistent across seventy years." He reversed himself on science, integration, the Talented Tenth, American patriotism and socialism. The reversals are mostly responses to evidence.

"Du Bois has now been fully incorporated." Substantially more taught than a generation ago, and standard theory sequences still routinely present a Marx–Durkheim–Weber core. The gaps listed above are the measure of how recent the incorporation is.

Why it matters

It shows a mind changing under evidence over seventy years, which is rarer in the canon than it should be, and more instructive than consistency. The turns are traceable to specific events — a lynching in 1899, the failure of integration by 1934, the behaviour of an educated elite by 1948 — and each is a finding about the world as much as about the man.

It poses the scholar's dilemma at full strength. Weber said the scholar clarifies and does not advocate. Du Bois did exactly that, with unusual rigour, and watched it fail to matter. Neither answer is complete, and the tension between them is one every researcher who studies something that harms people will eventually face.

It supplies six concepts the discipline now cannot do without — the community study, double consciousness, second-sight, the global colour line, the public and psychological wage, and the propaganda of history — every one of which was available decades before it was used.

And it closes Part 4's largest argument. The founders built a sociology of modernity from inside Europe and largely missed that modernity was a global racial order. Du Bois saw it, said so, produced the evidence, and was not read. That is not a footnote to classical theory. It is the reason Part 6 exists.

Du Bois published from 1896 to 1962 and reversed himself on fundamental questions, each time in response to evidence.

From science to power: he believed "the ultimate evil was stupidity" and the cure was investigation. The lynching of Sam Hose (1899) — whose knuckles were displayed in an Atlanta shop window — ended that. "One could not be a calm, cool, and detached scientist while Negroes were lynched, murdered and starved."

From integration to self-organisation (1934): distinguishing enforced segregation, which he opposed absolutely, from voluntary self-organisation under conditions of exclusion. It cost him the NAACP.

From the Talented Tenth to the Guiding Hundredth (1948): a candid repudiation of his own doctrine on the ground that education does not produce social responsibility, and that selection should be by dedication and sacrifice rather than attainment.

From reform to socialism to communism: joining the CPUSA at ninety-three, after indictment, acquittal, and eight years without a passport; then Ghana, citizenship, and death in Accra on 27 August 1963, the day before the March on Washington.

And the honest note: his praise of the Soviet Union and of Stalin was wrong and should be stated as such. It does not invalidate the empirical work, and the empirical work does not excuse it — the same standard this course applied to Comte, Spencer, Marx, Durkheim and Weber.

The legacy, as dated claims: double consciousness (1903) to standpoint theory (1980s) — eighty years. The global colour line (1900) to world-systems and postcolonial theory (1974, 1978) — seventy-plus. The wages of whiteness (1935) to whiteness studies (1991) — fifty-six. The community study (1899) to recognition as a founding text — over a century.

Those gaps are the measure of the exclusion, and they are the reason Part 6 exists.

The Sam Hose lynching (1899) — the event Du Bois identified as ending his belief that scientific investigation alone would resolve the race question.

Voluntary self-organisation vs enforced segregation — Du Bois's 1934 distinction, which cost him the NAACP.

Talented Tenth — his 1903 doctrine of a college-educated leadership class, repudiated in 1948.

Guiding Hundredth — his replacement, selected by dedication and sacrifice rather than attainment.

Dusk of Dawn (1940) — his "essay toward an autobiography of a race concept," and the fullest account of his own intellectual changes.

The World and Africa (1946) — his statement that the modern world was made by the relation between Europe and the rest.

Encyclopedia Africana — the project he first proposed in 1909 and went to Ghana to direct.

One — read Dusk of Dawn. It is Du Bois's own account of how and why his views changed, subtitled "an essay toward an autobiography of a race concept." It is the best single book for understanding a mind revising itself.

Two — take the dilemma seriously. Write the strongest case for Weber's position that a scholar must not advocate, and the strongest case for Du Bois's turn. Then decide which you would follow if you had produced fifteen years of evidence that nobody acted on.

Three — check a repudiation. Find Du Bois's 1948 Talented Tenth memorandum. Compare it with the 1903 essay. Reading a major thinker dismantle their own famous idea is unusual and worth the hour.

Four — measure a gap. Take any concept in current use that you can date, and find out when it was first stated by someone the field was not reading. The exercise is uncomfortable and it is available in most disciplines.

Five — apply the standard evenly. Pick a thinker you admire and identify something they were badly wrong about. Then pick one you dislike and identify something they were right about. If you cannot do both, you are not yet reading — you are recruiting.