The dispute over economic determinism has run for more than a century without either side stating the proposition clearly enough to be argued about. Critics say Marxism holds that the economy determines everything, which no serious Marxist has defended since the 1890s. Defenders say the theory only claims that material conditions matter, which no serious critic has denied since Herodotus. Both statements are true about somebody and neither is about the position actually in contention, and the result is a debate in which the two sides swing at each other’s shadows.

Here is the proposition both sides can accept as the thing under dispute. Marxism holds that the way a society organises material production explains, in a strong sense that goes beyond mere influence, the general character of its legal, political, and intellectual arrangements and the direction of its historical change. The disagreement is about what the strong sense amounts to, whether the theory can specify it without either collapsing into a truism or making claims the historical record refutes, and what, if anything, the claim rules out.

That last clause is where this article lands. The test that resolves the dispute, and the one neither side usually applies, is whether a given version of the claim forbids anything. A determination claim is only meaningful if it names a historical outcome it rules out. Any version of economic determination that forbids nothing has bought its safety by ceasing to be a theory, and any version that forbids too much has been refuted several times over. Call this the forbidding test. Applying it to five positions on the spectrum, from mechanical determinism to bare material constraint, produces a verdict that is neither the critic’s nor the defender’s, and produces it from evidence rather than from assertion.

The parent theory this dispute is about is set out in the complete guide to historical materialism, and the model whose interpretation is directly in question is set out in the guide to base and superstructure. This article owns the determination question itself and does not restate either.

Why this question is worth the argument

Three things ride on it, and none is merely academic.

The first is whether historical materialism is a theory at all. A theory that permits any outcome explains none, and the standard charge against Marxism, made most carefully by Karl Popper, is that its determinist claims were testable, failed the tests, and were then rescued by qualifications until nothing could count against them. If the qualified version is empty, the tradition has spent a century defending a form of words.

The second is political and was decisive in practice. If the development of capitalism produces socialism with something approaching necessity, then the strategy of a socialist party is to build organisation, wait, and avoid risking its forces prematurely. If it does not, then the outcome depends on political action, and the strategic question of when and how to act reopens. The determinist reading was the official position of the largest socialist party in the world before 1914 and it underwrote a strategy. The anti-determinist readings that followed were produced by people trying to explain why that strategy had failed. This dispute has never been purely philosophical, and treating it as though it were misses why it was fought so hard.

The third is what the word commits a writer to. Describing an analysis as economically determinist is now a routine move in political argument and in journalism, usually as a criticism, and it is made by people who have not asked what the position they are naming actually holds. A writer who uses the term should be able to say which of the five positions below they are attributing to their target.

Was Marx an economic determinist?

Not in the strong sense the label implies, and he never used the phrase approvingly. His formulations range from statements that sound mechanical to statements that make historical outcomes depend on political action, and the honest reading holds both. The interesting question is which weaker version he did hold and what it forbids.

The case for the determinist reading, at full strength

The determinist reading is not a caricature invented by hostile critics. It was the mainstream reading inside the socialist movement for a generation, it was held by the most learned Marxists of that generation, and it has real textual support. Stating it properly requires putting the strongest evidence first.

The preface to the Critique of Political Economy

The preface to A Contribution to the Critique of Political Economy is the strongest determinist-sounding text in the corpus, and it was published by Marx himself, which places it in a different evidential class from drafts and letters. In a few paragraphs it states that in the social production of their existence people enter into relations that are indispensable and independent of their will; that these relations constitute the economic structure of society, the real foundation on which arises a legal and political superstructure; that the mode of production of material life conditions the general process of social, political, and intellectual life; and that it is not the consciousness of men that determines their being but their social being that determines their consciousness. It goes on to say that at a certain stage the material productive forces come into conflict with the existing relations of production, which turn from forms of development into fetters, and that an era of social revolution then begins.

Read straight, this is a strong claim. Relations of production are independent of will. The superstructure arises on the base. Being determines consciousness. The conflict between forces and relations produces revolution. Nothing in the passage says that political action decides anything, and the language of independence from will positively suggests otherwise.

The determinist reading also has the advantage of taking the text at face value rather than explaining it away, and its defenders can reasonably ask why Marx would write in this register if he meant something much weaker.

The scientific self-description

Marx and Engels both described their work as scientific in contrast to the utopian socialism they attacked, and the contrast was precisely that utopians proposed a desirable arrangement while they claimed to identify a development already in progress. Engels’s account of the transition from utopian to scientific socialism makes the point explicitly. Determinism of some kind is not an unfortunate accretion on this self-description; it is what the self-description means. A social science that identified only influences and tendencies with no determinate direction would not have been contrasted with utopianism in that way.

Marx’s own language reinforces this. He described the natural laws of capitalist production working with iron necessity towards inevitable results, in the preface to the first edition of Capital, in the course of explaining why a German reader should not dismiss a book about England. He described the country that is more developed industrially as showing the less developed the image of its own future. Both formulations are stronger than any qualified reading comfortably accommodates.

The political work the reading did

The strong reading was the official position of the Second International, and its most authoritative exponents were Karl Kautsky in Germany and Georgi Plekhanov in Russia. Kautsky’s exposition of the Erfurt programme presented the collapse of capitalism and the arrival of socialism as the outcome of economic development, with the party’s task being to organise, educate, and stand ready. Plekhanov’s writings on the monist view of history and on the role of the individual in history argued that individuals can accelerate or retard a development whose general direction is fixed by economic conditions.

This was not intellectual laziness. It was a solution to a real strategic problem. A mass party operating legally in an authoritarian state needed a reason not to attempt insurrection and a reason to believe that patience would be rewarded. A theory of inevitable development supplied both. It also supplied morale, which mattered more than any argument: a movement that believes history is on its side survives defeats that would otherwise scatter it.

The strength of the determinist reading, in short, is that it takes the published texts seriously, it makes sense of the scientific claim, and it was believed by people who knew the texts extremely well and had every reason to get them right.

The case against, at full strength

The anti-determinist case is equally well supported, and its evidence is not confined to letters and marginalia.

The Eighteenth Brumaire

The historical writings are the problem for any strong determinist reading, because they are the places where Marx applied his theory to events he had watched and had to account for outcomes that a determinist framework does not predict. The Eighteenth Brumaire of Louis Bonaparte opens with the formulation that has anchored every anti-determinist reading since: “Men make their own history, but they do not make it” as they please, under circumstances chosen by themselves, but under circumstances directly encountered, given, and transmitted from the past.

The sentence is a balance, and both halves are load-bearing. Making one’s own history is agency; the circumstances are constraint. Neither clause is decorative, and the whole of the analysis that follows depends on holding both, since Marx’s explanation of how a mediocre adventurer came to rule France turns on the specific political conduct of specific classes and factions, none of which was settled by the state of French agriculture.

The text also contains material that no mechanical reading can absorb. Marx describes a class of smallholding peasants that shares a common economic situation without forming a class in the political sense, because its members are isolated and cannot represent themselves. He describes political representatives whose ideas do not straightforwardly express their class position and explains the connection through the limits of their mental horizon rather than through direct correspondence. He describes a state apparatus that has acquired a degree of independence from every class in society. Each of these is a complication that a strong determinist reading would have to explain away, and Marx introduces them as central to his account rather than as concessions.

Engels’s letters of the 1890s

The four letters Engels wrote between 1890 and 1894 are the authorised statement of the qualified position, and they are frequently misrepresented in both directions. Critics treat them as a late retreat under pressure; some defenders treat them as a simple restatement of what had always been meant. Neither is quite right, and the letters repay reading in their own words.

To Joseph Bloch, in September of 1890, Engels wrote that according to the materialist conception of history “the ultimately determining element in history is the production and reproduction of real life”, and immediately added that neither he nor Marx had ever asserted more than this, and that anyone who twists it into the claim that the economic element is the only determining one converts the proposition into a meaningless abstract phrase. The letter goes on to describe an interaction in which political forms, legal forms, and even the traditions haunting people’s minds exercise their influence and in many cases preponderate in determining the form of historical struggles.

The letters to Conrad Schmidt of August and October 1890 develop the point institutionally, arguing that once a domain such as law or the state has separated out, it acquires its own internal logic and its own personnel with their own interests, so that its products cannot be read off from economic conditions directly. The letter to Franz Mehring of July 1893 introduces the notion that ideologists work with thought material they take over from their predecessors and are unaware of the real motives driving them, which is the passage where the phrase false consciousness appears. The letter of January 1894, long printed as addressed to Heinz Starkenburg and now identified as written to W. Borgius, states that political, legal, philosophical, religious, literary, and artistic development rests on economic development, and immediately says that all of these react on one another and on the economic base, and that it is not that the economic position is the only active cause.

Two features of this material are usually missed. Engels is not conceding under attack; he is answering correspondents who had asked how far the theory goes, and in each case he says that the vulgar version is a misunderstanding rather than an earlier position now abandoned. And the letters contain a diagnosis of why the misunderstanding arose. Engels writes that he and Marx, having to emphasise the economic side against opponents who denied it, did not always give the other factors their due place, and that the fault lay partly with them and partly with followers who took a formula for a theory. That is a considerably more honest account than either side in the modern argument usually allows.

The anti-determinist reconstructions

The twentieth century produced four serious attempts to rebuild the theory without the determinist commitments, and each is a position in this debate rather than a departure from it.

Antonio Gramsci’s reconstruction is the most influential. Writing in prison in the late 1920s and 1930s, he attacked what he called the primitive infantilism of a Marxism that expected economic crisis to produce political outcomes automatically, and replaced it with an account in which a class becomes dominant by winning consent across a range of institutions before it takes power, and holds power by continuing to win it. Economic conditions on this account set the terrain and determine what is possible; they do not settle what happens on it, and the outcome depends on organised political and cultural work whose success is not guaranteed. The complete guide to hegemony sets out the apparatus in full, and it is the single most consequential anti-determinist move in the tradition’s history.

Louis Althusser’s reconstruction runs in the opposite direction, towards more structure rather than less, and arrives at a similar refusal of simple determination. His argument is that a social formation is a complex whole in which the economic, political, and ideological levels each have their own effectivity, that any historical event is overdetermined by contradictions at several levels rather than expressing a single one, and that the economy is determinant only in the last instance. He then adds the observation that the lonely hour of the last instance never comes, which is either a devastating concession or a precise statement of what the theory claims, depending on who is reading.

E. P. Thompson’s reconstruction comes from the practice of writing history rather than from theory, and it is the most hostile to the whole framing. His argument, developed against Althusser, is that class is not a structure or a category but a historical relationship that happens when people, as a result of common experiences, articulate the identity of their interests against others, and that a model with no room for experience and agency cannot produce history that anyone would recognise. The theoretical claim that follows is that determination should be understood as the setting of limits and the exertion of pressures, a formulation Raymond Williams developed in parallel, rather than as a relation of cause to effect.

The analytical Marxists made the fourth attempt, and it is the most demanding technically. Their objection, associated above all with Jon Elster, is that any claim about societies must be reducible to claims about the actions of individuals and the mechanisms connecting them, and that a determination claim without such microfoundations is not an explanation but a correlation with a directional arrow drawn on it. G. A. Cohen’s defence of a strong version of the theory, which is the most rigorous statement any determinist reading has received, and Elster’s attack on it are treated in the case against historical materialism, which owns that exchange.

Where the evidence on each side is weak

Both cases have weak points, and naming them is what distinguishes a verdict from a preference.

The determinist case rests heavily on a preface, and prefaces compress. The 1859 passage is three pages summarising the results of years of study for a book on economics, written for a censor’s eye as well as a reader’s, and Marx never expanded it into the systematic statement it promises. Treating the most compressed page Marx ever wrote as the most authoritative one is a choice that requires defence, particularly when the historical writings, where the theory is actually applied, read so differently. The determinist case also has to explain why Marx spent the last decade of his life refusing to universalise his account, most explicitly in a letter drafted in 1877 to a Russian journal in which he denied that his sketch of the genesis of capitalism in western Europe was a historico-philosophical theory of the general path every people is fated to tread. That refusal is difficult to square with a theory of necessary stages.

The anti-determinist case has a weakness that is the mirror image. Its evidence is disproportionately drawn from letters, from prison notebooks, from posthumously published historical journalism, and from twentieth-century reconstructions, and its central move is to qualify the published statements by reference to material that is less formal, less considered, or not by Marx at all. That is legitimate historiography, but it is not obviously the reading Marx would have endorsed, and a critic who says the qualifications are a rescue operation is not making an absurd charge. The anti-determinist case also has to explain what remains of the scientific claim once every strong formulation has been softened, and this is the point where it is most vulnerable. If the theory says that material conditions set limits and exert pressures, then the question is what those limits forbid, and the answer is rarely supplied.

Does Marxism claim the economy determines everything?

No version defended by a serious Marxist since the 1890s claims this. Engels explicitly rejected it in 1890, calling the idea that the economic element is the only determining one a meaningless phrase. What the tradition claims varies from strong causal priority for production relations to the much weaker claim that material conditions set limits on what is possible.

The positions that try to dissolve the dispute

Three moves attempt to make the question go away, and each fails in an instructive manner.

The first says that determination and causation are different relations, so the whole dispute rests on a confusion. There is something to this. To determine a thing can mean to cause it, to fix its limits, to give it its definite form, or to settle which of several possibilities is realised, and German lends itself to the ambiguity as much as English does. Distinguishing these senses genuinely clarifies. But it does not dissolve the dispute, because the question simply reappears in the new vocabulary: if the economy fixes limits rather than causing outcomes, what are the limits, and what falls outside them?

The second says the whole framing is positivist and that a dialectical account rejects the separation of cause from effect on which the question depends. This move is common and it is close to useless. Stated carefully, the point is that base and superstructure are internally related rather than externally connected, so that neither exists in the form it does without the other, which is a serious metaphysical position with a real Hegelian pedigree. Stated loosely, as it usually is, it functions as a refusal to answer, and a theory that responds to a request for its content by objecting to the request is in trouble.

The third says the dispute is about the wrong object, because Marx offered a critique of political economy rather than a general theory of history, so asking what determines historical change misconstrues the project. This has textual support and considerable scholarly backing, and it is the strongest of the three. Its difficulty is that Marx and Engels both stated general historical claims repeatedly and in publications, and a reading that treats every such statement as a lapse is doing a great deal of work to protect a thesis.

The determination spectrum

Five positions can be distinguished, and almost every participant in this argument occupies one of them, usually without saying which. The table is the instrument this article exists to produce, and its most useful column is the fourth: what each position forbids. A position that forbids nothing is not a theory, whatever else it may be.

Position	Canonical holder	Characteristic formulation	What it predicts	What it forbids	Principal weakness
Mechanical determinism	Second International orthodoxy, Kautsky’s popular expositions and Plekhanov’s monist writings	Economic development produces political and ideological outcomes with law-like necessity	A determinate sequence of stages culminating in socialism in the most industrially advanced countries	Durable political outcomes that run against the direction of economic development; revolution in a backward economy	Refuted by the record: revolutions occurred in agrarian societies, advanced economies stabilised, and the sequence did not hold
Productive force primacy	G. A. Cohen’s defence of the theory	The relations of production obtain because they develop the productive forces; forces have explanatory primacy	Relations persist while they develop the forces and are replaced when they fetter them	Long-run persistence of relations that systematically block technical development	Functional form: it explains the cause by its effect without specifying a mechanism that selects for it
Last-instance determination	Althusser and the structuralist reading; the letters of Engels read strongly	Levels are relatively autonomous; the economy is determinant in the last instance	Structural limits on which political and ideological forms can be reproduced over time	Explanations of a durable institutional order by ideas alone with no account of material reproduction	The last instance is never specified operationally, which is the force of Althusser’s own remark about the lonely hour
Limits and pressures	Raymond Williams; E. P. Thompson’s practice; much current historical work	Material conditions set limits and exert pressures on what can be thought, built, and sustained	Systematic patterns in which arrangements survive and which do not	Nothing sharply; it forbids only arrangements that cannot be materially reproduced	Purchased safety: it is compatible with almost any historical outcome, which is the unfalsifiability charge in its strongest form
Bare material constraint	The version most critics attribute to Marxism when arguing that it is trivially true	Economic circumstances constrain human action	That people cannot do what they lack the means to do	Nothing at all	Not a distinctively Marxist claim, and not a theory

The spectrum should be read as a trade-off rather than a menu. Moving down the table buys immunity from refutation at the cost of content, and moving up buys content at the cost of exposure to the record. That trade-off is the actual structure of this dispute, and neither side usually acknowledges that they are choosing a point on it.

What is the difference between determination and conditioning?

Determination fixes an outcome or the range of possible outcomes; conditioning alters the probability, cost, or form of outcomes without fixing them. Marx’s published language uses both kinds of verb, which is why translation choices matter here more than in most passages, and why the dispute survives careful reading of the texts.

The forbidding test and the verdict

The verdict this article reaches is that the strong determinist readings are false, the weakest readings are empty, and a defensible position exists between them, but only if it is willing to say what it rules out. The test that locates it is simple to apply and almost never applied.

Ask of any determination claim: name a historical outcome this forbids. Not an outcome it finds surprising, not one it would explain differently, but one that, if it occurred and persisted, would count against the claim. If no answer is available, the claim is not doing explanatory work, whatever its rhetorical force. If an answer is available, the claim can be assessed against the record, which is what a theory is for.

The reason this test is decisive here is that the entire century-long qualification of historical materialism has proceeded by weakening the claim without ever restating what the weakened claim excludes. Reciprocal action was conceded. Relative autonomy was conceded. The independent effectivity of law, of the state, and of ideology was conceded. Each concession was made in response to a genuine difficulty and each improved the theory’s fit with the evidence. What was never done, except by Cohen and by a handful of others, was to say what the resulting theory still denies.

Three things the qualified theory does forbid

The exercise is worth performing rather than merely recommending, because performing it shows that the qualified theory is not empty, which is the conclusion most critics draw and most defenders fear.

The first prohibition concerns durable institutional orders without material reproduction. The theory forbids the explanation of a stable institutional arrangement by reference to ideas, values, or beliefs alone, with no account of how the arrangement is materially sustained, who bears its costs, and what interests are served by its persistence. This is a real constraint on historical explanation and it is violated constantly in popular history. An account of the persistence of caste, of slavery, of guild restriction, or of gendered divisions of labour that appeals only to what people believed, without asking who was extracting what from whom and how the extraction was secured, is ruled out by this theory. That such accounts are often wrong is not a coincidence; the prohibition has a good record.

The second prohibition concerns the long-run persistence of relations that systematically block the development of productive capacity in a competitive environment. The theory forbids the indefinite survival of an arrangement that leaves its bearers systematically less able to produce than their neighbours and rivals, where those neighbours can invade, undersell, or absorb them. The word competitive is doing necessary work: the claim is not that inefficient arrangements collapse, since many have lasted for centuries in isolation, but that they do not survive contact with more productive rivals indefinitely. This prohibition has a good record too, and it is the core of what is defensible in the primacy thesis without the functional apparatus.

The third prohibition concerns the autonomy of ideological change from material circumstance. The theory forbids an explanation of a large-scale, durable change in what a population believes and how it lives that makes no reference to changes in how that population produces, works, and reproduces itself. This one is weaker than the other two and is the most often violated by Marxists themselves in the direction of overreach, but as a prohibition it still bites: it rules out histories of ideas in which doctrines succeed doctrines with no reference to the world, which was precisely the target of the manuscripts of 1846 and which remains a recognisable genre.

Those three prohibitions are what a qualified historical materialism forbids. They are not nothing. They are also considerably less than the tradition has often claimed, and a reader who says that the theory has been reduced from a science of history to a set of methodological constraints on historical explanation has described the outcome accurately. Whether that reduction is a defeat depends on what one expected. Compared with the promise of a law-governed sequence of stages, it is a substantial retreat. Compared with the alternative frameworks on offer, it is a serviceable research programme with a real record.

What would change this verdict

Two things would. If a defender specified a mechanism connecting economic structure to political and ideological outcomes that was precise enough to generate predictions and general enough to apply across cases, the position could move back up the spectrum without losing its immunity to the standard refutations. Cohen’s functional account was an attempt at this and the mechanism problem defeated it, but the attempt is not obviously impossible in principle.

If, conversely, historical work established that the three prohibitions above are routinely violated in the record, that durable institutional orders regularly persist with no material basis, that systematically unproductive arrangements survive competitive contact indefinitely, and that large-scale ideological transformations occur with no relation to changes in production, then the qualified theory would join the strong one in refutation. Nothing in the current historical literature suggests this, which is why the verdict lands where it does.

Popper’s objection, stated accurately

The most serious version of the unfalsifiability charge is Karl Popper’s, and it is misstated more often than any other argument in this field, usually by the people invoking it.

Popper did not argue that Marxism was always unfalsifiable. He argued the opposite, and the distinction is the whole point. In The Poverty of Historicism and The Open Society and Its Enemies, and more compactly in the essay on science and pseudoscience later collected in Conjectures and Refutations, his claim is that Marx’s theory was originally scientific in the strict sense: it made risky predictions, chief among them that revolution would come first in the most industrially advanced countries, that the condition of the working class under capitalism would deteriorate, and that the middle strata would be squeezed into the two great classes. Those predictions were testable, and they were falsified by events. What happened next is what Popper objected to. Rather than accepting the refutation, followers of Marx reinterpreted both theory and evidence in order to make them agree, introducing auxiliary hypotheses that saved the framework at the cost of its testability. He called this a conventionalist twist, and his charge is that the procedure destroyed the theory’s scientific status, which it had genuinely possessed.

Two things follow. The charge is historically specific and is aimed at a practice rather than at a doctrine, which makes it much harder to answer than the crude version, since a defender cannot simply point to a testable claim in Marx; that is Popper’s own premise. And the charge is partly an accusation of bad faith, which means the reply has to address how the tradition handled disconfirmation rather than only what the tradition believed.

The best reply concedes the practice and disputes the inference. Auxiliary hypotheses that save a theory from refutation are not automatically illegitimate; every research programme in the sciences protects its core assumptions in this way, and Imre Lakatos built an entire account of scientific rationality on the observation that programmes are judged by whether their protective modifications generate new predictions or merely absorb anomalies. The uneven development argument, which explains revolution in Russia rather than Britain by the position of each within a world economy, is the test case. If it merely says that revolution happens where it happens, it is degenerate and Popper’s charge lands. If it generates the further expectation that revolutionary crisis will concentrate at the points where advanced and backward forms of production are most violently combined, that is a prediction with content and it can be assessed. Judgments differ on which it did, and the honest position is that the auxiliary-rescue charge lands against a considerable amount of twentieth-century Marxist practice and does not by itself establish that the framework is unfalsifiable in principle.

The full engagement with Popper’s attack on Marxist method works through the historicism argument, the essentialism argument, and the political conclusions Popper drew, and this article deliberately does not duplicate it.

Does economic determinism make Marxism impossible to test?

The strong versions were testable and several were tested and failed. The problem is at the qualified end: once reciprocal action and last-instance determination are conceded, no test is specified. The theory becomes testable again only when someone states what it forbids, which is the exercise most defences skip.

Weber and the standard counter-case

Max Weber’s study of the Protestant ethic is invoked in this dispute more than any other work, and the version invoked is usually a caricature that would have appalled him.

The popular account has Weber proving that religion caused capitalism, thereby refuting Marx by reversing the arrow. That is not the argument. Weber’s claim is narrower and more interesting: that a particular religious ethic, ascetic Protestantism in its Calvinist form, produced a distinctive attitude to work, to profit, and to consumption, and that this attitude was one causally relevant condition in the emergence of the specific rational, methodical, profit-seeking conduct that characterises modern capitalism. He is explicit that he is not explaining the emergence of capital, of markets, of trade, or of acquisitive behaviour, all of which he says existed in many times and places. He is explaining the spirit, meaning a particular disciplined orientation, and he treats it as one strand in a multi-causal account.

He is also explicit about the methodological point that matters most here. At the close of the study he warns against replacing a one-sided materialist causal interpretation of culture and history with an equally one-sided spiritualist one, and says that each is equally possible and each equally unhelpful as a conclusion rather than a preparation for investigation. That sentence is the reason Weber cannot be used as a straightforward anti-Marxist weapon. He is objecting to monocausal explanation as such, and his objection applies with the same force to any account that would derive the Protestant ethic from the interests of a rising bourgeoisie.

What Weber does establish, and it is substantial, is that a causal story can run from a religious development to an economic one without absurdity, that the direction of explanation is a question for evidence rather than for doctrine, and that ideas can have consequences their holders neither intended nor foresaw. His account of the transformation of a religious motivation into a secular occupational discipline, surviving after the belief that produced it has gone, is a mechanism, and the presence of a mechanism is exactly what most determination claims on either side lack.

The comparison of the two thinkers on the specific question of class is owned by the comparison of Marx and Weber on class, and this article does not restate it. What belongs here is only the warning that a writer who deploys the Protestant ethic against economic determinism is relying on a reading that Weber rejected in the text itself.

Does religion shape the economy or the other way round?

Both directions are documented, which is why the interesting question is not direction but mechanism. Weber traced a route from religious discipline to occupational conduct; Marxist historians have traced routes from changing property relations to changing religious practice. An account that names a mechanism beats one that only names a direction.

The translation problem at the heart of the dispute

A surprising amount of this argument turns on the choice of two German verbs and how they have been rendered into English, and a reader who does not know this will find the textual evidence more decisive than it is.

The 1859 preface uses bedingen and bestimmen in adjacent sentences. Bedingen carries the sense of conditioning, of supplying the conditions for, of determining in the way that a set of circumstances determines a range. Bestimmen is stronger, carrying the sense of fixing, settling, or determining in the way that a cause determines an effect or an authority determines a rule. The famous sentence about the mode of production of material life uses the conditioning verb; the sentence about social being and consciousness uses the stronger one. English translations have not been consistent, and several render both as determine, which flattens a distinction present in the original and makes the passage read as more uniformly deterministic than it is.

This is not a reason to conclude that Marx was a soft conditioner rather than a hard determinist. The stronger verb is there, in the sentence that matters most for the philosophy of mind and history, and the distinction between the two verbs in nineteenth-century German philosophical usage is not as sharp as the contrast in this paragraph suggests. What it is a reason for is caution. Any argument that rests on the exact force of determine in the English preface is resting on a translator’s choice, and a writer making such an argument should name the German and the rendering. That discipline is the standing rule for the whole corpus and the research guide to Marx’s works and the order to read them in sets it out in general.

The same problem recurs at the other end of the dispute. Althusser’s determination in the last instance renders a French phrase that itself renders a German one, and the chain of translation has smoothed a formulation that was already imprecise in Engels’s original. When a debate is conducted in a phrase that has passed through three languages, the participants are unlikely to be disagreeing about exactly the same thing.

Where the phrase economic determinism comes from

Neither Marx nor Engels used economic determinism as a description of their own position. The phrase entered circulation from outside the tradition, largely in the early twentieth century, and it has done two quite different jobs since.

Inside academic history and social science it named a methodological position, held by writers who were often not Marxists at all, that economic factors are the primary determinants of historical change. In the United States in particular, the economic interpretation of history had an independent life associated with progressive historiography, and its practitioners were frequently reformers with no interest in class struggle or revolution. Treating economic determinism as a synonym for Marxism erases this, and a historian who does so will misdescribe a large body of work.

Outside the academy the phrase is almost always pejorative, and it functions as an accusation of crudity: the charge that a writer has reduced culture, politics, or belief to money. Used this way it is rarely accompanied by any specification of which determination claim is being attributed, which is what makes it so easy to deploy and so hard to answer.

The consequence for a careful writer is that the phrase should either be defined at first use or avoided. Naming the position on the spectrum is more informative and much harder to dismiss. Writing that an analysis assumes productive force primacy, or that it treats ideology as having no autonomous effectivity, says something a reader can evaluate. Writing that it is economically determinist does not.

Which Marxists rejected economic determinism?

Most of the significant ones, and from early on. Engels rejected the strong version explicitly in his correspondence of the 1890s, and diagnosed the movement’s tendency to treat a formula as a theory. Gramsci made the rejection central, attacking what he considered the fatalism of the orthodox reading and building an account in which political and cultural work decides outcomes that economic conditions leave open. Lukacs’s early work made class consciousness the pivot on which historical possibility turns, which is incompatible with any mechanical reading. Althusser rejected the expressive model in which every level of a society straightforwardly expresses its economic essence, and replaced it with relative autonomy and overdetermination. Thompson, Williams, and the historians around them rejected it from the side of historical practice. What unites these is not a shared alternative, since they disagree with one another profoundly, but a shared judgment that the strong reading is both textually and empirically indefensible. The tradition’s internal history is largely a history of arguing about how much determination to retain.

Can ideas change the economy?

The question is asked constantly and is usually answered badly by both sides, so it is worth answering carefully.

Nothing in any defensible version of historical materialism denies that ideas have effects. The theory is not a claim about causal traffic in one direction; if it were, it would be refuted by any instance of a belief changing behaviour, which is to say instantly. What the theory claims is something about explanatory priority and about the conditions under which ideas become historically effective.

The claim has three parts, and separating them prevents most of the confusion. Ideas act, but they act through people whose situations dispose them to find some ideas compelling and others absurd, so the question of why a particular idea took hold in a particular period cannot be answered by the idea’s merits alone. Ideas that require material conditions that do not exist do not become effective however widely they are held, which is why proposals for the reorganisation of society have littered every century and only a few have gone anywhere. And ideas that serve the reproduction of a dominant set of relations have institutional support that ideas threatening those relations do not, so the competition between ideas is not conducted on level ground.

None of that denies causal power to ideas; it specifies the conditions of that power. The version that would be denied by the theory is the one in which an idea sweeps a society because it is persuasive, independently of who is positioned to spread it and what it costs to act on it. That version is not held by many people once it is stated plainly, which is a sign that the disagreement here is narrower than the volume of argument suggests.

The interesting cases are the ones where an idea appears to have altered the organisation of production directly. Technical innovation is the obvious family, and it is not a problem for the theory: knowledge is one of the productive forces on any standard account, so a new technique changing production is the theory working rather than a counter-example. Religious transformation is the harder family, and Weber’s is the best-worked case. Legal and constitutional innovation is harder still, since a change in the form of property rights alters the relations of production themselves, which is why the status of law in the base has been argued about for a century. The treatment of base and superstructure works through the law problem, which is the sharpest boundary case in the whole model.

Does Marxism deny free will?

The determinism at issue in this debate is historical rather than metaphysical, and running the two together produces confusion that neither philosophy nor politics can clear up afterwards.

A metaphysical determinist holds that every event, including every human decision, is fixed by prior states of the world. A historical determinist holds that the large-scale course of history follows a determinate path. These are independent positions. One can hold the first and deny the second, since a world in which every event is fixed need not exhibit any large-scale pattern that a historian could state. One can hold the second and deny the first, since a pattern in the aggregate is compatible with indeterminacy in the parts, as every statistical regularity in a population of free agents demonstrates.

Marx’s own commitments on the metaphysical question are not clearly stated anywhere and were not a concern of his. The historical claim is what he argued about. When the Eighteenth Brumaire says that people make their own history under circumstances they do not choose, it is making a claim about the relation between agency and constraint at the historical level, not entering a debate about the causal closure of the physical world.

The practical upshot for a writer is that the objection “Marxism denies free will” is a category error in the form usually stated. The serious version of the objection is different and is worth taking: if the outcome is determined by economic development, then political and moral deliberation about what to do is pointless, and a movement that believes this will not act. That objection is about the practical incoherence of fatalism rather than about metaphysics, it was made forcefully inside the socialist movement by people watching the German party wait, and it is one of the reasons the determinist reading lost its grip after 1914.

What the historical record does to the strong reading

A debate about determination has to face the record, and the record is where the strong reading breaks.

Four patterns are decisive. Revolutions carried out in the name of the theory occurred in agrarian economies at the periphery of the world market rather than in the most industrially advanced countries, which is the reverse of what the strong reading predicted and what Marx himself expected for most of his life. The condition of the working class in the industrial core improved over the long run rather than deteriorating towards subsistence, though the measurement disputes here are genuine and the timing and distribution of the improvement are contested. Class identity proved weaker than national and religious identity at several decisive moments, most starkly in the summer of 1914 when the socialist parties of Europe voted for war credits against their own international commitments. And the postwar settlement in western Europe produced a durable class compromise, with rising wages, expanded welfare provision, and stable parliamentary politics, that no version of the strong reading anticipated.

Each of these has a reply and the replies are not worthless. Uneven and combined development explains why the chain broke at its weakest link rather than its strongest, and it was formulated before the fact rather than after. The improvement in working-class conditions can be located in the returns from imperial position and from productivity growth, and its distribution across the world economy is a different picture from its distribution within one country. The failure of internationalism in 1914 was a political failure with identifiable causes, including the integration of party and trade union leaderships into national institutions. The postwar compromise was conditional on a growth rate and a balance of forces that later shifted.

What the replies establish is that the record does not refute historical materialism as a research programme. What they cannot establish is that the strong reading survives, because each reply works by introducing a factor that the strong reading said was not decisive. Once national institutions, political leaderships, imperial position, and the balance of organised forces are doing the explanatory work, the position occupied is somewhere in the middle of the spectrum, not at the top of it. The case against historical materialism sets out these objections at full strength with the best available replies to each, and it owns that survey.

Why Marx’s late refusal matters more than it is usually allowed to

In 1877 Marx drafted a reply to a Russian critic who had presented the account of primitive accumulation in Capital as a universal scheme of development that Russia would have to undergo. Marx denied it flatly, saying that his historical sketch of the genesis of capitalism in western Europe was being transformed into a historico-philosophical theory of a general path prescribed by fate for all peoples, and that this did him too much honour and too much discomfort. He returned to the point in drafts of a letter to Vera Zasulich in 1881, where he considered whether the Russian village commune might provide a route to a different development.

The determinist reading has to treat these as either eccentric or unrepresentative, and neither is easy. They are late, they are considered, and they concern precisely the question at issue. What they establish is that Marx did not hold the theory of necessary universal stages that his followers attributed to him within a decade of his death, and that the strongest version of the determinist reading is not merely undersupported by the texts but explicitly repudiated in them.

What they do not establish is the weak reading, and this is where anti-determinist writers overreach. Denying that western European development is a universal fate is compatible with holding a strong theory about what determines development within a society once capitalist relations are established. Marx’s refusal is about the applicability of a sequence, not about the strength of determination.

How the argument shaped socialist politics

The dispute’s political history is the part most philosophical treatments omit, and it explains why the positions hardened as they did.

Eduard Bernstein’s revision at the end of the 1890s attacked the determinist reading from the right, arguing that capitalism was not producing the crises, the concentration, and the polarisation the theory predicted, that the working class was not becoming uniformly immiserated, and that socialism would therefore have to be argued for on ethical grounds and achieved through reform. His conclusion, that the movement was everything and the final goal nothing, was a repudiation of the determinist framework rather than a modification of it.

Rosa Luxemburg’s reply attacked Bernstein while refusing the fatalist reading he was attacking. Her argument was that the objective development produces the conditions of crisis and the possibility of revolution but not its accomplishment, that the working class must be capable of taking power when the opening comes, and that a party which treats the outcome as guaranteed will not develop that capacity. Her formulation of the alternative facing society, between a transition to socialism and a regression into barbarism, is the sharpest statement in the tradition that the outcome is not settled, and it comes from someone who was defending orthodoxy against revision.

Lenin’s contribution to the same argument was practical and is often misread as voluntarist. His insistence that political consciousness does not arise spontaneously from economic struggle, and that a revolutionary organisation is necessary to produce it, is an anti-determinist move regardless of what else one thinks of the conclusions he drew from it. The economists he was attacking in that argument held a position that was determinist in exactly the sense this article has been describing: that the economic struggle would generate political conclusions of its own accord.

What all three had in common, across enormous political differences, was the recognition that the strong reading was strategically disabling. That practical judgment did more to end the determinist consensus than any philosophical argument, and it is worth remembering when the dispute is presented as a purely textual one.

Is anyone still an economic determinist today?

Very few argue for the strong version by name, and almost nobody defends mechanical determinism. Productive force primacy retains serious defenders in the analytical tradition, and versions of last-instance determination remain common in structural analyses of politics and culture. The commoner phenomenon is unacknowledged determinism, where an analysis assumes that economic interests explain political outcomes without stating the assumption or defending it, and that habit is at least as widespread outside Marxism as within it.

Five misreadings that keep this dispute confused

Each of these is common enough to derail an otherwise sound argument, and each has a short correction.

Marx never used economic determinism approvingly, and neither did Engels. The phrase is not part of the vocabulary either man wrote in, and attributing it to them as a self-description is a factual error rather than an interpretative one. Writers who need a phrase from inside the tradition should use the materialist conception of history, which is what Engels called it.

Engels’s letters of the 1890s are not a late retreat. They are a restatement made in response to correspondents who asked how far the theory extends, and in them Engels says that the vulgar version was never the position, while conceding that polemical emphasis had encouraged it. Treating the letters as a deathbed climbdown is a rhetorical move that both critics and orthodox defenders have found convenient, and it does not survive reading them.

The Eighteenth Brumaire and the 1859 preface are not in flat contradiction, though they are in tension. The preface describes the structural conditions under which historical change becomes possible; the historical writings describe how particular outcomes were produced within those conditions. A theory can hold both, and the interesting work lies in specifying how tightly the first constrains the second, which is what the spectrum in this article is for.

The alternative to determinism is not idealism. This is the most common false dichotomy in student writing on the topic. Rejecting the claim that economic structure fixes historical outcomes does not commit anyone to the claim that ideas do, and the space between those positions contains almost every serious account of historical causation, Marxist and otherwise. The relation between the materialist and idealist positions properly understood belongs to the comparison of historical materialism with Hegelian idealism, which owns that distinction.

Determination and causation are not the same relation, as the earlier discussion set out, and treating them as interchangeable makes the textual evidence look more one-sided than it is. Where an argument turns on this, name which relation is meant.

How to place a writer on the spectrum

The method is worth stating because it converts this article from a survey into an instrument, and because it works on writers who never announce their position.

Ask first what the writer treats as needing explanation and what they treat as given. A writer who explains political outcomes by economic interests and never asks what explains the interests has placed the economic level outside the analysis, which is a determinist move regardless of how many qualifications appear elsewhere.

Ask second what happens in the writer’s account when the economic conditions are held constant. If the account permits substantially different political and ideological outcomes on the same economic base, the writer is at or below last-instance determination on the spectrum. If it does not, the writer is above it.

Ask third whether the writer has a mechanism. Naming a relation between an economic structure and a political outcome is not the same as showing how the one produces the other, and the presence of a mechanism, through interests, through institutional capacity, through selection over time, or through the reproduction requirements of a system, is what separates an explanation from a correlation.

Ask fourth what the writer says would count against their account. This is the forbidding test applied to a person rather than a proposition, and it is the fastest way to distinguish a serious analyst from someone using the framework as a vocabulary.

The method is not confined to Marxists. It applies with equal force to rational choice accounts that treat preferences as given, to cultural accounts that treat values as self-explaining, and to institutionalist accounts that treat institutions as prior to the interests they organise. Every framework has something it declines to explain, and finding it is the first move in evaluating any of them.

The dispute outside Marxism

Marxists did not invent this question and do not own it, and seeing the parallels helps a reader avoid the impression that the difficulty is peculiar to one tradition.

Economic historians argue about whether institutions or geography or factor endowments explain long-run divergence in national income, and the structure of that argument, with each side accusing the other of treating its favoured factor as exogenous, is recognisably the same structure as this one. Sociologists argue about the relation between social structure and agency, and the vocabulary of structuration was developed precisely to avoid a choice between a structure that determines and an agency that floats free. Historians of ideas argue about whether the context or the argument explains the emergence of a doctrine. In international relations, materialist and constructivist accounts divide over whether the distribution of capabilities or the distribution of ideas explains state behaviour.

The parallel is not decorative. It shows that the difficulty is general to the explanation of social phenomena rather than a defect in one theory, and it means that solutions developed elsewhere may transfer. The most useful transfer is the discipline of specifying mechanisms, which the social sciences have generally moved towards over the past several decades and which is exactly what the qualified versions of historical materialism most conspicuously lack.

Two cases where the test does real work

Abstract discussion of determination becomes useful when applied to a dispute historians actually have, and two disputes show the forbidding test earning its keep.

The transition to capitalism

The question of why capitalist relations emerged in some parts of Europe and not others is the oldest live argument in Marxist historiography, and it is a dispute between people who all accept the framework, which makes it a clean test of what the framework settles.

One family of explanations gives priority to the growth of trade, the expansion of markets, and the dissolving effect of commerce on feudal obligation. On this account the development of exchange erodes the manorial economy, and the emergence of capitalist agriculture and industry follows. A second family denies that trade dissolves anything by itself, pointing out that expanding commerce in eastern Europe produced an intensification of serfdom rather than its collapse, and locates the decisive variable in the outcome of struggles over rent and tenure between lords and peasants, which produced in one region a class of capitalist tenant farmers and in another a peasantry secure in its holdings and a nobility dependent on state office.

Notice what has happened. The same framework has generated two accounts that predict different outcomes from the same commercial development, and the second explicitly denies that the economic conditions settled the result, making it turn instead on the balance of forces in a conflict whose outcome was not given in advance. That is an internal argument about where on the spectrum the theory should sit, conducted by historians using archival evidence, and it is settled the way such arguments should be settled: by asking which account predicts the divergence between eastern and western Europe, and the second does while the first does not. The case against historical materialism treats this internal critique in more detail, since it is one of the strongest objections to productive force primacy and it comes from inside the tradition.

The welfare state

The second case is nearer in time and is instructive because the determinist and anti-determinist readings make different predictions that can be checked.

A strong reading explains the expansion of social provision in the industrial countries as a functional requirement of capital: an educated, healthy, and reproducible workforce is needed, so the state supplies what individual employers will not, and the resulting institutions serve accumulation. A weaker reading explains the same expansion by the strength of organised labour and the political concessions extracted from employers and states under pressure, with the specific form of provision reflecting the balance of forces at the moment of settlement.

The two accounts diverge sharply on what to expect. The functional account predicts that provision should track the requirements of production and should be similar in countries at similar levels of development. The balance-of-forces account predicts that provision should track union density, party strength, and the timing of enfranchisement, and should differ substantially between countries at similar levels of development. The comparative record, in which countries of similar productivity built systems that differ enormously in coverage, generosity, and administration, is difficult for the functional account and comfortable for the other. This is what the forbidding test looks like when applied rather than described: the strong version made a prediction, the prediction can be checked, and the checking pushes the defensible position down the spectrum.

Reading the preface passage closely

Since the passage carries so much of the determinist case, it repays a slower reading than it usually gets, and four features of it are routinely missed.

It is written in the past tense of an intellectual autobiography. Marx introduces it as the general conclusion at which he arrived and which then served as the guiding thread of his studies. He is describing a research heuristic that organised his work, not announcing a law he has demonstrated. That framing does not make the claims weaker, but it changes what kind of claim they are, and readers who treat the passage as the statement of a proven theory are reading against its own presentation.

It distinguishes two things that most summaries merge. The economic transformation, Marx says, can be determined with the precision of natural science; the legal, political, religious, artistic, and philosophical forms in which people become conscious of the conflict and fight it out cannot. The second half of that sentence concedes a great deal. It says that the ideological level is not amenable to the same kind of determination as the economic one, and it describes that level as the medium in which conflict is fought out, which requires it to be active rather than reflective.

It contains a clause about the conditions of change that is stronger than anything else in the passage and is almost never quoted. No social order perishes, Marx says, before all the productive forces for which it is sufficient have developed, and new relations never appear before the material conditions of their existence have matured within the old society. This is the passage that generates the theory’s actual prohibitions and it is the one a defender should reach for, since it names a condition that could in principle be shown not to hold.

It ends by naming four epochs in a single sentence and by saying that the bourgeois form is the last antagonistic one. Everything after that sentence is a claim about the future rather than a summary of research, and it belongs to a different evidential category from what precedes it.

The everyday version of the argument

The dispute has a popular form that most readers meet long before they meet the texts, and recognising it is part of understanding the debate.

The everyday determinist claim is that some conflict or policy is really about money, and that the stated reasons are cover. It appears in commentary on wars, on legislation, on institutional decisions of every kind, and it is made by people across the political spectrum who would be surprised to find themselves grouped with Marx. Its attraction is that it is often partly right: interests do shape conduct, stated reasons are often incomplete, and the question of who benefits is always worth asking.

Its weakness is the one this article has been tracing. In the form usually stated it forbids nothing, because any outcome can be redescribed as serving somebody’s interest after the event, and the redescription is unfalsifiable in exactly the way Popper described. It also tends to conflate two very different claims: that an outcome serves an interest, which is often demonstrable, and that it occurred because it served that interest, which requires a mechanism showing how the interest was translated into the decision. The gap between the two is where most bad political analysis lives.

The disciplined version of the same argument is much stronger and is worth learning. It identifies the interest, identifies the channel through which the interest was brought to bear, whether through direct representation, financial dependence, ideological formation, institutional selection, or structural constraint on what any government can do, and states what would have been observed if the interest had not been decisive. That version can be checked. It is also, as it happens, what the more careful Marxist work on the state and on policy has always tried to supply, which is why the accusation that such work is crudely determinist so often mistakes the disciplined version for the everyday one.

The reverse error is equally common and less remarked. An analysis that treats stated reasons as the whole explanation, and refuses on principle to ask who benefits, is not neutral; it has simply chosen a different thing to leave unexplained. The four questions set out earlier apply to it with the same force.

Why quotation cannot settle it

One reason this argument has lasted so long is that both sides can win the quotation contest, and neither side notices that winning it proves nothing.

The corpus is enormous, written across four decades, in registers ranging from the compressed summary to the newspaper article to the private letter to the unfinished draft, and by two authors whose collaboration makes attribution uncertain in places. A body of writing that size, produced under those conditions, will contain formulations pointing in several directions, and it would be remarkable if it did not. Assembling a list of strong statements proves that strong statements exist; assembling a list of qualifications proves that qualifications exist. Neither list establishes what the theory claims, because the theory is not the sum of its author’s sentences.

What can settle it is the reconstruction of a position that makes the best sense of the corpus as a whole, states its claims precisely enough to be evaluated, and is then assessed against the historical record. That is a harder exercise than quotation and it is the exercise this article has attempted. It also has an uncomfortable implication that both camps resist: the resulting position is a reconstruction, and the question of whether Marx held it is separable from the question of whether it is true.

Readers who find that separation unsatisfying should notice that it is standard practice in the study of any large body of thought. The strongest defensible version of a position is worth stating whether or not its supposed author would have recognised it, and the historical question about what Marx believed is a question for intellectual history rather than for social theory. Confusing the two is the reason so many arguments about determinism end with both sides producing more quotations.

A final observation about the shape of the argument. Every position on the spectrum was arrived at by someone trying to solve a problem that the position above it had failed to solve, which means the sequence is a record of learning rather than a menu of preferences. Mechanical determinism failed against the record of 1914 and after. Productive force primacy was constructed to give the theory a defensible core and ran into the mechanism problem. Last-instance determination was constructed to preserve the core while conceding the autonomy of other levels and ran into the difficulty of specifying the last instance. Limits and pressures was constructed to describe what working historians actually do and ran into emptiness. Nobody chose these positions out of temperament, and nobody arrived at one without first watching a stronger version fail against something it had been confident about. That is the ordinary shape of a research programme under pressure, and it is not by itself evidence of decline. Reading the spectrum as a history of successive repairs, each buying something and costing something, is more accurate than reading it as a range of opinions, and it also explains why the dispute has not converged: each repair solves the problem it was designed for and creates a new one.

Studying this dispute for an examination

Every serious examination question on historical materialism has an evaluation component, and this dispute is where that component is won or lost. The mark-earning move is not to declare that Marx was or was not a determinist, which is what most candidates do, but to distinguish the versions and then answer the question about a specified one.

The single distinction that earns marks is between the strong reading and the qualified reading, stated with evidence for each. A candidate who writes that the 1859 preface supports a strong reading while the historical writings and Engels’s correspondence support a qualified one, and who then says which reading the question is about, has demonstrated command of the material rather than a position on it. Adding the forbidding test, by asking what the qualified reading still rules out, lifts an answer further, because it converts an evaluation into an argument.

The standard trap is the assumption that showing Marx was not a crude determinist amounts to defending the theory. It does not, and a marker will notice. Once the theory has been qualified, the burden shifts to saying what it still claims, and an answer that stops at the qualification has left the theory undefended rather than defended. The mirror trap is the answer that attacks the strong reading and treats the whole tradition as refuted, which ignores a century of work that abandoned that reading before the candidate was born.

Two pieces of evidence are worth carrying in exact form: the conditioning language of the 1859 preface and the ultimately-determining formulation from Engels’s letter to Bloch. The answer structures in which these are deployed belong to the exam answer guide for historical materialism, which this article links to rather than duplicating.

Running this as a seminar

This dispute produces the best seminar of the term if the motion is set correctly and the worst if it is not. The motion that fails is “Was Marx an economic determinist?”, which produces two hours of textual ping-pong and no conclusion. The motion that works is “Name one thing historical materialism forbids”, and it works because it is answerable, because the answers can be assessed, and because students discover the trade-off between content and safety by walking into it themselves.

The misconception that dominates such a class is that determinism and free will are the question. Students arrive with the metaphysical debate in mind, having met it in philosophy or in a religious context, and they will run the historical question into it within ten minutes if not prevented. The question that surfaces this productively is to ask whether a statistical regularity in a population, such as the reliability with which a proportion of a cohort will do something, tells you anything about whether any individual member had a choice. Once the class sees that aggregate patterns and individual freedom are compatible, the historical question can be asked in its own right.

The extract that resolves the substantive issue best is the pairing of the opening paragraphs of the Eighteenth Brumaire with the central paragraph of the 1859 preface, read aloud in that order. Students who have been told that Marx was a determinist find the first text surprising, and students who have been told he was not find the second text surprising, and the discussion starts from genuine puzzlement rather than from positions. For a politically mixed class this is unusually safe ground, because the question is about what a theory claims rather than about whether it is admirable.

What a writer must verify before using the word

A journalist or analyst reaching for economic determinism under deadline should check three things.

Check whether the target actually holds the position. The label is applied to any analysis that gives weight to material interests, which is most serious analysis, and using it that way asserts nothing. If the target’s account permits different political outcomes on the same economic conditions, the label is wrong.

Check the attribution before quoting. The claim that Marx said the economy determines everything is not supported by any passage in his published work, and the most quotable evidence on the other side, the Engels letter, says the opposite in terms. The unsafe claim in general circulation is that Engels retracted the theory in old age; he restated it and rejected a caricature of it, which is a different event.

Check whether the argument needs the label at all. Naming the specific claim, that a writer treats interests as given, or that they assume institutional forms follow from ownership patterns, is more precise, more defensible, and harder to answer with a shrug. Building a set of citation cards that pairs each formulation with its source, its date, and its status as published work, letter, or draft is the practical defence against the errors above, and you can save your notes and build a citation-linked reading list free on VaultBook to keep them in one place.

What remains open

Three questions in this dispute are genuinely unsettled and are worth stating as such.

The mechanism question is open. No version of the theory has specified a general mechanism connecting economic structure to political and ideological outcomes that is precise enough to generate predictions and general enough to apply across cases. Interests, institutional capacity, selection, and functional requirement have all been proposed and each works in some cases and fails in others.

The scope question is open. Whether the theory is meant to apply to all societies, to class societies, to societies in contact with capitalist ones, or only to capitalism itself has never been settled, and Marx’s own late refusals point one way while his general formulations point another.

The strength question, which is this article’s subject, is open in the sense that no position on the spectrum commands agreement, but it is not open in the sense of being undecidable. The strong readings are refuted by the record. The weakest are empty. The defensible positions lie between, and they earn their status only by saying what they forbid. That is the state of the argument, and a reader who can state it in those terms is ahead of most of the writing on the subject.

Determination inside the economics

The dispute is usually conducted over the theory of history, but Marx’s economic writing contains its own version of it, and the version there is instructive because it is more careful.

Capital is organised around what Marx called the laws of motion of modern society, and the phrase invites a determinist reading. What the text delivers is more qualified than the phrase suggests. The most famous of these laws, the tendency of the rate of profit to fall, is presented in the third volume together with a set of counteracting influences in the same part of the work, including the intensification of exploitation, the depression of wages below the value of labour power, the cheapening of the elements of constant capital, relative overpopulation, foreign trade, and the increase of share capital. The tendency and the counter-tendencies are stated as a single structure, which means that the theory does not predict a steady decline in profitability and never did. It predicts a systematic pressure in one direction that other systematic pressures oppose, with the outcome in any period depending on the balance.

That structure is a model of how a determination claim can be made without becoming either mechanical or empty. It names a mechanism, it names what opposes the mechanism, and it therefore identifies the conditions under which the predicted outcome would and would not appear. A great deal of the confusion in the theory of history would have been avoided if its claims had been stated in the same form.

It also supplies the answer to a common objection. When critics observe that profit rates have not declined steadily, they are refuting a claim the text does not make. When defenders observe that the theory is safe from such refutation, they are close to conceding that it forbids nothing, which is the trap this article has been describing throughout. The honest position is that the tendency claim is testable, that testing it requires measuring the tendency and the counter-tendencies separately, and that the measurement choices drive the result. The empirical literature on this is treated in the cluster on crisis theory, and the pattern it displays, in which the substantive disagreement turns out to be a disagreement about measurement rather than about theory, recurs across the series.

Technological determinism as a separate charge

A distinct accusation is frequently folded into this dispute and should be separated from it, because the evidence and the replies are different.

Technological determinism holds that the development of productive technique drives social change, with social relations adapting to the requirements of the technology in use. The charge against Marx rests on a small number of passages, above all the remark in the polemic against Proudhon that the hand mill gives you society with the feudal lord and the steam mill society with the industrial capitalist, and on the general priority given to the productive forces in the 1859 preface and in Cohen’s reconstruction.

Three replies are available and they have different strengths. The weakest is that the hand mill remark is a rhetorical flourish in a polemic; it is, but it is also consistent with other passages and cannot be dismissed. The strongest is that the productive forces on Marx’s account include labour power, skill, and organisation, not merely instruments, so that the priority of the forces is not the priority of machinery, and an account in which the development of human productive capacity drives change is not technological determinism in the ordinary sense. The most interesting reply is historical: the detailed analysis in Capital of the transition from manufacture to machinofacture makes the introduction of specific machines depend on the state of the class struggle, on the length of the working day, and on the resistance of skilled workers, which reverses the causal arrow in exactly the places where a technological determinist would not expect it.

The charge is therefore weaker than it looks, but it is not baseless, and a writer who wants to use the priority of the productive forces should be clear about whether they mean technique or capacity. The distinction is set out further in the guide to the mode of production, which owns the forces and relations vocabulary.

The hardest case: law

If one wanted to test the determination claim on a single institution, law is the place to do it, and the case is instructive because it is genuinely difficult for the theory.

The difficulty is structural rather than empirical. Relations of production, on any standard account, are relations of effective control over productive resources and over labour, and in a developed society those relations are constituted by legal categories: ownership, contract, employment, corporate personality, and the rest. If the base is constituted by law, then law cannot straightforwardly be part of a superstructure that the base explains, because the explanation would be circular.

The two standard solutions each cost something. The first, associated with Cohen, describes relations of production in terms of powers rather than rights, so that the base can be specified without legal vocabulary and law can then be explained as the form in which those powers are stabilised and enforced. This works logically and is the most rigorous solution available, at the price of a rather artificial description of ordinary social relations, since nobody actually experiences effective control over a factory in abstraction from the legal title to it. The second denies the sharp separation, treating law as internal to the relations of production rather than derived from them, which is closer to how a working historian would describe an eighteenth-century property regime, at the price of weakening the model considerably.

The case matters here because it shows what a determination claim has to survive. Any position on the spectrum must say something about how a system of rules that constitutes economic relations can also be explained by them, and positions that have not addressed law have not been tested at their weakest point. The treatment of base and superstructure works through the boundary questions in detail.

Frequently Asked Questions

Q: Was Marx an economic determinist?

Not in the strong sense the label carries, and he never used the phrase to describe himself. The evidence pulls both ways and an honest answer holds both halves. The preface of 1859 states that relations of production are entered into independently of will, that a legal and political superstructure arises on the economic structure, and that social being determines consciousness, which is strong language. The Eighteenth Brumaire states that people make their own history under circumstances they did not choose, and the analysis that follows explains an outcome by the specific political conduct of specific factions. Late in life Marx explicitly refused to convert his account of western European development into a universal path prescribed for all peoples. The defensible conclusion is that he held a strong theory about the conditions under which historical change becomes possible and a much weaker one about which of the available outcomes occurs, and that he never reconciled the two systematically.

Q: Does Marxism claim the economy determines everything?

No version defended by a serious Marxist since the 1890s claims this, and Engels rejected it explicitly in his correspondence, saying that anyone who twists the materialist conception into the claim that the economic element is the only determining one turns it into a meaningless abstraction. What the tradition does claim varies widely. At the strong end it holds that relations of production obtain because they develop the productive forces, giving the economic level explanatory priority over every other. In the middle it holds that political, legal, and ideological levels have their own effectivity while the economic is determinant in the last instance. At the weak end it holds only that material conditions set limits and exert pressures. The everything claim is a position nobody occupies, which is why arguments against it never land. Naming the specific claim under attack is the only way to make the criticism connect.

Q: What did Engels say about economic determinism?

He rejected the vulgar version in four letters written between 1890 and 1894, and diagnosed how it had arisen. To Joseph Bloch he wrote that the ultimately determining element in history is the production and reproduction of real life, and immediately added that neither he nor Marx had asserted more, and that political forms, legal forms, and inherited traditions exert their own influence and often preponderate in determining the form of historical struggles. To Conrad Schmidt he argued that once law or the state separates out as a domain it acquires its own internal logic and its own personnel with their own interests. To Franz Mehring he described ideologists working with inherited thought material while unaware of their real motives. In the letter of January 1894 he wrote that everything reacts on everything else, and that the economic position is not the sole active cause. He also admitted that polemical emphasis had encouraged the misunderstanding.

Q: What does determination in the last instance mean?

It is the formulation that concedes real independence to political, legal, and ideological levels while retaining a claim of economic priority somewhere in the analysis. Its ancestor is Engels’s phrase about the ultimately determining element; its best-known form is Louis Althusser’s, who paired it with the idea that a social formation is a complex whole whose levels are relatively autonomous and whose events are overdetermined by contradictions at several levels rather than expressing one. The difficulty is that nobody has specified when the last instance arrives or what observation would show it operating, which is the force of Althusser’s own remark that the lonely hour of the last instance never comes. Read generously, the phrase names a structural constraint: over the long run, arrangements that cannot be materially reproduced do not persist. Read strictly, it postpones the determination claim to a moment that never arrives and therefore never has to be defended.

Q: Does economic determinism make Marxism impossible to test?

The strong versions were testable and several failed. Revolution did not come first in the most advanced economies, the condition of the working class in the industrial core improved over the long run, and class identity did not override national identity at the decisive moments. The problem is at the qualified end. Once reciprocal action, relative autonomy, and last-instance determination have been conceded, most statements of the theory specify nothing that would count against them, and Karl Popper’s charge was precisely that the qualifications were introduced to absorb refutations rather than to generate new predictions. The theory becomes testable again only when someone states what it forbids. Three prohibitions can be defended: durable institutional orders cannot be explained by ideas alone with no account of material reproduction; relations that systematically block productive development do not survive competitive contact indefinitely; and large-scale ideological transformation is not autonomous from changes in production.

Q: Can ideas cause economic change in Marxist theory?

Yes, and no defensible version of the theory denies it. What the theory claims is about the conditions under which ideas become historically effective rather than about the direction of causal traffic. Three claims do the work. Ideas act through people whose situations dispose them to find some ideas compelling and others absurd, so the spread of an idea is not explained by its merits alone. Ideas requiring material conditions that do not exist do not become effective however widely they are believed, which is why most proposals for reorganising society have gone nowhere. And ideas serving the reproduction of dominant relations enjoy institutional support that threatening ideas do not, so the competition is not conducted on level ground. Technical knowledge is the clearest case of ideas changing production, and it presents no difficulty for the theory, since knowledge counts among the productive forces on any standard account.

Q: What is the difference between determination and conditioning?

Determination fixes an outcome, or fixes the range within which outcomes must fall. Conditioning alters the probability, the cost, or the form of outcomes without settling which occurs. The distinction is not merely verbal, and a great deal of the textual argument turns on it, because the German of the 1859 preface uses a conditioning verb in one celebrated sentence and a stronger verb in the next, while several English translations render both as determine. That flattening makes the passage read as more uniformly deterministic than the original. The practical consequence is that any argument resting on the exact force of the English word determine in that text is resting on a translator’s choice, and a careful writer names the German and the rendering. It also means that showing Marx used conditioning language somewhere does not settle the dispute, since the stronger verb appears in the same paragraph.

Q: Which Marxists rejected economic determinism?

Most of the significant ones, and from early on, though they rejected it in incompatible directions. Engels rejected the strong version by name in the 1890s. Rosa Luxemburg attacked fatalism while defending orthodoxy against Bernstein, insisting that objective development produces the possibility of revolution but not its accomplishment. Antonio Gramsci made the rejection central, replacing automatic economic causation with an account of political and cultural work whose success is not guaranteed. Georg Lukacs made class consciousness the pivot on which historical possibility turns. Louis Althusser rejected the model in which each level expresses an economic essence, substituting relative autonomy and overdetermination. E. P. Thompson and Raymond Williams rejected it from the side of historical practice, recasting determination as the setting of limits and the exertion of pressures. What unites them is not an agreed alternative but a shared judgment that the strong reading is indefensible.

Q: Does religion shape the economy or the other way round?

Both directions are documented, and the more useful question is which account names a mechanism. Max Weber traced a route from ascetic Protestant discipline to a methodical orientation towards work and profit, and crucially described how a religious motivation could be transformed into a secular occupational discipline that outlived the belief producing it. That is a mechanism, which is why the study has lasted. Marxist historians have traced routes in the other direction, explaining changes in religious practice and organisation by changes in property relations, in urbanisation, and in the position of clerical institutions. Neither direction is ruled out by any defensible version of historical materialism. What Weber himself insisted on, at the close of his study, was that replacing a one-sided materialist interpretation with an equally one-sided spiritual one advances nothing, and that both are preparations for investigation rather than conclusions.

Q: Is anyone still an economic determinist today?

Very few argue for the strong version under that name, and mechanical determinism has effectively no defenders. Productive force primacy retains serious support in the analytical tradition descending from Cohen’s defence, and structural analyses of politics and culture continue to work with versions of last-instance determination. The commoner phenomenon is unacknowledged determinism, in which an analysis explains political outcomes by economic interests without stating the assumption, defending it, or asking what explains the interests. That habit is at least as widespread outside Marxism as within it, appearing in rational choice accounts that treat preferences as given and in policy analysis that treats economic incentives as the only real motives. Naming the assumption is more useful than naming the tradition, since the label attaches to a great deal of work whose authors would reject the theory entirely.

Q: Does Marxism say communism is inevitable?

Some Marxists said so and the texts do not compel it. The inevitability reading was the orthodoxy of the Second International, where it did political work by justifying patience, and it drew on Marx’s language about natural laws working with iron necessity and about the more developed country showing the less developed the image of its future. Against it stand Rosa Luxemburg’s formulation of an alternative between a transition to socialism and a regression into barbarism, which explicitly denies that the outcome is settled, and Marx’s own late refusal to treat his account of western European development as a fate prescribed for all peoples. The strongest textual claim is narrower than inevitability: that capitalism generates crises and conflicts it cannot permanently resolve. What follows from those crises is left open, and the record since has been considerably kinder to the narrow claim than to the broad one.

Q: Is Marxism compatible with free will?

The determinism at issue in this dispute is historical rather than metaphysical, and the two are independent. A metaphysical determinist holds that every event, including every decision, is fixed by prior states of the world. A historical determinist holds that the large-scale course of history follows a determinate path. One can hold either without the other, since aggregate regularities are compatible with indeterminacy in individual cases, as every statistical pattern in a population demonstrates. Marx did not address the metaphysical question and nothing in his work commits him on it. The serious version of the objection is different and worth taking: if the outcome is settled by economic development, deliberation about what to do is pointless and a movement holding this belief will not act. That is an objection about the practical incoherence of fatalism, it was pressed hard inside the socialist movement, and it did more than any philosophical argument to end the determinist consensus.

Q: How do you tell if an argument is economically determinist?

Four questions settle it, and they work on writers who never announce a position. Ask what the argument treats as needing explanation and what it treats as given; an account that explains politics by interests and never asks what explains the interests has placed the economic level outside the analysis. Ask what the argument permits when economic conditions are held constant; if substantially different political outcomes are possible on the same base, the position is a moderate one. Ask whether a mechanism is specified, since naming a relation between structure and outcome is not the same as showing how one produces the other. Ask what the writer says would count against their account. These questions apply beyond Marxism, to rational choice accounts treating preferences as given and to cultural accounts treating values as self-explaining, and every framework has something it declines to explain.

Q: Was the Second International’s Marxism deterministic?

Its dominant reading was, and the reasons were strategic as much as philosophical. Karl Kautsky’s expositions of the Erfurt programme presented the arrival of socialism as the outcome of economic development, with the party’s task being to organise, educate, and be ready, and Georgi Plekhanov argued that individuals can accelerate or retard a development whose general direction is fixed by economic conditions. This suited a mass party operating legally under an authoritarian state, since it supplied a reason for patience, a reason to avoid premature confrontation, and morale that survived defeats. It was not a misunderstanding produced by ignorance; its exponents knew the texts extremely well. Its collapse came from events rather than argument, above all from the socialist parties voting for war credits in 1914, which demonstrated that class position had not determined political conduct at the moment when the theory said it should.

Q: Is technological determinism the same as economic determinism?

They are distinct charges with different evidence, though they are frequently run together. Technological determinism holds that the development of productive technique drives social change and that social relations adapt to the technology in use. The case against Marx rests largely on the remark in the polemic against Proudhon linking the hand mill to feudal society and the steam mill to industrial capitalism, and on the priority given to the productive forces. The strongest reply is that the productive forces on Marx’s account include labour power, skill, and organisation rather than machinery alone, so the priority of the forces is not the priority of technique. The most interesting reply is that the analysis in Capital of the introduction of machinery makes it depend on the length of the working day and on the resistance of skilled workers, which reverses the arrow precisely where a technological determinist would not expect it.

Q: What do students most often get wrong about economic determinism?

Three errors dominate. The first is treating the alternative to determinism as idealism, which produces essays arguing that because Marx was not a crude determinist he must have thought ideas drive history. The space between those positions contains almost every serious account of historical causation. The second is assuming that demonstrating Marx was not a crude determinist amounts to defending the theory; once it has been qualified, the burden shifts to saying what it still claims, and an answer that stops at the qualification has left the theory undefended. The third is running historical determinism into the free will debate met in philosophy, which converts a question about historical explanation into a question about metaphysics that neither the texts nor the seminar can resolve. The question that surfaces the last error usefully is whether a statistical regularity in a population tells you anything about whether any individual had a choice.

Q: Does the theory apply to all societies or only to capitalism?

This is genuinely unsettled and is one of the reasons the dispute persists. The general formulations, particularly in the 1859 preface, are stated without restriction and read as claims about human history as such. The detailed analytical work is almost entirely about capitalism, where the mechanisms are specified with a precision found nowhere else in the corpus. Marx himself refused to universalise, denying that his sketch of the genesis of capitalism in western Europe amounted to a theory of the path every people must tread, and considering in his last years whether the Russian village commune might permit a different route. Some readers conclude that the theory is properly a theory of capitalist development that was overextended by its author and his followers. Others hold that the general claims are meant and that the difficulty is one of evidence rather than scope. Nothing in the texts settles it.

Q: Why does the determinism dispute matter for politics?

Because the answer changes what a political organisation should do. If capitalist development produces socialism with something approaching necessity, the rational strategy is to build organisation, avoid risking forces prematurely, and wait, which is precisely the strategy the largest socialist party in the world pursued before 1914. If the outcome depends on political action, the questions of when to act, what to build, and how to win consent become decisive, and failure becomes a political failure rather than a sign that conditions were not ripe. Every major strategic argument in the tradition, from the revisionist controversy to the disputes over reform and revolution to the arguments about hegemony, is at bottom an argument about how much the economic development settles. That is why the dispute has been fought with an intensity that a purely philosophical question would never have generated.