POLITICS AND POETICS

ROSA LUXEMBURG DIDN’T HAVE AN INSTAGRAM ACCOUNT

MARA VOSS

September 29, 2026

Rosa Luxemburg taught workers political economy because she believed they should be able to understand the forces shaping their lives and argue about how to change them. Today, political movements often promise to amplify people’s voices while supplying the language in which those voices are expected to speak. What was lost when political education gave way to representation, and arguments became slogans to share?

---

Rosa Luxemburg did not have an Instagram account. She had students. Beginning in 1907, she taught political economy and economic history at the Social Democratic Party school in Berlin, where each year a small group of party and trade union members came to study for several months. They were not, for the most part, people who would otherwise have found themselves at a German university. Among the students were carpenters, typesetters, tailors, bricklayers, party functionaries, and newspaper workers. They studied economics, political history, historical materialism, law, public speaking, and journalism. Luxemburg taught for two hours a day and was known to make herself available in the afternoons. She was demanding. One of her students remembered that she taught principally by asking questions, pursuing an argument until the class discovered for itself where its knowledge had become uncertain or contradictory. The object was not to memorize the conclusions of Karl Marx. It was to acquire the ability to think politically without requiring an instructor in the room.

It is difficult not to be struck by the ambition of this arrangement. A political party had decided that people who worked for a living should study political economy—not receive a summary of it, not be told what position it took on the important issues, not be supplied with a vocabulary through which to describe their grievances, but study it. Luxemburg’s course used Marx’s Capital (1867), and it contributed greatly to what eventually became her own books Introduction to Political Economy (1906–9, posthumously published in 1925) and The Accumulation of Capital (1913). Luxemburg’s theoretical work was not separate from her work in the classroom, one serious and the other charitable. Teaching workers forced Luxemburg to clarify problems that she herself had not resolved. The students were being educated by the theorist, but the theorist was also being educated by the difficulty of explaining capitalism to the students.

There was nothing particularly eccentric about the underlying idea. The German socialist movement had emerged partly from workers’ educational associations and, as it grew into a mass political force, constructed an impressive intellectual world alongside the official one: newspapers, journals, libraries, lecture programs, reading groups, publishing houses, and eventually the party school itself. The Social Democratic Party was not simply seeking the votes of workers. It was attempting to produce socialists, and producing socialists meant producing people capable of understanding the society they intended to change. One of the great ambitions of the old left, largely forgotten now, was pedagogical. The worker was not merely someone whose material interests socialism claimed to represent. The worker was expected to become an intellectual participant in politics.

There is no reason to sentimentalize this world. Political education can easily become another name for indoctrination, and socialist movements proved no more immune to stupidity, orthodoxy, vanity, factionalism, or obedience than anything else. The history that followed Luxemburg offers enough evidence of what could happen when a theory of emancipation hardens into a doctrine of historical certainty. Even within the Social Democratic Party, the school was controversial; some of Luxemburg’s opponents suspected that theoretical education was a means by which the party’s radical wing could reproduce itself. Luxemburg herself could be combative to the point of exhaustion. She spent much of her political life disagreeing with people who had previously been her allies, and occasionally with nearly everyone in the room. What remains remarkable is not that the socialist movement possessed the correct theory. It is that it believed ordinary people should possess theory at all.

This required a particular conception of what a political person was. Experience mattered, but experience alone was insufficient. A worker knew what it was to work for a wage, to be dismissed, to encounter a foreman, to pay rent, to watch prices rise, to strike, perhaps to be arrested. But socialism promised more than the validation of those experiences. It proposed that they could be understood as part of a structure. The individual experience of exploitation could become an inquiry into surplus value. The rent payment could lead to questions of property; unemployment to the organization of production; military service to nationalism and imperialism. Political consciousness was supposed to move in two directions at once, from the intimate outward toward the system, and from the system back toward the intimate. Luxemburg defended theoretical education precisely on these grounds: Workers already possessed a great quantity of knowledge from living under capitalism, but theory could organize those facts into an understanding of how the whole operated.

A room full of people expected to do their own thinking.

This is very different from one of the most persistent ideas in contemporary progressive culture, the idea of the “voice.” We are forever being asked to give people a voice—to center voices, amplify voices, make space for voices, and listen to voices, particularly those described as marginalized. There is generosity in this language and a history behind it. People excluded from political and cultural institutions have frequently been spoken about by people who never thought to listen to them. Yet the contemporary obsession with voice contains a peculiar condescension of its own. Someone, after all, owns the amplifier.

Around the marginalized person there has arisen an extraordinary professional apparatus dedicated to explaining the importance of allowing the marginalized person to speak. There are foundations, NGOs, academics, journalists, museums, consultants, activists, communications specialists, and politicians, all equipped with an increasingly standardized vocabulary through which other people’s experiences can be recognized and made publicly legible. The language is by now familiar enough to pass almost unnoticed: platforming, centering, allyship, visibility, representation, inclusion. None of these things is inherently undesirable. What is curious is how often a politics ostensibly concerned with allowing people to speak for themselves requires a professional class to explain the proper conditions under which their speech is to be heard.

Luxemburg’s classroom suggests a different ambition. She did not want merely to hear the worker speak. She wanted the worker to acquire an argument. The distinction is more important than it initially appears. A voice expresses an experience; an argument makes a claim about the world. An experience can be acknowledged; an argument can be contradicted. A voice asks to be heard; an argument enters a structure of propositions in which it can be tested, revised, rejected, or extended. The old socialist movement, at its most intellectually confident, imagined workers not as the authentic bearers of an experience to which others should defer, but as people capable of entering a more dangerous territory: They were supposed to become capable of disagreeing with their teachers.

There is something wonderfully unfashionable about this. Luxemburg did not appear before a room of workers and tell them that their lived experience was valid. She assumed it was valid and then asked them to study economics.

The difference points toward a larger transformation on the left. The political subject of the nineteenth and early twentieth centuries was generally defined by their place within a system of production: worker, peasant, bourgeois, capitalist. The political subject of the early twenty-first century is more often described through identity or marginalization. These categories are not mutually exclusive, nor did race, sex, nationality, or colonialism suddenly appear at the end of the twentieth century. The history of the left is full of people who understood that class never existed in isolation from other forms of domination. W. E. B. Du Bois hardly needed the twenty-first century to inform him that race and political economy might be related.

What changed was less the discovery of identity than the political language surrounding it. The worker was, at least theoretically, a political subject who was expected to become increasingly capable of analyzing the forces acting upon him. The marginalized person is frequently presented as possessing an experience that institutions must learn to recognize. One model moves toward political education, the other toward representation. One asks how a person might acquire the means to interpret the world; the other asks whether the world has adequately acknowledged the person’s experience within it.

The old left wanted to turn workers into intellectuals. The contemporary progressive world has become remarkably good at turning people into constituencies.

An instruction in search of an object.

This may explain something about the peculiar thinness of much contemporary political language. Walk through the culturally progressive parts of New York, London, Berlin, or Paris, and one encounters a small international vocabulary: Black Lives Matter, Housing Is a Human Right, Silence Is Violence, Defund the Police, Decolonize, Kill the Rich. The phrases concern serious matters, and their simplicity should not be confused with the absence of serious thinking behind them. There are libraries of scholarship on race, policing, colonialism, and housing. Indeed, this produces the paradox. The intellectual traditions behind contemporary progressive politics are often immensely complicated, while the public language in which that politics is expressed has become remarkably simple.

Consider “Housing Is a Human Right.” It is difficult to object to without appearing to object to housing or human rights, which is part of its rhetorical efficiency. But as a political proposition, it leaves almost everything unresolved. Who should own housing? Should land remain private property? Should rents be controlled? Should the state construct housing directly? Should tenants own buildings cooperatively? How should housing be allocated? What creates scarcity? What produces land value? What obligations follow from declaring something a right, and who assumes them? The slogan provides a moral destination while relieving us, at least temporarily, of the inconvenience of the route.

“Decolonize” is more interesting still. Decolonization once referred to an identifiable if enormously varied historical process involving empires, territories, sovereignty, and political power. The word now travels with remarkable ease. A country can be decolonized, but so can a museum, a university, a syllabus, a collection, a bookshelf, a diet, or a garden. The greater the word’s moral authority has become, the less certain its object sometimes appears. One need not oppose the reconsideration of colonial inheritances to wonder whether a concept that can describe everything continues to explain very much.

The grammar is revealing. “Black Lives Matter” is an affirmation. “Housing Is a Human Right” is an affirmation. “Silence Is Violence” is a moral equation. “Defund” is an imperative. “Decolonize” is an imperative. These are closed linguistic forms. They do not naturally invite the questions with which Luxemburg tormented her students: Why? By what mechanism? Under what circumstances? What happens next? What would contradict this proposition? What if we are wrong?

There is a temptation to blame all of this on social media, which would be convenient and mostly false. Political slogans existed long before Silicon Valley. Karl Marx and Friedrich Engels were hardly incapable of a memorable phrase, and the socialist and communist movements became exceptionally proficient at banners, posters, chants, symbols, and reductions. Nor was the newspaper culture of the old left a permanent seminar in dialectical reason. Propaganda is not an invention of Instagram.

What Instagram and Twitter did was to provide an almost perfect environment for a political culture already moving toward compression, identification, and display. A pamphlet assumes that the reader will give it time. A newspaper article assumes that one proposition may have to follow another. A lecture assumes, somewhat optimistically, that the audience will remain in the room. Social media assumes none of these things. Its fundamental condition is interruption. The political statement appears between a photograph of someone’s lunch and a hotel in Sicily, between an advertisement for shoes and a friend’s child, and it must establish its meaning before the thumb moves upward.

Political progress, measured in likes.

Under these conditions, complexity is not forbidden; it is simply at a competitive disadvantage. The successful political message is recognizable before it is understood. It establishes allegiance immediately. It tells us not only what is happening but how a decent person ought to feel about what is happening. The distinction between communicating a political idea and displaying a political identity consequently becomes difficult to maintain. One posts the exhibition, the restaurant, the holiday, and the injustice. The political statement says something about the world, but inevitably it also says something about the person who has chosen to display it: This is who I am; this is where I stand.

Luxemburg understood political identity perfectly well. She devoted her life to organizations in which membership could determine friendships, livelihoods, exile, imprisonment, and, ultimately, survival. But for her, belonging to the left could never settle the argument. More often it began one. She argued with Eduard Bernstein about revisionism, with Vladimir Lenin about organization and democracy, with Karl Kautsky over political strategy and militarism, with her own party over World War I, and with fellow revolutionaries over what revolution itself should mean. Her socialism was not a collection of approved positions, but a method of understanding relations among capital, labor, imperialism, nationalism, democracy, and political power. Her disagreements were not unfortunate interruptions of political solidarity. They were evidence that politics remained alive.

This is also why “Silence Is Violence” is such an intriguing contemporary sentence. There are obviously circumstances in which silence assists injustice. Yet as a general political proposition, the phrase does something unusual: It transforms the absence of speech into an action, and assigns that action a moral character. Silence can no longer mean uncertainty, ignorance, hesitation, privacy, confusion, contemplation, or simply the possibility that one has not yet decided what one thinks. The silent person has already taken a position. Speech therefore becomes compulsory, and once speech is compulsory, the range of acceptable things to say tends to become rather narrow.

Luxemburg’s classroom proceeded in the opposite direction. Her former student Rosi Wolfstein recalled that Luxemburg questioned the class relentlessly, probing the limits of what the students thought they knew until they could hear, as Wolfstein beautifully put it, where their knowledge “sounded hollow.” Luxemburg did not promise ready-made answers. She expected students to discover contradictions and work their way through them. The political educator was successful when the student no longer required the answer from the educator.

A right in five words. The arrangements will take longer.

The contemporary politics of representation has produced a different relationship between political authority and its audience. The language is egalitarian, but the structure can be oddly hierarchical. Experts explain structural injustice. Institutions develop the appropriate vocabulary. Activists translate experience into claims. Communications departments refine those claims into language capable of circulation. The resulting message returns to the people in whose name it was produced, who can then share it. Everyone has a voice, but remarkably few people appear to have written the script.

This is where the political celebrity enters. The politician on Instagram does not simply communicate policy. He or she creates a continuous relationship with an audience. We see the apartment, the kitchen, the dog, the train ride, the preparation for an event, the joke, the outrage, the response to yesterday’s outrage. Political authority becomes intimate. The politician appears less remote, which can be genuinely democratic in one sense, but the relationship contains an amusing linguistic confession: We call the people on the other side “followers.”

One should be careful here. There are politicians whose use of social media coexists with substantial political programs, just as there were socialist politicians a century ago who could speak for hours without saying anything interesting. Bernie Sanders, for example, belongs in significant respects to an older political culture, one formed through decades of ideological and electoral activity before social media became the dominant environment of political communication. In Germany, Gregor Gysi represents another kind of historical continuity: Whatever judgment one makes of his politics, his public language emerges from the complicated inheritance of twentieth-century German socialism and its arguments about Marxism, the GDR, reunification, and the post-communist left. These figures are useful precisely because they prevent the argument from becoming generational. The problem is not that old politicians are intelligent and young ones are foolish. It is that the institutions that once joined political participation to sustained political education have largely disappeared.

The housing question, with the answer already painted in.

What replaced them is not ignorance. If anything, contemporary political culture is saturated with information. We know more, faster, about events occurring farther away than Luxemburg could possibly have imagined. The problem is the organization of that knowledge. Information arrives continuously while explanation struggles to keep up. We become extraordinarily well informed about what has happened and remarkably dependent on others to tell us what it means.

The loss, then, is not Marxism. There is no reason to imagine that reopening the SPD party school and assigning everyone three volumes of Capital would cure contemporary politics. Nor should Luxemburg’s faith in socialism be confused with a general solution to the failures of the present. The more interesting loss is the conception of citizenship implicit in her classroom: the belief that emancipation requires people to become intellectually less dependent on those who claim to represent them.

That idea is more demanding than “giving people a voice.” It requires giving people the tools with which they may say things we dislike. It means accepting that political education can produce disagreement rather than consensus. It means treating someone not as the representative of an experience but as a person capable of abstraction, historical understanding, contradiction, and error. Above all, it means assuming that difficult ideas need not remain the property of specialists.

There is a photograph of Luxemburg at the party school. Like most photographs of classrooms, it is not especially dramatic. The great events associated with her life are absent: the revolution of 1905, the mass strikes, the demonstrations against war, the Spartacus League, the German Revolution, the Hotel Eden, the rifle butt, the Landwehr Canal. Nothing in the photograph tells us that the woman standing among these students will become one of the most famous revolutionaries of the twentieth century or that she had little more than a decade to live. It is simply a group of people assembled for the purpose of learning.

The lecture tour meets the IG carousel.

Perhaps this is the Luxemburg we have the most reason to recover. Not Luxemburg the martyr, whose murder can too easily sanctify everything she believed. Not Luxemburg reduced to a quotation on democracy or revolution, Not Luxemburg reproduced on a poster as another radical icon. Luxemburg the teacher is more troublesome because her classroom contained a political demand that remains difficult to satisfy. She assumed that people who had been denied most of the privileges of intellectual life were nevertheless capable of intellectual life. Their circumstances deserved explanation, not merely acknowledgment. Their experiences could become arguments. Their arguments could become theory. And their theory could alter the movement that had set out to educate them in the first place.

We speak today, with considerable seriousness, about empowerment. The word appears in the mission statements of institutions that possess power and in grant applications written on behalf of people who possess considerably less of it. Luxemburg’s classroom suggests a less comfortable definition. To empower someone politically might mean making oneself progressively less necessary to that person’s understanding of politics. It might mean creating not followers but opponents, interlocutors, critics, people capable of answering back.

Rosa Luxemburg did not have an Instagram account. She had students.

Mara Voss (b. 1982 in Vaduz, Liechtenstein) once enrolled in a course on political economy and was asked to leave after she questioned the examination questions. She writes about politics, education, and the trouble with people who speak on behalf of others. She has no Instagram account, although three people have opened one in her name.

Cover image: Before the followers, there was an audience.