Intersectionality entered public language as a slogan, but it began as an argument about court cases. In 1989 a legal scholar published an article in a law journal contending that American anti-discrimination doctrine could not see a particular kind of plaintiff. A Black woman who had been treated badly at work, in a way that was neither the treatment meted out to white women nor the treatment meted out to Black men, was required by the structure of the doctrine to choose: she could bring a race claim, comparing herself with Black men, or a sex claim, comparing herself with white women, but the law had no comfortable category for the compound claim that described what had actually happened to her. The word coined for the missing category was intersectionality, and the argument was doctrinal and remedial, a proposal for repairing a body of law, rather than a theory of society in general. Almost everything confused in the later quarrel between intersectionality and Marxism flows from forgetting that origin, and almost everything illuminating in the quarrel flows from a single analytical question buried inside it: whether class names a relation of exploitation or one category of disadvantage among others.

That question is the subject of this article, and it is a real question, not a proxy for a political quarrel conducted by other means. On one side stands an objection from within Marxism, stated at full strength in what follows: a framework that places class on a list alongside race, gender, and other axes of disadvantage has already changed what class is, because class in the Marxist tradition is not a disadvantage a person carries but a position within a structure of production that other people also occupy from the opposite side. On the other side stands an objection from within the intersectional tradition, also stated at full strength: Marxist movements and Marxist theory have a documented record, running across more than a century, of treating race and gender as secondary matters to be settled after the main conflict is resolved, and the charge of class reductionism names that record rather than inventing it. Both objections are substantially correct about the targets they actually hit. The work of the article is to establish which targets those are.

Three recoveries organise the argument. The first is the legal origin itself: what the 1989 article argued, which employment case it used to expose the doctrine’s failure, and why an argument about the pleading rules of discrimination law is not interchangeable with the general social theory later built under the same name. The second recovery is an antecedent that the standard telling omits: in 1949, four decades before the term existed, a Trinidadian-born theorist writing inside an American communist party published an essay arguing that Black women in the United States faced a compounded oppression which neither the labour movement nor the women’s movement of her day was equipped to address, and she made the argument with Marxist tools. The third recovery is a distinction between two versions of intersectionality that are habitually run together: an additive version, in which oppressions are separate burdens that accumulate on the same shoulders, and a constitutive version, in which the systems in question produce one another’s concrete forms, so that class is always lived through race and gender and they through it. The Marxist objection is close to decisive against the first version and much weaker against the second, and the founding legal argument belongs, on a careful reading, to the second.

A note on method before the evidence. This article names no politician, party, campaign, or institution of the present day, and it enters no live dispute inside feminism or inside the left, because the debate it examines is a scholarly one with a documentary record, and the record is where its interest lies. Where a contested term such as identity politics appears, it appears as an expression whose usage is described, attributed to the writers who use it in a given way, never as this article’s own characterisation of anyone’s position. The reader who came looking for ammunition will find the article badly designed for the purpose, since its central finding is that each side’s objection lands on a real fault in the other’s weakest formulations and misses the other’s strongest ones.

A word coined for a courtroom problem

The term was introduced by Kimberlé Crenshaw, a scholar of civil rights law, in an article titled Demarginalizing the Intersection of Race and Sex, published in the University of Chicago Legal Forum in 1989. The venue matters as much as the date. This was a contribution to legal doctrine, addressed to judges, litigators, and scholars of discrimination law, and its complaint was precise: the doctrine organised claims along a single axis. A plaintiff alleging race discrimination was measured against the experience of men of her race; a plaintiff alleging sex discrimination was measured against the experience of white women; and a Black woman whose treatment resembled neither pattern found that the comparison classes the law supplied dissolved her claim between them. Crenshaw called this the single-axis framework, and her argument was that it distorted the analysis of discrimination, the design of remedies, and the politics of the movements that relied on the law.

Where did the term intersectionality come from?

From a law journal article of 1989 arguing that anti-discrimination doctrine failed Black women by requiring them to bring either a race claim or a sex claim, when their treatment was compound. It began as a legal and remedial argument, not as a general social theory.

The illustration at the centre of the article was an employment case from the previous decade, DeGraffenried v. General Motors, decided in 1976. Five Black women sued the company over a seniority system whose last-hired, first-fired logic wiped out the employment of Black women as a group, because the company had not hired Black women at all before 1964 and had therefore left every one of them exposed when layoffs came. The court’s handling of the claim became Crenshaw’s exhibit. The plaintiffs could not proceed as women, because the company had hired women, white women, before 1964; they could not comfortably proceed as Black workers, because Black men had been hired earlier and the race claim was expected to be made on behalf of all Black employees; and the court refused to let them proceed as Black women, a combined class, partly from an expressed fear of opening a door to endless compound categories. The result was that a policy whose burden fell with demonstrable precision on Black women fell, in the eyes of the doctrine, on no one the law could recognise.

Crenshaw set two further cases beside it, and the pattern across the three is the argument. In one, a court declined to let Black women represent all women in a sex discrimination action, on the ground that their experience, marked by race, might diverge from that of the class they sought to represent. In another, the same judicial logic ran in the opposite direction, requiring Black women plaintiffs to distinguish their claim from a race claim brought by Black men. Read together, the decisions showed doctrine protecting the purity of its categories at the expense of the claimants the categories were built to serve: where race and sex coincided in one person’s treatment, the law’s demand that the claim be one thing or the other made the treatment itself legally invisible. The point was never that judges were unusually obtuse. It was that a framework built on single-axis comparisons generated the invisibility automatically, whoever operated it.

The image Crenshaw offered for the structure of the problem gave the term its name. Discrimination, she suggested, works like traffic moving through an intersection: it can flow from several directions at once, and an injury sustained in the middle of the crossing may have been caused by cars arriving along any of the roads, or by several together, in ways that an inquiry insisting on a single road cannot reconstruct. The image is worth dwelling on because of what it quietly assumes. The roads exist before any particular traveller reaches the crossing; the harm is compound because the systems producing it are distinct in origin and intersecting in effect; and the failure being diagnosed is a failure of an observer, the law, whose categories force a compound event into a single-cause file. Nothing in the original figure presents the traveller’s identity as the object of the theory. The object is the traffic: structures of race and sex discrimination whose interaction a remedial system had failed to model.

That remedial character is the first thing later usage discarded, and recovering it changes the evaluation of nearly every claim made in the term’s name. The 1989 article proposed no general account of how societies are stratified, offered no ranking of oppressions, and issued no programme. It identified a mismatch between the shape of a body of law and the shape of the harms the law existed to remedy, and it argued, within the professional conventions of legal scholarship, for doctrinal repair: compound claims should be cognisable as such, comparison classes should not be allowed to dissolve them, and the experiences of those who stand where categories cross should be treated as evidence about the categories rather than as anomalies to be managed. A reader who approaches the founding text expecting a manifesto will find instead something closer to a technical memorandum with radical implications, and the gap between the text and its reputation is itself a datum about how terms travel.

Was the founding argument about identity?

No, and the distinction carries the whole later dispute. The plaintiffs in the cases the article examined were not asking the law to affirm who they were; they were asking it to register what had been done to them, by a seniority system, an employer’s hiring history, and a set of doctrinal rules.

Their position in the structure of the workplace was the substance of the claim. Identity, in the sense of self-understanding or cultural membership, formed no part of the legal reasoning, and the theory built directly on the article can dispense with the concept entirely: what intersects, in the founding argument, are systems of treatment, not selves. Later usage blurred exactly this point, recasting an argument about structures as a vocabulary for describing persons, and much of the Marxist objection examined below is aimed, accurately, at the blurred version rather than at the argument of 1989.

The article’s second concern is often missed because the first is so quotable. Crenshaw criticised not only courts but also the political movements that had organised themselves around the single axes the doctrine supplied. Feminist theory, she argued, had generalised from the experience of women who were otherwise privileged by race and class, and antiracist politics had generalised from the experience of men, so that the agenda of each movement was set by the members least burdened by the other system. The critique was thus double-headed from birth: it indicted a legal framework and the movement frameworks that mirrored it. This matters for the debate with Marxism because the same double structure reappears there. The intersectional objection to Marxist movements, examined at full strength later in this article, is an application of the 1989 pattern to a third movement: an emancipatory politics whose categories, generalised from a part of its constituency, schedule the compound burdens of another part for later attention.

The second article and the widening of the claim

Two years later, in the Stanford Law Review of 1991, Crenshaw published Mapping the Margins, which extended the analysis from doctrine into the politics of violence against women and gave the framework a threefold internal structure that later writers have used ever since. She distinguished structural intersectionality, concerning the way differently situated women encounter institutions such as shelters, immigration rules, and labour markets from unequal positions; political intersectionality, concerning the way women of colour are situated between two political agendas, feminist and antiracist, each of which can subordinate their interests to its own coherence; and representational intersectionality, concerning the cultural images through which such women are depicted and the contests over those images. The distinctions are analytical instruments, not a creed, and the article’s concrete materials were rape crisis provision, battered women’s services, and the treatment of immigrant women, domains in which the cost of single-axis thinking was measured in unmet need rather than in lost lawsuits alone.

Even in this widening, the argument remained anchored in institutions and their failures. The structural strand asked why a shelter system designed around an abstract woman could be unusable by a woman whose immigration status, language, or employment made its routines inaccessible. The political strand asked why raising violence within a minority community could be treated as disloyalty by one movement and as the only permissible topic by the other. These are questions about the design of provision and the discipline of coalitions. They are not, in themselves, a metaphysics of the self, and the distance between Mapping the Margins and the term’s later career as a free-floating synonym for the acknowledgement of multiple identities is the distance the rest of this article keeps in view.

The journey from law reviews into general social theory was made by many hands over the following two decades, and the mechanism of the journey deserves description because it explains both the term’s reach and its deformation. A concept coined for a specific institutional failure proved portable: any field whose categories dissolved compound cases could use it. Sociologists took it up as a paradigm for studying stratification; political scientists applied it to representation and policy; education and public health built research programmes on it. Portability, however, taxed the concept in a specific way. In the legal original, the intersecting systems were specified, race and sex discrimination as administered by employers and courts, and the compound harm was a demonstrable event. In transit, the list of systems lengthened, the specification loosened, and intersectionality could shrink into a gesture: the acknowledgement that many things matter at once. A gesture cannot be falsified, cannot guide a remedy, and cannot distinguish a strong analysis from a weak one. The Marxist critics who attack the gesture are attacking something real; the error is to suppose the gesture exhausts the term’s content, when the founding texts supply criteria by which the gesture itself can be judged a degeneration.

An argument with a Marxist antecedent

The standard history of intersectionality begins in 1989, and the standard history is incomplete in a way that reverses the usual story of two traditions meeting as strangers. In June 1949, forty years before the term was coined, the journal Political Affairs, the theoretical organ of the Communist Party of the United States, published an essay by Claudia Jones titled An End to the Neglect of the Problems of the Negro Woman. Jones, born in Trinidad and raised in Harlem, was a leading figure in the party’s work among Black Americans and a journalist of long standing. Her essay argued that Black women in the United States occupied a position compounded of three subordinations at once, as workers, as Black people, and as women, and that this compounding was not a sum of three disadvantages but a distinct location with its own intensity and its own political consequences. Later scholarship has given the structure Jones described a name she did not use: triple oppression.

The essay’s complaints were aimed at her own movement, which is what makes it an antecedent rather than merely a parallel. Jones charged the labour movement and the party itself with neglecting the specific conditions of Black women workers, who were crowded into the worst-paid corners of domestic and agricultural labour, and with treating their organisation as an afterthought to the organisation of men. She charged the women’s movement of her day with an equal and opposite neglect. And she argued that the neglect was not only unjust but strategically ruinous, since the most oppressed layer of the working class, once organised, would be its most militant: the compounding that doctrine elsewhere dissolved was, in her analysis, a source of political capacity. Every element of this structure reappears in the later history of the concept: the compound position, the indictment of movements organised on a single axis, and the claim that the standpoint of the multiply burdened is epistemically privileged for understanding the whole.

Jones’s method was Marxist in the strict sense that her categories were drawn from the organisation of labour. Her analysis of Black women’s position began from the occupations they were crowded into and the pay attached to those occupations, and treated their super-exploitation, her word for exploitation intensified past the norm by race and sex, as a fact about the labour market’s structure rather than about attitudes. She was deported from the United States in 1955, under legislation aimed at Communist Party members, and continued her work in London, where she founded a newspaper serving the Caribbean community and helped to establish the carnival that became an annual institution of that community’s public life. She died in 1964 and is buried in Highgate Cemetery, near Marx. The biographical details matter here only for one reason: the essay of 1949 was not the work of a critic outside the Marxist tradition lobbing objections at it, but of one of its organisers, applying its method to a failure she documented from inside its own institutions.

Who made the compound argument before the term existed?

The clearest antecedent is Claudia Jones, whose 1949 essay in a communist party journal argued that Black women faced a threefold oppression, as labourers, as Black people, and as women, which the labour movement and the women’s movement alike had failed to address. Her method was Marxist: the argument begins from the jobs Black women were confined to and the wages those jobs paid.

Between Jones and Crenshaw stands the statement that supplied the later framework with its most durable phrase. In 1977 a group of Black feminists in Boston, the Combahee River Collective, named for a river in South Carolina where Harriet Tubman had led an armed raid that freed hundreds of enslaved people during the Civil War, issued a statement that has become one of the most reprinted documents of American feminism. The Collective described the systems bearing on Black women’s lives as interlocking: racial, sexual, heterosexual, and class oppression, in their formulation, were not separable layers but a single fabric, and they declared that the synthesis of these oppressions created the conditions of their lives. The statement also coined the usage that later attached itself, usually unfairly, to intersectionality as such: the Collective used the expression identity politics to describe a politics grounded in one’s own material position, the conviction that the most radical politics grow directly out of the experience of those who live the conditions to be changed. In context the phrase named a materialist method, beginning from one’s location in a structure, and its later career as a term of abuse for politics indifferent to material structure would have surprised its authors.

The Collective was explicit about its intellectual allegiances, and they complicate any tidy genealogy. Its members described themselves as socialists, held that the liberation of Black women required the destruction of the political and economic systems they named as capitalism and imperialism as well as patriarchy, and criticised both the feminist movement’s racism and the left’s sexism from a position inside both arguments. The statement is therefore best read not as a rival founding document to Jones’s essay but as its continuation by other means: the compound analysis, the double indictment of movements, and the insistence that the systems interlock rather than accumulate all descend in a line that runs through the Marxist tradition even where its bearers were in open conflict with that tradition’s organisations. When Crenshaw, a decade later, gave the structure a legal formulation and a name, she was formalising an analysis that Black feminist socialists had been developing for a generation.

Other antecedents crowd the nineteenth century, and a full genealogy would give them chapters. Anna Julia Cooper, in A Voice from the South of 1892, argued that the Black woman occupied a position from which the condition of both her race and her sex could be judged as from no other. Sojourner Truth’s interventions in the women’s conventions of the 1850s pressed the same point against audiences disposed to a single axis. These earlier figures asserted the compound standpoint; what Jones added in 1949 was its derivation from the structure of employment, and what Crenshaw added in 1989 was its translation into a critique of legal doctrine with cases attached. The line of descent is a progressive formalisation of one insight, and recognising the line dissolves the framing, common on both sides of the later quarrel, in which intersectionality arrives from outside Marxism to accuse it. Part of the accusation was composed inside the tradition, in its own journals, before the term existed.

The matrix of domination

One year after the legal article, and independently of its doctrinal setting, the sociologist Patricia Hill Collins published Black Feminist Thought in 1990, the book that carried the analysis furthest into general social theory and supplied its most ambitious organising concept. Collins proposed that race, class, and gender be understood not as separate systems that occasionally collide but as a single matrix of domination, a historically specific organisation of power in which each of the intersecting structures is arranged through the others, and in which every person is positioned as both oppressor and oppressed along different dimensions of the matrix. The concept did specific work against the additive picture. In a matrix, there is no bedrock oppression to which the others are added as modifiers; the configuration itself is the object, and two people who share a class position but differ by race and gender do not experience that position in a common core plus decorations, but differently in kind.

Collins’s book also fixed the epistemological strand of the tradition. Black feminist thought, on her account, was an intellectual tradition produced by Black women’s location, carried in their work, their music, their literature, and their everyday strategies of resistance, and systematically devalued by the institutions that certify knowledge. The claim paralleled the standpoint arguments being made inside Marxist feminism in the same decades, and the parallel is not accidental: both hold that a social order is most legible from positions where its mechanisms press hardest, and both must then explain why legibility is not automatic, why a location supplies questions rather than answers ready-made. Collins’s answer, like the Marxist feminists’, ran through collective work: standpoint is an achievement of organised reflection on a location, not a birthright conferred by it. That refinement, present in the founding texts of both traditions, is routinely dropped by their popularisers, to the detriment of each.

For the argument with Marxism, the matrix concept sharpens the real difficulty rather than dissolving it, and the sharpening is a service. If the systems interlock constitutively, then isolating the causal contribution of any one of them is a modelling choice, not a discovery; yet Marxist analysis characteristically claims more than that for class, claiming that the organisation of production explains why the matrix takes the historical forms it does, why its other axes are recruited, bent, and periodically rebuilt around the needs of accumulation. Whether that explanatory priority can survive the matrix picture, or whether the matrix picture demotes class to one structure among peers, is the substantive question underneath the polemics, and it is a question about explanation, not about respect. The two versions table later in this article is designed to make it askable in a form students can use: the additive and constitutive versions give different answers to it, and conflating them is what allows each side to refute the version the other does not hold.

Class as a relation, class as an attribute

The Marxist side of the dispute cannot be understood until its central category is stated with the precision the tradition insists on, because the whole objection turns on a distinction that everyday language erases. In ordinary usage, class functions as an attribute: a person is working class much as a person is tall or left-handed, a property carried from situation to situation, available for inclusion in a description of who someone is. Stratification sociology often formalises the attribute view, sorting populations by income, occupation, and education into bands whose members share a level of advantage. In Marx’s usage, class is none of these things. It is a relation. A wage labourer and a capitalist are not two kinds of person distinguished by their holdings; they are the two poles of a single relation, the purchase of labour power and its use in production, and neither pole exists without the other. The category is defined by position within a process, and the process is exploitation: the extraction of surplus labour from those who must sell their capacity to work by those who own the means with which work is done.

The linked article on the Marxist theory of class develops the relational definition in full; what matters here is the consequence for the dispute. A relation cannot be added to a list of attributes without changing its logical character. To write class into a series, race, class, gender, sexuality, and to treat each entry as a dimension along which a person may be advantaged or disadvantaged, is to convert the first entry, silently, into the attribute sense: class becomes a quantity of disadvantage, a rung on a ladder of resources, and the exploitative relation that defined it, with its implication that one pole’s advantage is the other pole’s product rather than a parallel misfortune, drops out of the analysis. This is not a pedantic point about vocabulary. The remedy implied by a disadvantage is redistribution toward parity; the remedy implied by an exploitative relation is the abolition of the relation itself. A framework can of course choose the first remedy knowingly. The Marxist objection insists only that the choice be noticed, since frameworks that believe they have included class while converting it into disadvantage have not included it at all.

What kind of category is class?

In the Marxist tradition it is a relation: a position within the organisation of production, defined against the opposing position it presupposes, as wage labourer is defined against employer. Treating class as an identity, one attribute of a person alongside others, converts a position into a characteristic, and the conversion is the hinge on which the debate with intersectionality turns.

The distinction also clarifies what the strongest intersectional writers were and were not claiming. When Crenshaw’s plaintiffs were dissolved between comparison classes, the harm was located in a structure, an employer’s seniority system interacting with a legal doctrine, not in anyone’s self-conception. When Jones analysed the triple position of Black women, she began from their confinement to specific labour markets. Neither analysis needs class to be an identity, and the constitutive version of intersectionality, in which systems produce one another’s forms, can accommodate a relational account of class without strain: the wage relation in a given society is always already organised through racial and gendered divisions of labour, so that the relation as it concretely exists is never found in a pure state. The additive version cannot accommodate it, because addition requires commensurable items, and a relation is not commensurable with an attribute. Here, in miniature, is the article’s thesis: the dispute is genuine against one version of intersectionality and largely a misunderstanding against the other, and the failure to distinguish the versions is the single largest source of wasted argument in the literature.

The Marxist objection to intersectionality, stated at full strength

The objection must now be given in its strongest form, attributed to the writers who have actually made it, because the debate’s literature is littered with refutations of versions nobody holds. Its most systematic recent statement by a Marxist feminist is Martha Gimenez’s, in Marx, Women, and Capitalist Social Reproduction, published in 2018. Gimenez argues that intersectionality, whatever its merits as a description of how disadvantages coincide, is not and cannot be a theory of the society that produces them, because it lacks an account of capitalism as a system with its own dynamics. In her formulation, the framework catalogues the intersections among pre-given categories, race, gender, class, whose existence it assumes, while the Marxist question is how the categories themselves, and the relations among them, are produced and reproduced by the organisation of production. A list, however refined, does not explain its own entries.

An earlier and sharper version of the same objection, aimed at the concept’s popular career, is Eve Mitchell’s essay I Am a Woman and a Human: A Marxist Feminist Critique of Intersectionality, published in 2013. Mitchell’s target is the use of intersectionality as a theory of identity, and her argument is that the framework, so used, inverts the direction of explanation. Where Marxist analysis begins from the objective structure of production and asks how consciousness and identity are formed within it, an identity-first framework begins from subjective self-understanding and treats the structure as background scenery. The political consequence Mitchell draws is the one the brief for this debate requires to be stated plainly: frameworks organised around identity generate demands for recognition, for the acknowledgement and representation of the groups the framework enumerates, while a framework organised around exploitation generates demands about ownership, about who controls the means of production and on what terms labour is performed. Recognition can be granted by the existing order without cost to its structure; a change in ownership cannot. The migration of the term into institutional life, where it functions fluently in the vocabulary of diversity administration, is for Mitchell not an accident of reception but evidence about the concept’s centre of gravity.

Behind these named statements stands the objection’s analytical core, which can be separated from any polemic and tested on its merits. It has four components. First, the conversion argument already developed: listing class among axes of disadvantage converts a relation into an attribute and loses the mechanism of exploitation. Second, the aggregation problem: an additive framework has no principle for weighting its axes, and so cannot explain why, in a given historical setting, one line of division organises the others, why the colour line in one society and the confessional line in another becomes the form through which class is lived. Third, the absence of a theory of production: the framework describes the distribution of burdens with great sensitivity but is silent about the process that generates the surplus over which the burdens are distributed, and silence about production is, for a materialist, silence about the engine of the whole arrangement. Fourth, the direction of remedy: because the framework’s categories are categories of persons, its natural political output is the fairer representation of persons within existing institutions, rather than the transformation of the institutions’ economic basis.

Each component deserves the respect of an answer, and the answers exist; they are assembled in a later section. But the objection’s force should be registered before any answer is attempted, because it is routinely understated by the framework’s defenders. The objection is not that attention to race and gender is a distraction, a claim that belongs to a cruder tradition the intersectional side is entitled to refuse. Gimenez is a feminist of fifty years’ standing in the movement she criticises; Mitchell writes as a Marxist feminist, not as an opponent of feminism. Their charge is that a specific theoretical architecture, however morally compelling its motivating cases, cannot do the explanatory work it is asked to do, and that in the attempt it dissolves the one category, class as a relation, that connects feminist analysis to a theory of how the social order reproduces itself. Whether the charge succeeds depends, as the next sections show, on which version of intersectionality it is aimed at. Against the additive version, the version of posters and training materials in which oppressions are stacked like weights, the charge is decisive, and the honest defenders of the framework have generally conceded as much by retreating to the constitutive version whenever pressed.

What exactly does the Marxist objection deny?

It does not deny that race and gender structure people’s lives, nor that their burdens compound. It denies that a framework listing class as one axis of disadvantage among others has preserved what class means to Marxism, and it denies that it can explain the system producing the disadvantages it catalogues. The objection concerns architecture, not whose suffering counts.

There is also a historical edge to the objection, less often articulated but present in the named statements, concerning the concept’s timing. Intersectionality rose to dominance in the academy during the decades in which organised labour contracted and class language receded from public life, and its critics ask whether a framework that disperses class into a plurality of identities could have achieved its institutional triumph in a period when class organisation was strong. The question is legitimate as sociology of knowledge and illegitimate as refutation: a concept is not false because its career was convenient. The distinction matters because the debate repeatedly commits the genetic fallacy in both directions, intersectionality dismissed as a creature of foundations and universities, Marxism dismissed as a creature of parties and states, and in both cases the dismissal substitutes an account of a theory’s patrons for an examination of its claims. This article declines the substitution in both directions, and notes it here because the objection in its strongest form, Gimenez’s and Mitchell’s, does not need it.

A further clarification protects the objection from its own vulgarisation. The claim that class is a relation does not entail that race and gender are unreal, secondary, or destined to dissolve when the relation is transformed. Relations of production in every actual society are instituted through racialised and gendered divisions of labour; the wage relation arrived in the Americas entangled with enslavement, and in industrial Europe with a legal regime that made married women’s property their husbands’. A Marxism that treats those entanglements as decoration on a pure economic structure has misunderstood its own founder, whose account of primitive accumulation is an account of violence organised by race and colony from the first page. The objection, properly stated, is narrower and stronger than the caricature: it objects to a grammar, the grammar of the list, and to the theoretical losses the grammar imposes. It does not object to the phenomena the list attempts to honour.

The intersectional objection, stated at full strength

The other objection is equally owed its strongest form, and it is the one the Marxist side more often understates, because its evidence is historical rather than analytical and therefore harder to dismiss by definition. It says that Marxist theory and Marxist movements have a documented record of subsuming race and gender into class and of deferring them until a future settlement, and that the charge of class reductionism describes a real tendency with a real archive, not a caricature invented by rivals. The archive is the point: unlike the Marxist objection, which is a claim about a concept’s logical structure, the intersectional objection is a claim about what organisations did, and it is answered, if at all, by evidence about what organisations did.

The record begins with the movement Jones wrote to in 1949. Her essay is itself the evidence, since its occasion was the neglect she charged. But the pattern she named is visible across the movement’s institutions. Socialist parties of the early twentieth century routinely treated women’s suffrage as a demand to be supported in principle and postponed in practice, on the ground that dividing the working class by sex weakened its unity, a logic that protected male workers’ privileges in the name of unity and discovered, when pressed, that unity meant the postponement of women’s claims. The American labour movement’s record on race followed the same logic: the great industrial unions of the 1930s admitted Black workers and built a genuine interracial movement, and then, in the postwar decades, allowed seniority rules, segregated locals, and apprenticeship practices to reproduce racial hierarchies inside the movement itself, while the leadership’s answer to complaint was, characteristically, that raising race would divide the class. The linked article on whether Marxism neglected gender assembles the evidence on the gender side of this record in detail; the pattern on the race side is symmetrical.

Angela Davis documented the pattern for the women’s movement in Women, Race and Class, published in 1981, and her argument deserves the attribution it usually receives, since it is the standard scholarly statement of the intersectional objection’s historical half. Davis showed that the American women’s movement, from the suffrage campaigns through the feminism of the nineteen seventies, repeatedly constructed its universal woman from the experience of white women of the middle class, and that its central symbols, the entry of women into the professions, the liberation of the housewife from the home, spoke to women whose labour had never been confined to the home in the first place. Black women had been working in other people’s homes and in the fields for generations; the demand to leave the kitchen for the office misdescribed their condition at the level of the description itself. Davis’s target was liberal feminism, and her book was written from a position within Marxist politics, as a member of the American Communist Party. Its implication for Marxism is nonetheless direct: a class analysis that takes the male industrial worker as its representative figure, and treats the household, the field, and the informal sector as marginal to production, commits exactly the error of generalisation from a part that the 1989 legal article identified in doctrine.

The deferral pattern has a temporal signature that recurs across the century, and naming it is what makes the charge falsifiable rather than atmospheric. It runs: acknowledge the secondary claim, affirm its justice, and schedule its resolution after the primary conflict is won, whether the primary conflict is named as the revolution, the capture of state power, or the rebuilding of the economy. The scheduling does two kinds of work at once. It protects the movement’s unity in the present by removing a divisive item from the agenda, and it protects the movement’s theory in the future by ensuring that the item is never allowed to test the theory’s claim that it can absorb every other division. The Combahee statement’s fury at this pattern is intelligible once the pattern is seen as a structure rather than a sequence of accidents: the Collective’s members were writing after a decade in which the women’s movement had treated their concerns as a distraction and the Black liberation movement had treated their feminism as a betrayal, and their statement is a refusal to be scheduled twice over.

Does the reductionism charge have an archive?

Yes, and the evidence is documentary. Socialist and labour movements repeatedly acknowledged race and gender claims, affirmed them, and scheduled them after the primary conflict, in a pattern that protected the movement’s unity in the present and its theory alike. The charge names that pattern, not an insult traded between camps.

It is met, if it is met at all, with the record of organisational practice, not with a redefinition of terms.

It must be conceded, in the interest of the same fairness, that the Marxist tradition contains its own resources against the pattern, and that the objection’s strongest form does not pretend otherwise. Jones’s essay is itself the demonstration that the tradition’s tools could diagnose its own failure. The tradition’s theory of social reproduction, developed in the domestic labour debate and carried into the present by the unitary framework examined below, was built precisely to answer the charge that Marxism treats the household as outside production. And the socialist states’ record, examined in the linked article on Marxist feminism’s tradition, contains genuine advances, in women’s employment, childcare provision, and legal equality, that no honest version of the objection can omit. The objection at full strength is therefore not that Marxism is incapable of addressing race and gender. It is that its institutions, across a century of practice, showed a repeated tendency to defer them, that the tendency is documented in its own journals and memoirs, and that a theory which cannot explain why its own bearers kept doing this has a gap in its self-understanding. A Marxist answer that treats the charge as a slur, rather than as a description of a tendency with an archive, repeats the gesture the charge identifies.

Two versions, one word

Most of the published quarrel becomes intelligible, and much of it evaporates, once a distinction between two versions of intersectionality is kept in view. The versions have no official names, and different writers mark the difference differently, but the difference itself is stable across the literature and is captured here by the pair additive and constitutive. The additive version treats systems of oppression as separate burdens that accumulate in a life: a person may be disadvantaged by class, and then additionally by race, and then additionally by gender, each burden distinct in nature and summing with the others to a total weight. The picture is intelligible, portable, and morally vivid, which is why it dominates the term’s popular career. The constitutive version holds that the systems do not merely coincide in a life but produce one another’s concrete forms: the class structure of a given society is built through racial and gendered divisions of labour, so that no worker’s class position exists apart from the racial and gendered organisation of the labour market that allocates it, and conversely race and gender take forms specific to the class order they run through.

Dimension	Additive version	Constitutive version
How oppressions relate	Distinct burdens, summed on a single person	Systems that produce one another’s concrete forms
How class is treated	One attribute of disadvantage, listed beside the others	A relation that exists only in racialised and gendered forms
What the Marxist objection shows	That the list converts a relation into an attribute, decisively	That an explanatory priority for production must be argued, not assumed
What it implies for analysis	A fuller inventory of the burdens a person carries	An account of how each structure shapes the others it intersects

The two versions table orders the debate because it relocates every standard move in it. The Marxist objection built on the list’s grammar is aimed at the left-hand column, and against that column it succeeds: addition presupposes commensurable and independent items, and a relation of exploitation is neither. The intersectional objection from the historical record is aimed at a Marxism that in practice treated race and gender as the right-hand column’s secondary effects, and against that practice it succeeds. But the constitutive version is not vulnerable to the grammar objection at all, because it makes no list: it asserts a mutual formation among structures, which a relational theory of class can state as readily as anyone. And a Marxism that studies how the wage relation is instituted through race and gender, as the best of its own literature has always done, is not vulnerable to the deferral charge in the way its institutions were, because the charge was about scheduling, and the constitutive literature of both traditions refuses the schedule.

The distinction also explains a curiosity of the debate that otherwise looks like bad faith on both sides: each side’s canonical refutations consistently target formulations the other side’s leading theorists also reject. Marxist critics quote training materials and popular essays in which oppressions are stacked like weights, and refute the stack; intersectional theorists answer with Crenshaw and Collins, in which the structure is constitutive from the start. Intersectional critics quote party resolutions and polemics in which race is dissolved into class by declaration, and refute the dissolution; Marxist theorists answer with Jones and with the social reproduction literature, in which no dissolution occurs. The quarrel persists because the weak versions are genuinely widespread, in circulation they may even be the majority usage, while the strong versions, being harder, are less often met. An honest guide must therefore grade on two scales at once: what the best statements of each position claim, and what the term does in the wild. The finding of this article is that the strong versions are compatible to a degree their partisans rarely admit, and that the weak versions richly deserve each other’s criticisms.

What separates the additive version from the constitutive one?

Addition treats each oppression as a separate burden, complete in itself before it meets the others, so that a life can be described by listing and summing them. The constitutive version denies that the burdens exist in separable form at all.

It holds that class, in any society that actually exists, takes shape through racial and gendered divisions and is shaped by them in return, so the analysis must begin from their mutual formation rather than from a list.

It follows that disputes about intersectionality are often disputes about nothing until the version in play is specified, and the first question to ask of any use of the term, in a seminar room or in a policy document, is which version the use presupposes. The question is diagnostic in both directions. A use that enumerates axes and calls for their joint acknowledgement, without any account of how the axes produce one another, is additive whatever its author’s allegiance, and is exposed to the Marxist objection in full. A use that traces how a labour market’s racial segmentation constitutes the class structure of an industry, and how that structure in turn fixes the forms racial disadvantage takes, is constitutive, and a Marxist who attacks it with the grammar objection has mistaken the target. The table is not a courtesy device. It is the debate’s missing instrument panel, and teaching the debate through it, rather than through the polemics, is the single change that would most improve its conduct.

The relation and identity distinction

The distinction that does the most work in this debate deserves its own statement, since the article’s namable claim rests on it. Ask of any analysis that uses the word class which of two things the word is doing. In the first use, class names a relation: the analysis is about a structure of production, about who must sell labour power and who buys it, and the persons who appear in the analysis appear as occupants of positions the structure defines. In the second use, class names a category of disadvantage: the analysis is about a population sorted by its level of resources or standing, and the word functions as it functions in census tables and ministry reports. The two uses are not stronger and weaker versions of one idea. They are different ideas sharing a word, and they license different inferences. From class as a relation it follows that the advantaged pole is implicated in the disadvantaged pole’s condition, since the relation produces both; from class as a category it follows only that the distribution is uneven, which might be anyone’s fault or no one’s.

The relation and identity distinction grades the whole intersectionality debate once it is applied consistently. The founding legal argument passes the test in a way that surprises readers who meet the term only in its popular form: Crenshaw’s plaintiffs were located by their position in an employment structure, and the compound harm she identified was an interaction between that structure and a body of doctrine, not a property of the plaintiffs’ self-understanding. Jones’s 1949 essay passes it in the same way, beginning from labour markets. The additive popular version fails it, because addition requires that class be one measurable property among others. And the test cuts against careless Marxism with equal readiness: a Marxism that treats the working class as a cultural identity, a matter of accent, dress, and inherited loyalty, has committed the same conversion in the opposite direction, turning a relation into a badge, and is correspondingly defenceless when the badge fails to predict behaviour. The distinction is tradition-neutral. It asks a question about grammar that any framework must answer, and the frameworks that answer it well converge, while those that answer it badly share their faults across the supposed divide.

This is why the article’s claim is that most of the argument between the two traditions resolves, on inspection, into a dispute about the additive version which neither tradition’s best theorists defend. The residue that survives the resolution is genuine and should not be wished away. It concerns explanatory priority: whether the organisation of production explains the historical forms taken by race and gender, as the Marxist tradition characteristically claims, or whether the matrix of mutually constituting structures has no such centre, as the constitutive version in Collins’s formulation suggests. That residue is a hard question in social theory, comparable in difficulty to the questions the tradition argues about internally, and it is not settled by anything said here. But it is a much smaller and much more interesting disagreement than the public quarrel suggests, and it is the disagreement the two traditions’ serious writers are actually having, in the works cited in this article, while the larger noise proceeds around them.

Attempts at synthesis

The compatibility of the strong versions has not gone unnoticed by their holders, and the attempts at synthesis are part of the record rather than a hope appended to it. Three deserve description. The first is the unitary framework of social reproduction theory, whose core text is Lise Vogel’s Marxism and the Oppression of Women, published in 1983 and recovered for a new generation by the collection Social Reproduction Theory, edited by Tithi Bhattacharya in 2017. Vogel’s proposal refuses both the list and the dualism of two separate systems: there is one process, the reproduction of capitalist society, within which the renewal of labour power, carried on largely by women’s unpaid work in households, is internal to the capital relation rather than a second system beside it. On this account, women’s oppression is neither an added axis nor a separate patriarchy but a structural feature of how the wage relation perpetuates itself, and racial divisions enter the same analysis through the organisation of the labour markets and households the process actually uses. The linked article on social reproduction theory expounds the framework in full; its relevance here is that it demonstrates, from inside Marxism, that a constitutive treatment of gender and class is available without abandoning the relational definition of class, which is exactly the combination the debate declares impossible.

The second attempt is Cinzia Arruzza’s Dangerous Liaisons, published in Italian in 2010 and in English in 2013, which surveys the marriages and divorces of Marxism and feminism and argues, against both dual systems theory and the dispersal of class into identities, for a unitary account in which gender and race are internal differentiations of the capital relation’s concrete forms. Arruzza’s book is instructive for the debate because its polemics run in both directions at once: against feminists who treat class as one oppression among others, and against Marxists who treat the unity of the theory as a licence to ignore the differentiations. The third attempt is the book-length reconciliation by Ashley Bohrer, Marxism and Intersectionality, published in 2019, which works through the historiography of both traditions and argues that the constitutive reading of intersectionality and a relational Marxism are not merely compatible but require each other: intersectionality supplies the account of how class is concretely differentiated, Marxism supplies the account of the system within which the differentiation is produced. Bohrer’s title names the synthesis the debate’s noise obscures, and the book’s existence, from a scholar formed in both literatures, is evidence that the resolution described in this article is a position in the field and not an editorial wish.

Can the two frameworks be combined?

Yes, on the conditions the synthesis literature specifies. The combination requires class to remain a relation rather than becoming an item on a list, and requires the intersectional side to remain constitutive rather than additive, an account of mutual formation rather than of accumulated burdens.

Social reproduction theory and the book-length reconciliations of the last decade show the combination at work; whether it satisfies either tradition’s partisans is a separate question from whether it is coherent.

The neighbouring debate over racial capitalism, which asks whether racism is internal to capitalism’s formation rather than an inheritance it uses, runs parallel to this one and shares several of its figures and texts. It belongs to its own article, on Black Marxism and racial capitalism, and is flagged here only to prevent a conflation the seminar room makes constantly: intersectionality is a claim about how structures compound in lives and in law, racial capitalism is a claim about the historical constitution of a mode of production, and the two claims, though allied, are tested against different evidence. Keeping the debates separate is another application of the article’s general discipline, which is to refuse the amalgam in favour of the specific argument, since it is only specific arguments that either side has ever succeeded in answering.

The doctrine after the founding cases

Because the founding argument was legal, its fortunes can be traced with unusual precision in the case law that followed, and the trace is instructive for readers who meet the term only in its theoretical career. The doctrinal problem the 1989 article exposed was never wholly solved, and its partial solutions show what the repair of a single-axis framework looks like in practice. American courts had already developed, before Crenshaw wrote, a workaround known as the sex-plus doctrine, under which a plaintiff could proceed not as a woman but as a woman plus some further characteristic, mothers of young children being the standard example. The workaround conceded the principle that compound classes exist while confining the concession to an exception, and it pointed in two directions at once: compound claims were cognisable, and the framework that required them to dress as exceptions remained intact. Intersectionality’s legal career has largely been the slow, uneven expansion of the exception, as courts in several jurisdictions came to accept that a claim of discrimination against Black women as such could be pleaded and proved without being disassembled into its elements.

The stubbornness of the original problem is worth understanding, because it is structural rather than attitudinal and recurs wherever categories administer remedies. A discrimination regime built on comparisons must specify the comparator, and every specification builds a theory of the harm into the procedure: to compare Black women with white women is to define sex as the operative axis and race as noise; to compare them with Black men is to define race as the axis and sex as noise; to compare them with white men is to capture the compound harm while attributing it to neither axis, which for purposes of doctrine and of statistical proof is its own kind of dissolution. The choice among comparators is unavoidable in litigation, and no choice is innocent, which is why Crenshaw’s argument was about the framework and not about the sympathies of its operators. The same structure appears in any administrative system that distributes remedies by category, in welfare eligibility, in university admissions accounting, in the design of equality audits, and the concept’s migration into those domains followed the path of the problem rather than the fashion of the academy.

A second legal lesson concerns the direction of the argument’s radicalism. The founding article is often described as though it were a demand for the multiplication of categories, an ever-finer sorting of persons into compound classes, and the fear expressed in the DeGraffenried decision, of a super-remedy available to a proliferating list of protected subgroups, is quoted as though the article endorsed the proliferation. It did not. The article’s point was that the compound claim of Black women exposed a defect in the framework’s logic, and that taking the defect seriously might require rethinking the framework’s reliance on comparison classes altogether. The difference between multiplying categories and questioning the categorical machinery is the legal form of the difference between the additive and constitutive versions: the additive lawyer pleads more boxes, the constitutive argument asks why the boxes decide. Readers who keep the legal origin in view are protected against the commonest misdescription of the theory, and the protection is one reason the origin matters beyond antiquarian interest.

The method question

When sociologists took the concept up, their first problem was methodological, and the debate they conducted about method is the part of the literature most useful to students, because it forces the two versions apart in operational terms. Leslie McCall’s article The Complexity of Intersectionality, published in the journal Signs in 2005, remains the standard map of the methodological options. McCall distinguished three stances toward the categories the framework uses. The anticategorical stance treats the categories themselves as products of power and seeks to dissolve them, regarding any analysis that sorts persons by race, class, and gender as complicit in the sorting it studies. The intracategorical stance interrogates a single neglected location in depth, the compound position in its full texture, in the manner of the founding legal cases. The intercategorical stance retains the categories provisionally in order to document the patterns of inequality across them, accepting the categories as instruments while disclaiming their naturalness.

Each stance operationalises a different version of the theory, and the operationalisation exposes what programmatic statements conceal. The intercategorical stance, which is the one quantitative research can perform, slides toward the additive picture whenever its instruments are crude, because a statistical model that estimates the separate contribution of each category and then their interaction has rebuilt the list inside the mathematics: the categories must be specified in advance, treated as attributes of persons, and combined by addition and multiplication of their effects. The intracategorical stance, by contrast, preserves the constitutive insight at the price of generalisation, since the texture of one location, however deep, does not by itself yield an account of the structure. McCall’s own conclusion, that the stances are complementary instruments rather than rivals, is the methodological form of the finding this article reaches by the historical road: the versions are answers to different questions, and the quarrel among them is sustained by pretending that one instrument must do every job. For the debate with Marxism, the methodological literature adds a point neither side’s manifestos make, that the relational account of class is itself difficult to operationalise, since relations do not sit still to be coded as variables, and some of what looks like intersectionality’s vagueness and some of what looks like Marxism’s retreat into philosophy are the same difficulty, the resistance of structures to the instruments built for attributes, met from opposite directions.

Why is the concept hard to measure?

Because measurement instruments are built for attributes, and the constitutive version of the theory is about relations and mutual formation. A statistical model must fix its categories in advance and treat them as properties of persons, which rebuilds the additive picture inside the mathematics.

The difficulty is real, it is shared by every structural theory including the Marxist account of class, and it explains much of the vagueness charged against the field.

The qualitative literature answered the difficulty with a discipline of its own, and its findings repay attention because they are the empirical core the polemics float above. Studies conducted in the intracategorical manner, of domestic workers organising across racial divisions, of women in racially segmented labour markets, of migrants whose legal status cuts across every other category, repeatedly find that the compound position is not an accumulation but a distinct organisation of constraints, in which each structure alters the operation of the others: a labour market that reserves certain jobs for women of one racial group creates a class fraction whose experience cannot be reconstructed by adding a race effect to a gender effect to an employment effect. The finding is modest, cumulative, and almost never cited by either side’s polemicists, for the polemical reason that it supports the constitutive version, which inconveniences both the additive populariser, whose list it refutes, and the Marxist objector, whose grammar argument it bypasses. A reader who wants to know what the theory looks like when it is working should read this literature before the manifestos, and the linked study materials, including the VaultBook study notes prepared for this series, are organised to make that order of reading easy to keep.

The standpoint strand and its limits

Running through the whole tradition, from Jones through the Combahee statement to Collins, is an epistemological claim that needs separate treatment because it attracts a misunderstanding that damages both the claim and its critics’ aim. The claim is that social structures are differentially visible from different positions within them, and specifically that the compound position, where several structures press at once, affords a view of their joint operation unavailable from positions where only one presses, or where the pressure is cushioned. In its defensible form the claim is about evidence and attention: a worker whose life is organised by the interaction of a racialised labour market, a gendered division of household labour, and an immigration regime will have occasion to notice the interactions, because she must navigate them, in a way that a theorist modelling one structure at a time may not. Jones made exactly this claim in 1949 when she argued that the most oppressed would prove the most militant and the most clear-sighted about the system as a whole, and she derived the clarity from the position’s demands, not from the occupants’ essence.

The misunderstanding converts a claim about the distribution of evidence into a claim about the authority of utterance: that an analysis is validated by the position of its author, and criticism of it answered by locating the critic. That conversion is a corruption of the standpoint strand by the attribute view of identity, and it has the same structure as the corruption of class examined earlier: a relation to a structure, which supplies questions and evidence, is flattened into a property of a person, which is then treated as a credential. The founding writers resist the conversion explicitly. Collins insists that a standpoint is achieved through collective critical work on a location, not conferred by occupancy; the Combahee statement grounds its politics in a material analysis its members expected to argue for, not merely to announce. Marxist critics who attack the credential version are therefore attacking a position the tradition’s major texts disown, while defenders who practise the credential version are betraying the texts they cite. The symmetry is the article’s recurring finding in a new key, and it suggests that the debate’s real frontier is not between the two traditions but between the structural and attributive readings available inside each.

There is a reciprocal lesson for the Marxist side, which possesses its own standpoint tradition and its own corruption of it. The claim that the proletarian standpoint affords knowledge of the social whole, developed by Lukács in the text examined elsewhere in this series, suffers the identical flattening when the standpoint of a class is treated as the property of whoever claims to speak for it, and the history of parties that converted a theory of structural visibility into a certificate of their own authority is the political form of the error. A debate conducted with the distinction in hand would find the two traditions auditing the same failure in themselves and in each other, which is a more useful exercise than the exchange of denunciations, and it is available to any reader who keeps the relation and identity distinction in view.

Does occupying a position confer the standpoint?

No. In the founding texts of both traditions a standpoint is an achievement of collective work on a location, not a property conferred by occupying it. A location supplies evidence and questions, because those who live a structure’s crossings must navigate them; the analysis still has to be made, argued for, and exposed to correction.

The credential version, in which an author’s position validates an argument and answers its critics in advance, is therefore a corruption that the texts disown, and it is the attributive error in epistemological dress: a relation to a structure, flattened into a property of a person and then spent as authority.

Coalition and unity: the political form of the question

The analytical dispute has a political form, and stating it completes the article’s map, because the two traditions’ theories of how emancipatory forces combine differ as their theories of structure differ. The classical Marxist form is unity: a class, divided in experience, is to be unified by the recognition of a common position and a common opponent, and divisions of race and sex figure in that project as obstacles to the recognition, to be overcome by the demonstration that they serve the other side. The intersectional form is coalition: constituencies whose positions genuinely differ, and whose interests partly diverge, combine on terms that keep the differences on the table, because a unity purchased by suppressing the compound claims will be paid for, in the Combahee statement’s finding, by the defection or the silence of those whose claims were suppressed. Crenshaw’s political intersectionality is the coalition form stated as doctrine: a politics adequate to people situated between agendas must be built as a negotiation among them, not as the subsumption of any under the most general.

Each form has its characteristic failure, and the failures are the historical record once more. Unity’s failure is the schedule: the difference that cannot be recognised without breaking the unity is deferred, and the deferral compounds until the unity is a shell held by those it served. Coalition’s failure is the veto and the inventory: a combination in which every constituency’s claims must be acknowledged before any can be advanced can dissolve into the additive list at the level of organisation, a programme assembled by enumeration, in which the mechanism that joins the claims, and might rank or transform them, is never faced. The packinghouse union examined earlier is instructive precisely because it found, in practice, a third shape: a unity built by fighting the divisions as class questions inside the common organisation, so that the compound claims were neither suppressed nor merely listed but made into the organisation’s work. Whether that shape generalises beyond its industry and decade is an open historical question, and it is the political form of the article’s analytical finding: the strong versions of the two traditions describe the same organisation from different ends, and their slogans describe its two failures.

How does a coalition differ from a unity?

A unity presupposes a common position beneath the differences and asks the differences to dissolve into it. A coalition presupposes that the differences are real and partly divergent, and builds the combination by negotiation among them, keeping each claim on the table. The first risks the schedule; the second risks the inventory.

The difference matters for the reading of the debate because the camps’ political accusations are usually the analytical ones in this dress. The Marxist charge that intersectional politics fragments the movement is the coalition failure described from the unity standpoint; the intersectional charge that class politics dissolves its allies is the unity failure described from the coalition standpoint. Both failures are documented, both charges are therefore partly true, and the organisations that escaped them did so by the discipline this article has recommended throughout: they treated the compound claims as the test of the combination rather than as its obstacle, and they built mechanisms, grievance procedures, caucuses with power, programmes aimed at the joints of the structure, through which the test could be applied continuously rather than at the moment of crisis.

The dispute that remains: explanatory priority

Once the additive version is set aside, the compound claims are restored to their structures, and the standpoints are returned to their evidence, one dispute survives intact, and intellectual honesty requires stating it as a dispute rather than dissolving it into the general reconciliation. It concerns explanatory priority. The Marxist tradition, in its strong form, claims that the organisation of production is not merely one structure among others but the structure that explains the historical career of the others: why racial division takes the form of a segmented labour market in one period and of a colour bar in another, why the gendered division of household labour is reorganised when the wage relation expands, why forms of subordination that long predate capitalism are recruited, reshaped, and sometimes dismantled by its requirements. The claim is not that production causes attitudes, a vulgarity no serious writer holds, but that the reproduction of the economic order sets the constraints within which the other structures operate and supplies the mechanisms by which they are rebuilt.

The constitutive intersectional position, in its strong form, declines the priority claim. On Collins’s matrix, the structures are mutually constituting without a designated centre, and the request for a centre is itself suspect, as the residue of a theory that habitually finds its own favourite structure at the bottom of every pile. Between these positions there is a real difference of explanatory strategy, with real consequences: the priority claim licenses a research programme that traces each form of subordination to its function and fate within accumulation, and predicts, sometimes wrongly, that subordinations obstructing accumulation will be eroded by it; the matrix claim licenses a programme that maps configurations without presuming their hierarchy, and predicts, sometimes wrongly, that no single transformation will dissolve the configuration. The historical evidence can be mustered honestly on both sides. The erosion of formal racial bars under the pressure of labour market integration supports the first programme; the persistence and mutation of racialised outcomes through every transformation of the economic order supports the second. The persistence of the gendered division of household labour through both capitalist and socialist industrialisations supports the second; its dramatic reorganisation under the pressure of women’s mass entry into wage labour supports the first.

This residue is not a scandal, and the article’s refusal to award it to either side is a finding, not a failure of nerve. It is the normal condition of a live research frontier, and it resembles the unresolved disputes inside each tradition, between Marxists over the motor of history, between intersectional theorists over the status of their categories, more than it resembles the war of slogans conducted in the frontier’s name. What the article claims to have established is the location of the frontier. Readers who reach it, equipped with the legal origin, the antecedents, and the two versions, are in a position to read the serious literature of both traditions as contributions to a shared problem: how structures of subordination are joined in fact, and whether the joining has a centre. Readers who skip the recoveries will continue to meet the slogans, and the slogans will continue to deserve each other.

Claudia Jones and the founding essay

The antecedent deserves a fuller hearing than the summary given earlier, because the essay’s argument is more exact than its reputation and because its recovery by scholars is itself part of the debate’s history. Claudia Jones wrote the essay from a specific location: she sat on the party’s national committee, responsible for work among women and among Black Americans, writing in the party’s theoretical journal at a moment when the party’s influence was contracting under state persecution. The essay is therefore an internal memorandum as much as a public argument, and its tone, exhortatory and indicting by turns, is that of an organiser reporting a failure to the institution responsible for it.

Its economic core is a claim about the labour market. Black women, Jones documented, were concentrated in domestic service and agricultural labour, the two sectors least protected by the labour legislation of the period, least organised by unions, and worst paid in absolute terms. Their earnings stood below those of Black men, which stood below those of white women, which stood below those of white men, an ordering she treated not as a curiosity of prejudice but as a structure with functions: the confinement of Black women to the cheapest labour subsidised the households and farms that employed them, and the availability of that labour disciplined the wages of everyone else. She used the vocabulary of her tradition, calling the arrangement super-exploitation, exploitation beyond the standard rate because the worker’s bargaining position was undermined from three directions at once. The analytic point survives translation out of the vocabulary: the compounding she described was produced by institutions, employers’ hiring practices, unions’ exclusions, legislation’s exemptions, and not by the sum of three attitudes in the minds of the powerful.

From the economic core the essay drew its political conclusions, and they were aimed in three directions. Against the party and the unions, Jones charged a practical neglect dressed as principle: the specific organisation of Black women was perpetually postponed on the ground that general organisation would reach them in due course, a deferral whose consequence was that the most exploited layer of the class remained the least organised. Against the women’s movement, she charged that its demands were framed for women whose problem was exclusion from the professions, while Black women’s problem was super-exploitation within employment they already performed. Against a certain complacency in Black politics, she argued that the liberation movements of her people treated their women as auxiliaries. The triple direction of the critique is the essay’s most modern feature: no camp is treated as the compound position’s natural home, and the compound position is not treated as a sum of the camps’ concerns.

Jones then made the move that connects her essay to the standpoint strand examined above. The layer that bears the compounded structure, she argued, is positioned to lead the struggle against it, because its members cannot mistake any single axis for the whole: the Black woman worker who wins a race claim and remains super-exploited as a woman and a worker has been given an education in the limits of single-axis remedies that no pamphlet can supply. The argument is a claim about political epistemology, that clarity about a structure is distributed by the structure itself, and it predates by decades the academic formulations for which standpoint theory became known. That the essay’s author was a communist organiser, making the argument in a party journal with the tradition’s own categories, is the fact that the later framing of intersectionality as an external criticism of Marxism cannot digest. The criticism was internal first.

The essay’s afterlife completes the point. Jones was imprisoned and then deported in the anti-communist campaigns of the nineteen fifties, and her work dropped out of the American record, to be recovered by scholars generations later: Carole Boyce Davies’s study Left of Karl Marx, published in 2007, restored Jones to the history of Black radical thought, and Erik McDuffie’s Sojourning for Freedom, published in 2011, documented the tradition of Black left feminism to which the essay belonged, a tradition whose members developed the compound analysis, under names like triple oppression and double jeopardy, inside and around the communist movement of the nineteen forties and fifties. The recovery matters for the debate because it rewrites the chronology on which both sides rely. If the compound analysis was being made, with Marxist tools, in 1949, then the story in which intersectionality corrects Marxism from outside in 1989 is false, and so is the story in which the correction was always available and nothing was neglected. What the chronology actually shows is a tradition that produced the analysis, marginalised its producers, forgot the analysis, and was then confronted with it again in legal form: a sequence that the intersectional objection, the charge of documented deferral, describes with precision, and that no amount of definitional care on the Marxist side can argue away.

The longer prehistory

Behind Jones stands a nineteenth-century record that the scholarly literature has assembled piece by piece, and a sketch of it protects the article’s genealogy from the charge of inventing a lineage. When Sojourner Truth addressed the women’s rights convention at Akron in 1851, in the speech transmitted to posterity through Frances Gage’s later transcription, her argument proceeded by exhibiting a life that the convention’s categories could not parse: she had ploughed, planted, and gathered into barns, she had borne children and seen them sold, and she asked, in the refrain the transcription made famous, whether she was not a woman. Whatever allowances are made for the transcription’s dialect and mediation, the argument’s form is unmistakable and it is the compound form: the category woman, as the movement used it, had been built from lives unlike hers, and the exhibit of her life was a counterexample that the category had to absorb or confess. The form recurs because the problem recurs, wherever an emancipatory category is generalised from its most audible members.

Maria Stewart, lecturing in Boston in the early 1830s, two decades before Akron, addressed Black women directly and located their degradation in their confinement to domestic service, urging education and collective organisation as the remedy: an analysis in which race, sex, and occupation are already a single knot. Frances Ellen Watkins Harper, poet and lecturer of the later century, made the priority argument against the suffrage movement’s scheduling: when the white women of the reform world spoke of rights, she observed, they meant their own, and she refused the choice between her race and her sex that the movement’s leadership tried to impose at the moment of the Reconstruction amendments. Anna Julia Cooper gave the standpoint its first systematic statement in A Voice from the South in 1892, arguing that the Black woman, confronting both a race problem and a sex problem, was the only figure positioned to testify about both, and that the condition of a civilisation could be measured by the condition of those its categories served worst. Cooper’s book is a work of philosophy, complete with a theory of value, and its presence in the genealogy refutes the notion that the compound analysis waited for the twentieth century’s institutions to become thinkable.

What the twentieth century added was not the insight but its successive formalisations, each adapted to a new institutional setting. Jones formalised it for the labour movement, in the vocabulary of exploitation. The Combahee Collective formalised it for the feminist and Black liberation movements of the nineteen seventies, in the vocabulary of interlocking systems. Crenshaw formalised it for the law, in the vocabulary of doctrine and remedy. Collins formalised it for sociology, in the vocabulary of the matrix. The series of translations is itself evidence about the insight’s robustness: it survives changes of idiom because it answers a structural feature of emancipatory politics, the tendency of categories to be set by their least burdened members, which reasserts itself in each new institution that takes the categories up. A genealogy of this shape also disciplines the claims made for any single formalisation. The legal version is not the theory’s essence, nor is the sociological one; each is an instrument built for a site, and the debate goes wrong when an instrument’s local features, the law’s preoccupation with comparison classes, sociology’s with measurable attributes, are mistaken for the insight itself and criticised, or celebrated, in its place.

The theory side of the historical charge

The intersectional objection has so far been documented at the level of movements and institutions, but its full strength includes a claim about theory, and fairness requires the claim to be faced. The charge is that classical Marxist theory itself, and not merely its organisational custodians, supplied the habits that produced the deferrals: that the tradition’s master narrative, in which history’s motor is the conflict of classes defined by production, structurally invites its holders to treat other divisions as epiphenomenal, as forms in which the real conflict appears and through which it will be resolved. On this reading, the deferral pattern is not a betrayal of the theory but its faithful execution, and the remedy would have to be theoretical, a rebuilding of the framework, rather than a change of personnel.

The evidence within the founding texts is genuinely mixed, and the article reports it mixed. Against the charge stands the method itself: Marx’s mature work analyses the concrete forms through which the capital relation is lived, and the chapters on the working day, on primitive accumulation, and on the Irish question show divisions of nationality, and the colonial relation, treated as mechanisms internal to the system’s operation, not as decorations on it. His political judgements moved with the analysis, most famously on Ireland, where he reversed his earlier view and concluded that the English working class would accomplish nothing until it had separated itself from the oppression of Ireland, because a people that oppresses another forges its own chains: a dictum that makes a supposedly secondary division into the precondition of the primary struggle, and so refutes, in the founder’s own practice, the scheduling logic attributed to the theory. For the charge stands the tradition’s other inheritance: the evolutionary scaffolding of nineteenth-century anthropology that Engels built into his account of the family, the confidence of the movement’s popularisers that economic development would dissolve subordinate questions in sequence, and the base and superstructure metaphor in its crude form, which generations of readers took as a licence to rank phenomena by reality and to attend to the unreal ones later.

The honest verdict, which the strongest writers on both sides approach, is that the theory contains both resources and both temptations, and that which of them governed was decided historically, by institutions under pressure, rather than logically, by the texts. That verdict is unsatisfying to partisans because it denies them a clean weapon, but it has the merit of fitting the record, in which the same canon authorised Jones’s compound analysis in 1949 and the deferrals Jones was protesting. It also locates the practical question correctly. If the theory underdetermines the practice, then the debate’s stakes are not settled by exegesis, and both traditions are returned to the evidence: the Marxist side to demonstrate, in its analyses, that the relational account of class can hold the compound phenomena without a list; the intersectional side to demonstrate that its constitutive version can explain, and not only exhibit, the configurations it maps. The demonstrations are the research programmes described in the section on explanatory priority, and the debate is best understood as the competition between them, conducted, at its best, in a shared evidentiary court.

The answers assembled

The Marxist objection was stated at full strength earlier, and the integrity of the exercise requires its answer to be assembled with equal explicitness, point by point, rather than left to accumulate by implication. To the conversion argument, that listing class dissolves the relation, the answer is that the objection holds against the additive version and that the constitutive version, which is the version of the founding texts, performs no listing: it claims that structures interlock, a claim about the world, not an inventory of a person’s attributes, and a relational account of class can be one of the interlocking structures without ceasing to be a relation. To the aggregation problem, that the framework cannot explain why one division organises the others in a given setting, the answer is a concession and a counterclaim: the concession is that intersectionality as such supplies no such principle, and the counterclaim is that it was never designed to, since its founding question was not why configurations take their shapes but why remedial categories fail the people at the crossings; the explanatory principle must come from a theory of the society, and the synthesis literature’s proposal is that Marxism is that theory, an offer the objection’s holders are free to refuse but not to describe as unmade. To the absence of a theory of production, the answer is the same in structure: the framework is a corrective to frameworks, a test that any theory of a stratified order must pass, namely that its categories not dissolve the compound cases; a test is not refuted by the observation that it is not itself a theory, any more than a proofreading is refuted by not being a novel. To the direction of remedy, the answer concedes the most: the popular career of the term has indeed issued in demands for recognition and representation more often than in demands about ownership, and the founding legal argument, being about the repair of doctrine, supplies no resistance to that career. The concession is the price of the distinction, and this article pays it deliberately: the additive and institutional version of intersectionality does what Mitchell’s objection says it does.

What the exchange establishes, when both the objection and the answer are put at full strength, is narrower than either side’s publicity. The Marxist objection defeats the additive version and the institutional usage. The intersectional answer defends the constitutive version and the founding argument. Between the defeated usage and the defended theory lies the actual territory of the term’s life, most of it closer to the usage than to the theory, which is why the debate’s temperature is so badly calibrated: the critics are describing the territory, the defenders are describing the texts, and both descriptions are accurate. A reader equipped with the two versions table can perform the calibration in any particular case, asking of any deployment of the term whether it is the theory or the usage that is on the page, and grading the objections accordingly. That reader is the one this article has been written to produce.

The intersectional objection, being historical, is answered by evidence rather than by distinction, and its answer has been given in the sections on the record. But it deserves the same explicit assembly. The answer cannot be that the deferral pattern did not happen; it did, and the documents are quoted above. The answer that remains available to the Marxist side is threefold. First, that the pattern is a tendency of movements under the pressure of coalition politics generally, not a specialty of Marxist ones, as the feminist and civil rights movements’ own internal records of deferral, of women by race movements and of race by women’s movements, attest; the compound position was scheduled by every movement it belonged to, which is the Combahee statement’s own testimony. Second, that the Marxist tradition uniquely generated, from within itself, both the diagnosis of the pattern and the theoretical instruments against it, in Jones’s essay, in the social reproduction literature, and in the self-criticism that runs through its feminist writing. Third, that the alternative traditions have their own archives to answer for, and that the comparative audit of movements, as opposed to the audit of one movement, is the exercise the evidence supports. None of this acquits the record. It converts the objection from a verdict on a tradition into a finding about emancipatory organisation as such, a finding from which the compound analysis itself historically grew.

The internal critics

A debate is healthiest when its participants criticise their own instruments, and the intersectional literature contains an internal criticism of the first rank that any honest guide must include, because it disciplines the framework from the position of its founding constituency. Jennifer Nash’s Black Feminism Reimagined, published in 2019, argues that the academic field built around intersectionality has developed a defensive posture toward its founding figures and texts, a posture Nash calls intersectional originalism: the treatment of the founding legal articles as a sacred origin whose correct interpretation settles present disputes, policed by the field’s custodians against rival claimants, including other Black feminist traditions that predate or bypass the legal formulation. Nash’s charge is striking because it comes from within Black feminism and is aimed at the field’s institutional habits rather than at the concept: a framework that began by criticising how categories dissolve inconvenient cases had built, she argues, a custodial practice that dissolves inconvenient genealogies, treating Jones and the Collective, among others, as anticipations of the founding text rather than as theorists in their own right.

The relevance to the present debate is direct. The genealogy recovered in this article, in which the compound analysis descends through Jones and the Collective and is formalised, not founded, in 1989, is the genealogy Nash’s criticism makes room for, and it is the genealogy least convenient to both polemical camps: to the intersectional custodians, because it dissolves the founding moment into a longer Black feminist and Marxist history; to the Marxist objectors, because it shows the analysis arising inside their own tradition’s institutions and journals. Nash pairs the criticism with an ethic she calls love politics, a demand that the field hold its founding figures with affection and criticism at once, and the pairing generalises into the method this article has tried to practise toward both traditions: attribution instead of ventriloquism, the strongest form instead of the convenient one, and the documentary record instead of the camps’ memories of it.

Sirma Bilge’s work supplies a second internal criticism, aimed at the framework’s institutional career. In a widely cited article of 2013 on the whitening of intersectionality, Bilge argued that the concept’s uptake in the European and North American academy had systematically stripped it of its commitments: its origin in Black feminist critique, its orientation to social justice as transformation rather than inclusion, and its attention to the material structures of neoliberalism, leaving a depoliticised methodology of diversity that could be practised, in her phrase’s implication, without disturbing anything. The criticism converges with Mitchell’s from the opposite camp: a Marxist feminist and an intersectionality scholar agree that the institutional usage betrays the founding argument, disagreeing about what the founding argument, kept whole, would imply. Convergence of that kind, between critics who share no camp, is the most reliable signal available to a reader that the object of criticism, the additive and administrative version, is genuinely there to be criticised, and that the debate’s defenders waste their strength when they answer the signal as though it were an attack on the texts.

Why the additive version spreads

If the additive version is the weakest form of the theory, its dominance in circulation requires explanation, and the explanation is instructive about how concepts live outside books. The additive picture has three advantages in transmission, none of them intellectual. It is pictorial: burdens can be drawn, stacked, and counted, and a picture survives compression into a slide, a poster, or a paragraph in a way that a claim about mutual constitution does not. It is administratively legible: an institution that must demonstrate its attentiveness can enumerate the axes it has attended to, and the enumeration is auditable in a way that an account of how its own structures interlock is not. And it is morally undemanding of its user: to acknowledge accumulated burdens requires no decision about their causes and no commitment about remedies beyond inclusion, whereas the constitutive version, by asking how the structures produce one another, points at mechanisms somebody owns. A version of a theory that can be drawn, audited, and affirmed without consequence will out-replicate a version that cannot, in any environment that selects for those properties, and the environments in which the term now mostly circulates select for exactly those properties.

The Marxist tradition has no standing to find this process exotic, since its own concepts underwent the same selection. The base and superstructure metaphor, a heuristic Marx used sparingly and Engels qualified in his late letters, became in transmission a diagram of two storeys that generations of students met as the theory itself, and the diagram’s defenders and attackers alike spent decades on a construction the texts support only as a rough first approximation. Class itself suffered the fate the objection describes: converted in ordinary usage from a relation into a stratum, a band of income or a style of life, long before intersectionality existed to be blamed for the conversion. The sociology of concepts therefore supplies a shared premise the debate lacks: every structural theory has a weak version, produced by the mechanics of transmission, and a polemic that cannot distinguish the weak version from the theory will always find targets and never find the argument. The discipline of distinguishing them, practised symmetrically, is most of what this article means by reading the debate well.

How can a reader tell which version is in use?

Apply the substitution test. Replace the term’s appearance with the claim it is making in that sentence. If the claim survives as an enumeration, that several disadvantages coincide and should all be acknowledged, the additive version is in use here.

If the claim is about a mechanism, how one structure shapes the operation of another in the case at hand, the constitutive version is in use. The theory is the second; the first is its travelling shadow.

The grammar of the list in practice

The conversion of class into an attribute is easiest to see at the level of the sentence, and a short exercise in grammar does more to fix the distinction than another round of definitions. Consider the sentence pattern that recurs in policy documents and movement literature alike: the programme, it is said, must serve working-class women of colour, naming a constituency by the accumulation of its descriptors. Nothing is wrong with the sentence as politics; constituencies must be named to be organised. But notice what the grammar does to the analysis that might follow. The descriptors present themselves as parallel properties of the constituency, and the analysis is silently commissioned to proceed property by property: a measure for the class property, a measure for the race property, a measure for the sex property, each administrable by the office that owns that axis. The relational content of the first descriptor, that the constituency’s members stand in a structured relation to employers whose interests the programme may have to oppose, has been grammatically demoted to one characteristic among three, and no measure for it appears in the programme, because programmes have no office for relations. The additive version is not, in such cases, a theory anyone chose. It is the default setting of the administrative sentence, and it reasserts itself in every document that nobody wrote with the distinction in mind.

The constitutive analysis writes different sentences, and their shape is recognisable. It says that the labour market of a given city allocates migrant women to night cleaning and domiciliary care through the joint operation of immigration law, the gendered definition of care as unskilled, and employers’ preference for a workforce whose residence rights depend on employment, so that the class position in question, night cleaner on a precarious contract, is manufactured by the three structures in one process. There is no list in the sentence, because the phenomenon does not decompose: remove the immigration regime and the job changes hands and character; remove the gendered definition of the work and its wage and status change; remove the employer’s preference and the allocation dissolves. Analysis of this form is what the founding legal argument was reaching for, since the seniority system in the DeGraffenried case likewise manufactured a compound position in one process, and it is what Jones was reaching for in describing a labour market that confined Black women to its least protected sectors. It is also, not incidentally, the form of analysis that classical Marxism at its best produced about its own central cases, the factory, the colony, the household. The two traditions’ strong versions share a grammar, and the shared grammar, more than any negotiated platform, is the ground of their possible synthesis.

Double jeopardy and the organisational line

Between Jones’s essay and the Combahee statement, the compound analysis lived in organisations whose history has only in the last decades been written, and the middle chapter of the genealogy deserves its place, since it shows the analysis doing organisational work rather than awaiting its theoretical formulation. In 1968 and 1969, Black women within the Student Nonviolent Coordinating Committee formed a caucus that became the Black Women’s Alliance and then, widening its membership, the Third World Women’s Alliance. Frances Beal, a founder of that current, published in 1969 the pamphlet Double Jeopardy: To Be Black and Female, a title that fixed the period’s vocabulary for the compound position. Beal’s argument ran on the now familiar rails with an organiser’s concreteness: Black women were exploited as workers in the lowest-paid sectors, subordinated as women within their own movement, where the demand that they defer to male leadership was being dressed in the language of racial unity, and targeted as Black people by the state and the labour market together. Her pamphlet is also an early site of the debate’s permanent tension, since Beal wrote as a socialist for whom the economic analysis was primary in method while refusing, in practice, every attempt to schedule the other claims behind it.

The Alliance’s newspaper, Triple Jeopardy, took its name from the threefold structure Jones had named in 1949, with imperialism in the third position, and the continuity of the vocabulary across the decades is the genealogy in miniature: triple oppression in the party journal, double and triple jeopardy in the movement pamphlets, interlocking systems in the Collective’s statement, intersectionality in the law reviews. Each renaming adapted the analysis to a new arena and a new opponent, and each retained the core refusal, that the compound position would not be decomposed into a primary claim and a residue. The scholarly recovery of this organisational line, in works such as McDuffie’s history of Black left feminism and Dayo Gore’s Radicalism at the Crossroads of 2011, has a polemical consequence the debaters have been slow to absorb. The compound analysis was not a seminar-room invention of the late twentieth century but the working doctrine of organisers across four decades, most of them formed in or near the Marxist movement, and both camps in the present quarrel are its heirs: the intersectional camp by direct descent, the Marxist camp by the descent it declined to acknowledge when it let the tradition’s custodians be deported, marginalised, and forgotten.

The domains of power

Collins’s later reformulation of her own framework supplies a vocabulary that helps to state the constitutive version with operational precision, and it deserves inclusion because it is the version most often taught. In the second edition of Black Feminist Thought, published in 2000, the matrix of domination is analysed through four interrelated domains of power. The structural domain organises oppression through the large institutions of a society, its labour markets, legal systems, and schools. The disciplinary domain administers it, through the rules, routines, and bureaucratic practices by which institutions manage populations. The hegemonic domain justifies it, through the ideologies and images that make the arrangement appear natural or deserved. The interpersonal domain enacts it in everyday interaction. The scheme’s analytical value is that it prevents a characteristic slippage: an analysis that locates oppression only in the hegemonic domain, in images and attitudes, can be answered by an institution that reforms its imagery while leaving its structures untouched, while an analysis that keeps all four domains in view must specify, for any claimed oppression, where it is organised, how it is administered, by what story it is justified, and how it is lived.

The domains scheme poses the priority question, where the Marxist dispute is concerned, in a usefully concrete form. A Marxist reading of the scheme accepts the four domains and asks which of them sets the terms for the others: whether the structural domain, in which the organisation of production sits, does not in practice govern the disciplinary routines, supply the hegemonic stories with their plausibility, and script the interpersonal encounters. A Collinsian reading answers that the domains are interrelated without a governor, and that the appearance of governance is an artefact of asking the question with a hierarchy already in mind. The exchange is the explanatory priority dispute at working scale, and it shows why that dispute will not be settled by definition: it is a disagreement about how social orders hold together, on which the evidence is the whole of social history, read through rival habits of attention. What can be settled, and what teaching should insist on, is that the dispute is this one, a dispute between a centred and a decentred model of mutual constitution, and not the cartoon in which one side acknowledges only class and the other acknowledges everything except class.

The founder’s own clarifications

Crenshaw has returned to the concept repeatedly in later writing and interviews, and her clarifications form a useful control on the debate, since they show the founding author resisting both of the popular deformations. She has described intersectionality as a provisional concept and a heuristic, a lens for seeing where categories fail, rather than a completed theory of society or a doctrine to be applied. She has resisted the reading of the framework as a theory of identity, remarking in substance that it was always about structural and political dynamics, about how systems converge in institutions, and that its reduction to a synonym for diversity of persons is a fate that befell it in transit. And she has resisted, from the other direction, the custodial use of the founding texts as a settlement of disputes the texts did not address, the originalism Nash criticised. The three resistances line up with the distinctions this article has drawn: against the additive reading, against the attributive reading, and against the treatment of a corrective instrument as a closed system.

The clarification about the concept’s provisional status deserves emphasis, because it answers a version of the Marxist objection that the answer section above left standing. If intersectionality is offered as a test that theories of stratification must pass, a check against the dissolution of compound cases, then the objection that it is not itself a theory of capitalism, while true, is addressed to a claim its founder does not make in that form. The objection retains its force against the many deployments that do make the claim, in which intersectionality is presented as a self-sufficient account of the social order, and it retains its force against the institutional usage. But the debate’s centre of gravity shifts once the founder’s own framing is admitted as evidence: the question becomes whether the test is a good one, and whether a Marxism that passes it, as the social reproduction synthesis attempts to show, is weakened or strengthened by the passing. That is a question the two traditions can investigate jointly, which may be why it is so seldom asked in the polemical literature, whose economy runs on the other kind.

A term of abuse, described

The phrase identity politics shadows this entire debate, and the neutrality this article has promised requires its usage to be described rather than adopted. The expression entered the literature in the Combahee statement of 1977, where it named the Collective’s own method: a politics whose warrant was the members’ material position and whose starting point was their own lives as Black women. In that usage the phrase carried no pejorative charge and implied no indifference to structure; the Collective’s socialism was stated in the same document. Over the following decades the expression migrated, and its dominant usage inverted: it came to name, in the mouths of its critics across the political spectrum, a politics alleged to substitute the recognition of group identities for material redistribution and universal programmes. Between the two usages there is no stable referent, and most arguments that employ the phrase trade on the slide between them, the descriptive sense lending the phrase its documentary air, the polemical sense doing the work.

For the present debate the slide has a specific cost. When the Marxist objection is phrased as the claim that intersectionality is identity politics, the argument’s analytical content, the conversion of a relation into an attribute, is packed into a label whose polemical sense pre-judges it and whose original sense contradicts the packing, since the phrase’s coiners were socialists making a compound class analysis. The objection is stronger without the label, as Gimenez’s and Mitchell’s best passages show, because it then has to specify the mechanism it alleges and can be answered at the level of mechanism. The same discipline applies in reverse: intersectional writing that dismisses class analysis as economism, without specifying which reduction is alleged and where it occurs, is performing the symmetrical slide. This article has therefore used the contested phrase only as an object of description, and it recommends the practice to the debate: contested labels are where arguments go to be replaced by allegiances, and a literature that wants its disagreements to be about the world should ration them.

The term’s public career

Outside the literatures this article has surveyed, the word leads a second life as a token in public argument, and the linked article on Marxism in political discourse traces the general pattern of which this career is a case. In public usage the term functions most often as a signal of allegiance or of accusation, deployed without reference to any text, and its semantic content thins to the gesture the methodological section described: on one side, an affirmation that many disadvantages exist and matter; on the other, a charge that an opponent cares about nothing else. The two public usages deserve each other, and neither survives contact with the founding argument, which was specific, remedial, and addressed to a documented failure of a legal framework. Journalists who must use the word, and the persona section below returns to their case, perform a service each time they restore one fact to it: that its first home was a law journal of 1989 and its first subject a rule about who is laid off last.

The public career also feeds back into the scholarly one, and the feedback explains some of the debate’s peculiar ferocity. A scholar defending the constitutive theory finds herself conscripted as the defender of the public gesture, and a Marxist criticising the gesture finds his argument received as a criticism of the theory, and of the constituency the theory arose from. Positions harden in proportion to the stakes imputed to them, and the stakes imputed to a term of public combat are always larger than the stakes of the analytical question it carries. The analytical question, whether class is a relation or an attribute and whether structures have a centre, is genuinely important and entirely unfit to bear the charge of the public quarrel. One function of a reference article is quarantine: to hold the analytical question in a space where it can be examined at the temperature it deserves. The historical method is the quarantine’s mechanism, since a debate that must pass through 1949 and 1976 and 1989 before it may proceed to its conclusions has already been slowed to the speed of evidence.

One case, two analyses

The difference between the versions can be made concrete by running a single case through each, and the founding case serves again, since its facts are settled. Take the seniority system at the plant in the DeGraffenried litigation: a workforce in which Black women had been hired only after 1964, a last-hired first-fired rule neutral on its face, and a recession in which the rule extinguished the jobs of Black women as a group. The additive analysis describes the outcome as the sum of a race burden and a sex burden: the women were disadvantaged as Black workers, because their hiring came late, and as women, because the jobs open to them were marginal, and the two disadvantages accumulated to produce their exposure. The description is true and insufficient, and its insufficiency is diagnostic. It cannot explain why the burdens arrived fused, why the specific women extinguished were not simply the most burdened individuals scattered through the workforce but a category produced as a category by the interaction of a hiring history and a seniority rule: the plant’s racial history determined when women of one group could be hired at all, and its gendered job structure determined which jobs they could be hired into, so the seniority rule fell on a position that the other two structures had already built for it to fall on.

The constitutive analysis begins from that interaction. It asks how the hiring history and the job structure jointly manufactured the exposed position, and it finds the compound not in the women but in the mechanism: a seniority rule is a class instrument, distributing security among workers by tenure; a hiring bar is a racial instrument; a sex-typed job ladder is a gender instrument; and the outcome was the product of the three instruments meshed, so that no reform addressed to one instrument alone, not the seniority rule’s amendment, not the hiring bar’s removal a decade too late, could have repaired the position. The analysis is also, notice, a piece of class analysis of a classical kind: it is about the distribution of security and insecurity among workers, the segmentation of a workforce, and the way segmentation is produced by the joint action of institutions. A Marxist who reads the constitutive analysis finds a familiar method at work on unfamiliar materials. An intersectional theorist who reads the additive analysis finds her own theory’s travelling shadow. The case is the debate in miniature, and it is why the article has insisted that the versions, and not the camps, are the units of the argument.

Does the additive version ever suffice?

For some purposes, yes, and the concession keeps the distinction honest. Where the task is remedial triage, the enumeration of the burdens a population carries, in order to direct provision toward the most burdened, an additive inventory is the right instrument.

Its crudity is a fair price for its speed. The error is not using the instrument but mistaking it for the theory, and then defending or attacking the theory on the instrument’s performance.

The evidence question

This debate is conducted almost entirely without numbers, and a reader schooled by the data articles in this series will want to know whether that is an evasion. It is, in the first instance, a property of the subject: the claims at issue concern the architecture of frameworks, the grammar of categories, and the historical record of movements, and none of these is settled by a coefficient. But the deeper reason deserves stating, because it is the contested figures problem in a new dress. The quantities that might seem to arbitrate the dispute, the share of an outcome attributable to class as against race or gender, are not measurements awaiting collection; they are outputs of the very frameworks under dispute. To decompose a wage gap into portions assigned to each axis, a model must first treat the axes as separable attributes with separable effects, which is the additive assumption; the decomposition therefore cannot be offered as evidence for the additive version without circularity. The constitutive version predicts, in advance, that such decompositions will be unstable across specifications, since the effect of each structure runs through the others, and the prediction is borne out often enough to keep the argument alive without ever being allowed to close it.

What evidence can do, and what the serious literature uses it for, is more modest and more valuable. Historical evidence disciplines the deferral dispute, which is a claim about what organisations did. Case evidence disciplines the legal dispute, which is a claim about what doctrines do to claimants. Ethnographic and institutional evidence disciplines the constitutive claim, by exhibiting mechanisms in which structures demonstrably manufacture one another’s forms, or fail to. Conceptual evidence, the exhibition of what a category does inside an argument, disciplines the conversion dispute. Each kind of evidence has its court, and the debate goes wrong when a verdict from one court is read into another: when the historical record of deferral is taken to refute the relational definition of class, which it does not address, or when the logical critique of the list is taken to refute the historical record, which it does not touch. Keeping the courts separate is the last of the article’s disciplines, and it may be the one with the widest application beyond this debate, since most interdisciplinary quarrels are, on inspection, two verdicts from different courts being read as one.

Four tests for any use of the term

The article’s instruments can be gathered into a short protocol for grading any use of intersectionality, in a text, a syllabus, a judgment, or a programme, and the protocol is offered as the practical residue of the whole argument. The first test is the version test, already given: substitute the claim the term is making, and determine whether it is an enumeration of coincident burdens or an account of mutual formation. The second is the relation test: ask what the use takes class to be, a position within a structure of production or a level of disadvantage, and note that a use may pass the version test and fail this one, since a writer can speak fluently of interlocking systems while treating every system, class included, as a distribution of burdens among persons. The third is the mechanism test: ask the use to name the institution through which the intersection operates in its case, an employer’s rule, a statute’s comparison class, a labour market’s segmentation, a household’s division of work. The founding texts always can; the gesture never can; and a use that cannot name a mechanism has not yet begun the analysis it advertises. The fourth is the remedy test: ask what would count, on the use’s own terms, as the problem’s solution. Recognition, representation, and parity of outcomes across the enumerated categories indicate the administrative version; a change in the mechanism the third test named indicates the structural one. Neither answer is forbidden, but the use should know which it is giving.

The protocol is deliberately double-edged, and it cuts Marxist usage as cleanly as any other. A Marxist text can be put through the same four tests with the terms transposed: does it treat race and gender constitutively, as forms through which the class structure actually exists, or additively in reverse, as effects to be acknowledged after the economic analysis is complete; does it treat them as relations and mechanisms or as cultural colouring; can it name the institutions through which they operate on its cases; and does its remedy schedule them, once again, for after the transformation of the mechanism it does name? The deferral pattern, restated as a grading protocol, is simply the repeated failure of the fourth test. A literature in which both sides graded themselves before grading each other would not end the dispute about explanatory priority, which is genuine, but it would retire the greater part of the published quarrel, which is not.

The misreadings, corrected in one place

Reference articles earn their keep by correcting the errors a reader is most likely to bring to them, and this debate has five that recur with enough regularity to deserve a consolidated correction. The first is that intersectionality was formulated as a general theory of society. It was formulated as an argument about the failure of anti-discrimination doctrine to process compound claims, and its generalisation into social theory was the work of the following decades, principally in sociology. The second is that the term means the addition of identities, a reading its founding argument contradicts in terms, since the compound harm of the founding cases was precisely what addition could not represent: the plaintiffs were not more disadvantaged than the comparison classes along a shared scale, they were positioned where the scale’s construction made their treatment invisible. The third is that Marxists reject the framework wholesale. The record shows Marxist feminists among its antecedents, its critics, and its synthesisers, often in the same bibliography, and the tradition’s internal argument about the framework is older than the framework’s name.

The fourth misreading is that class reductionism is a slur without a referent. It has a referent, documented above: a recurring institutional pattern in which race and gender claims were acknowledged, affirmed, and scheduled for later, across a century of movements that included Marxist ones prominently and others beside. To use the term without the archive is to weaken it; to deny the archive is to forfeit the argument. The fifth misreading is that the term has no Marxist antecedents. It has one of the best-documented antecedents in the series: a 1949 essay, in a communist party’s theoretical journal, by an organiser who grounded the compounded position of Black women in the way their labour was bought and sold, and who sent the bill for the movement’s neglect to the movement itself. Each of the five corrections is available from the texts and dates given in this article, and a reader who carries them into the debate will find that they dissolve, between them, most of the exchanges that pass for it in public.

What a synthesis costs each side

Since syntheses are on offer in the literature, their price should be stated, because a reconciliation that costs nobody anything is a slogan rather than a position. The intersectional side, in accepting the synthesis of the social reproduction and reconciliation literature, gives up the list: it accepts that class enters the analysis as a relation with a determinate structure, not as one attribute in an enumeration, and that the framework’s own contribution is the compound test and the constitutive claim, not a replacement theory of the social order. It gives up, with the list, the additive version’s administrative career, which is no small institutional sacrifice. The Marxist side gives up the schedule: it accepts that race and gender are not scheduled business, that the constitutive entanglement of the structures is a premise of its own method and not a concession to fashion, and that the historical record of deferral must be carried as a liability of its tradition and answered in each generation’s practice rather than rebutted as a misunderstanding. It gives up, with the schedule, the comfort of treating the compound cases as someone else’s specialism.

What neither side gives up is the residue. The Marxist keeps the priority claim, that production’s organisation explains the configuration’s historical forms; the intersectional theorist keeps the decentred matrix, in which the claim of a centre is itself interrogated. The synthesis literature lives inside that tension rather than abolishing it: Vogel’s unitary theory is a priority claim executed with constitutive care, Bohrer’s reconciliation is a matrix with a Marxist engine, and each is legible as a victory by the other side’s partisans, which is the usual sign that a synthesis is real. The debate’s future, if the literature’s best current work is the guide, lies in the competition of such syntheses on evidence, the mechanisms of particular labour markets, households, and legal systems, rather than in the defence of the slogans. That is a less dramatic future than the quarrel promises, and a considerably more useful one.

The unhappy marriage: the debate before the term

The argument this article has staged did not begin in 1989, and its rehearsal inside Marxist feminism a decade earlier explains the speed with which both sides recognised their positions when the term arrived. In an essay first published in 1979 and widely reprinted in the collection Women and Revolution in 1981, Heidi Hartmann described the relation between Marxism and feminism as an unhappy marriage, in which, adapting a formula from English family law, the two were one, and that one was Marxism: feminist questions were answered, when they were answered, by being dissolved into the categories of the partner theory. Hartmann’s remedy was the dual systems proposal, in which patriarchy and capitalism are distinct systems, each with its own material base, interlocking in practice. The essay is the intersectional objection stated from inside Marxist feminism before the vocabulary existed, complete with the archival charge, that the tradition’s categories had been sex-blind in their construction, taking the male worker’s situation as the situation of labour as such.

Dual systems theory is also, for this article’s purposes, the additive temptation arising inside Marxism itself, and recognising that fact disciplines any triumphalism about the list. To posit two systems, each complete with its own logic, and then to study their interaction, is to build the list at the level of systems rather than of attributes: the items are grander, but the grammar is the same, and the same difficulties follow, since the interaction of two self-sufficient systems is as mysterious as the addition of two self-sufficient burdens. The unitary theorists of the following years, Vogel above all, made exactly this criticism of Hartmann from the Marxist side, arguing that the dualism reinstated at the level of theory the separation that the analysis of social reproduction had to overcome. The debate between dual systems and unitary theory, conducted entirely within Marxist feminism in the nineteen eighties, is the template of the later debate with intersectionality, down to the distribution of the objections: the dual systems writer charges the unitary theory with dissolving gender into class, the unitary writer charges the dual systems with a list. When intersectionality arrived with its own version of the interlocking picture, the participants on the Marxist side had, whether they acknowledged it or not, been rehearsing their lines for a decade.

Iris Marion Young’s response to Hartmann, in the same 1981 collection, proposed the escape route that much later synthesis would take. Young argued that the remedy for a sex-blind theory of capitalism was not a second theory of a second system but a single theory whose central category, the division of labour, was gendered from the start: marginalisation, in her analysis, is a process by which capitalism constitutes categories of persons as secondary workers, and gender and race are the historically given forms the constitution takes. The proposal is the constitutive move executed inside Marxism before the constitutive vocabulary existed, and Young’s later work on structural injustice generalised its ethic: responsibility for structures attaches to all who participate in them, by position rather than by fault. A genealogy attentive to these rehearsals dissolves the last remnant of the two-camps picture. The debate of 1989 and after is the third staging of an argument the traditions had already had with themselves, and the positions were furnished, on both sides, before the term that now names the quarrel was coined.

Angela Davis and the suffrage archive

Davis’s Women, Race and Class has been cited above for its general argument; its documentary chapters deserve a paragraph of their own, because they show what the intersectional objection looks like when it is prosecuted as history rather than asserted as principle. Davis reconstructed the nineteenth-century American women’s movement’s entanglement with slavery and its aftermath: the abolitionist roots of the first feminists, the movement’s fracture over the Reconstruction amendments, when its leadership accepted and in some mouths demanded the deferral of women’s suffrage behind Black male suffrage while deploying racist arguments against the comparison, and the long consequence, a suffrage movement that by its final decades was courting the Southern states with the promise that white women’s votes would secure white supremacy. The archive is damning in detail, names and conventions and editorials, and Davis’s use of it is precise: the point is not that the suffragists were wicked but that a movement whose universal category had been built from a particular constituency behaved, at the moment of decision, as the category dictated. It is the single-axis failure in political form, documented forty years before Crenshaw named its legal form, by a writer working within the Marxist tradition and drawing the moral for all movements, her own included.

Her chapters on enslavement perform the complementary demonstration. Against the picture of the slave household as a matriarchy or a void, Davis argued that the enslaved community’s internal life, and women’s place in it, had to be understood through the labour regime that owned it: enslaved women worked as labourers on the same terms as men in the fields and were subjected, as women, to a sexual economy of coercion that was part of the property relation itself. The analysis is the constitutive method in historical practice, since the race relation, the property relation, and the sex relation are nowhere separable in the account, and each takes its specific form from the others. That a communist scholar produced the period’s most influential statement of both the objection and the method is the single most inconvenient fact in the debate for the two-camps framing, and this article has placed it at the centre for that reason.

Metaphors and their commitments

The debate’s three governing images repay a moment’s analysis, because each smuggles a version of the theory into the reader’s imagination before any argument begins. The intersection, Crenshaw’s own figure, pictures the structures as roads: built independently, crossing at points, with the traveller’s injury at the crossing attributable to traffic from several directions. The figure is constitutive about effects and agnostic about origins; the roads do not produce one another, and the analysis it sponsors asks who is hurt at the crossings and why the observer’s map fails there. The interlocking systems of the Combahee statement picture something closer to gears or to the fingers of folded hands: distinct elements whose operation is meshed, so that the motion of each is transmitted to the others. The figure sponsors a stronger claim than the intersection, since meshed elements condition one another’s movement, and a weaker one than identity, since the elements remain distinguishable. Collins’s matrix abandons the mechanical register for a spatial one: a field of positions in which every location is defined by its relations along every dimension at once, and in which the question of which road or gear came first cannot be posed.

The Marxist objection, translated into this register, is the demand for the engine: an account of what powers the machine, what sets the roads’ routes and the gears’ speeds, and its candidate answer is the accumulation of capital, which builds the roads. The intersectional answer, in Collins’s version, is that the demand for an engine is the residue of a mechanical picture the matrix has already left behind. The exchange of metaphors is not a substitute for the exchange of arguments, but it explains the arguments’ persistent mutual unintelligibility: the two sides are not disagreeing about the decoration of a shared picture, they are disagreeing about the picture, and the careful writer, this article included, does well to notice which figure she is inside at any moment. The two versions table was designed in the additive and constitutive vocabulary partly to escape the metaphors altogether, since versions can be tested against cases while pictures can only be preferred.

The ranking problem

A charge frequently laid against the framework in its popular career is that it licenses a competition of disadvantages, a ranking of persons by accumulated oppression in which authority follows the total. The charge deserves the article’s standard treatment: identify its target and grade it. Against the additive version the charge is substantially just, for addition invites comparison of sums, and a culture that has converted the framework into an inventory of attributes will predictably conduct the comparison the inventory affords. Against the founding texts the charge fails on inspection. Crenshaw’s argument assigns no rank; its plaintiffs are not presented as the most oppressed litigants in the system but as the litigants the system’s categories could not see, a distinction between invisibility and magnitude that the whole argument depends on. Collins’s matrix explicitly denies that the dimensions can be summed into a hierarchy of suffering, since each position in the matrix is a distinct configuration, not a score. Jones’s argument, the oldest of the three, converts the compound position into a claim about leadership and clarity, which is a claim about the position’s relation to the whole, not about its place in a table of wounds.

The ranking charge also rebounds, and the rebound is instructive. Any politics that allocates attention, resources, or solidarity must prioritise somehow, and the Marxist tradition prioritises by structural position, holding that the claims of those at the point of production have a strategic weight the claims of others do not. That prioritisation is defensible, and this article’s tradition sections defended its logic, but it is a ranking, executed by a theory rather than by an inventory, and its holders are ill placed to treat ranking as such as the rival framework’s disgrace. The honest difference is over the principle of prioritisation, structural centrality against compounded burden, and over whether either principle should be permitted to silence the claims it ranks lower. On the second question the two traditions’ founding figures are, once again, in unexpected agreement: Jones and Crenshaw both built their arguments against the silencing use of a priority, and the classical Marxist defence of priority, from the Irish dictum onward, insists that the primary struggle is disabled, not advanced, by the subordination it tolerates.

The segmented labour market: a Marxist precedent

The constitutive analysis of a labour market is not a foreign import into Marxist economics, and one precedent deserves recovery because it shows the method operating, with the tradition’s own tools, on the debate’s central materials a decade before the term was coined. The economist Michael Reich’s Racial Inequality, published in 1981, set out to test, against American data, the proposition that racism among workers was a ruling-class instrument whose benefits flowed to employers. Reich’s argument inverted the assumption that white workers gained from racial division: in the metropolitan labour markets he studied, greater racial inequality was associated with lower wages for white workers as well as Black, and with weaker unions, while the gains accrued to capital through the division of the workforce against itself. Whatever the verdict of later econometrics on the details, the argument’s form is what matters here. Race appears in it neither as an attribute added to a class analysis nor as a distraction from one, but as a mechanism internal to the class structure, segmenting the labour market, weakening the collective capacity of labour as a whole, and thereby entering the determination of the wage itself.

The segmentation literature of which Reich’s book is the Marxist exemplar, alongside the dual labour market studies of the same period, supplies the missing empirical middle of the present debate. Segmentation is precisely a claim that a labour market is not one market with frictions but several markets manufactured by institutions, and that the manufacturing runs along lines of race and sex through specific devices: job ladders closed to some groups, credential requirements calibrated to exclude, seniority systems of the kind the founding legal case examined. A Marxism equipped with the segmentation literature does not need to borrow the constitutive insight; it needs to remember that it has it. An intersectional analysis equipped with the same literature gains what its programmatic statements lack, a political economy of the mechanisms it exhibits. The precedent also corrects a lazy symmetry in the debate’s rhetoric, in which Marxism figures as the theory of one structure and intersectionality as the sensibility of many. At the level of the working literature, the Marxist economists of segmentation and the intersectional sociologists of segmented workforces have been studying the same factories with compatible instruments for forty years, and the quarrel is loudest precisely where the working literature is least read.

The household as crossing point

If one site had to be chosen where the debate’s abstractions become concrete, it would be the household, and the choice explains why Marxist feminism, of all the tradition’s branches, is the debate’s natural home. The household is where labour power is daily and generationally renewed, where the wage is converted into meals, rest, care, and raised children, and where the burdens of that conversion are distributed by sex with a consistency that no labour market reform of the last century has dislodged. It is also where class and race cross in their most intimate form: the households of the better paid purchase the labour of the worse paid to perform the renewal work, so that the emancipation of some women from the kitchen has been accomplished, in part, by the employment of other women in it, at wages set by the racialised and migratised structure of the care labour market. Davis made the point in 1981 about the housewife ideal’s irrelevance to Black women’s lives; the care economy literature has since documented its global form. Any framework that cannot hold the household in view will misdescribe all three structures at once, and the test discriminates usefully: the additive version gestures at the household as another site of burdens, the constitutive version and the social reproduction framework meet there as on common ground, since the renewal of labour power is precisely a mechanism by which the structures produce one another.

The linked article on social reproduction theory and its unitary claim develops the household analysis in full, and the cross-reference is not decorative. Of all the synthesis attempts surveyed in this article, social reproduction theory is the one whose central mechanism is a crossing point: it claims that the capital relation cannot be stated completely without the unwaged work that reproduces its indispensable input, and that the organisation of that work, by sex, and its purchase across race and citizenship lines, is therefore internal to the class relation and not an intersection with it from outside. Whether the claim succeeds is the priority dispute again, but its form should be noticed: it is the strongest available answer to the Marxist objection from within the intersectional problematic, and the strongest available answer to the intersectional objection from within Marxism, and it achieves both by refusing the metaphor of crossing roads for the metaphor of a single process with internal differentiations. A reader who follows one link from this article should follow that one.

Why the synthesis is contested

The synthesis literature surveyed earlier has not gone unchallenged, and completeness requires the challenges, since a guide that reports only the reconciliations would misdescribe the field as thoroughly as one that reports only the quarrel. From the intersectional side, the standing worry is domestication: that a synthesis on Marxist terms will preserve the vocabulary of interlocking structures while quietly restoring the priority of class in the analysis’s actual operations, so that race and gender are honoured in the premises and demoted in the conclusions. Bilge’s criticism of the academy’s whitening of the concept generalises into this worry, since a framework can be defused by absorption as readily as by exclusion. The worry is not paranoid; it is the deferral pattern translated into theoretical form, and its holders can point to syntheses in which the compound cases appear in the introduction and vanish from the case studies. From the Marxist side, the standing worry is dissolution in the other direction: that a synthesis on intersectional terms will preserve the vocabulary of exploitation while quietly converting it into a synonym for disadvantage, so that the theory’s distinctive claim, that the poles of the class relation produce each other, is honoured in the preface and abandoned in the analysis. Gimenez’s insistence that the categories must be derived from the system’s dynamics, and not merely juxtaposed with them, is the rigorous form of this worry.

Both worries are, at bottom, demands for a test that a claimed synthesis can be made to pass or fail, and the four tests given earlier supply it in a usable form. A synthesis that names its mechanisms, keeps class relational in its working chapters rather than only in its definitions, declines to schedule the compound questions, and states its remedies at the level of the mechanisms it names, has earned the title; a synthesis that performs the vocabulary in its framing and the old single-axis habits in its substance has not, whichever direction it travelled from. The debate’s maturity can be measured by the migration of its arguments toward such tests and away from the policing of allegiance, and on the evidence of the last decade’s literature the migration is under way, driven less by goodwill than by exhaustion: the slogans have been exchanged at full strength so many times that the field’s working scholars have begun, quietly, to grade one another’s mechanisms instead.

The concept in the world’s law

The legal career of the founding argument did not remain American, and its travel into international instruments is worth a brief record, since it shows the compound claim being institutionalised in the very decades when its popular meaning was thinning. The conventions against racial discrimination and against discrimination against women had been drafted, characteristically, as separate instruments, each with its own committee and its own single axis. Around the turn of the century the committees began to correct the architecture from inside: the committee supervising the women’s convention adopted language recognising that discrimination against women is compounded by race and other grounds, and its general recommendation of 2010 on the core obligations of states parties named intersectionality expressly as a basic concept for understanding the scope of those obligations. The world conference against racism at Durban in 2001 likewise placed the compounded position of women at the centre of its declarations. Whatever one’s estimate of such instruments’ force, the direction of travel is the opposite of the one the term’s critics usually describe: in the law, the concept moved from the journals into the texts that define states’ duties, because the administrative problem it names, the dissolution of compound claims between categories, is a problem every remedial system has and none can delegate.

The traveller’s tale has the usual moral. In the instruments, the term retained its remedial specificity because the setting demanded it: a committee examining a state’s report must ask which women, in which labour markets, under which rules, and the compound question arrives as a question about mechanisms. In the seminar and the slogan, the specificity was optional and was shed. The contrast is the best available evidence for the article’s recurring claim that the concept’s deformations are produced by its settings rather than by its content, and it returns the reader to the founder’s clarification with new respect: a heuristic built for remedial systems will behave like one wherever remedies are actually being designed, and like a gesture wherever they are not.

Du Bois and the wages of membership

The genealogy has one more major figure, chronologically before Jones and analytically indispensable, since he supplied the classic account of how a racial division is maintained inside a class by benefits that are real without being, in the last analysis, economic. In Black Reconstruction in America, published in 1935, W. E. B. Du Bois argued that the white labourer of the American South, though paid a low wage, was compensated in part by a public and psychological wage: access to public functions and public deference, admission to parks and schools from which Black citizens were barred, the courtesy title of citizen and voter in a polity that treated his whiteness as a credential. The wage was paid in status and in services rather than in money, and it was paid, Du Bois argued, precisely so that the white worker would not make common cause with the Black worker whose exploitation was deeper but whose enemy, on the analysis, was the same. The passage is one of the most consequential in American social thought, and its form should be registered: it is a materialist account of a racial ideology, explaining the ideology by the real, if meagre, benefits it distributes, and thereby explaining the division of a class by a mechanism internal to the class’s own situation.

For the present debate Du Bois’s analysis does double duty. It is, first, another antecedent from within the broad materialist tradition for the constitutive claim that race in America is not an attitude laid over the class structure but a load-bearing element of it, distributing real goods and thereby reproducing itself. It is, second, a standing answer to a crude form of the priority claim, since the psychological wage is a mechanism by which a supposedly superstructural division maintains the base’s segmentation against the base’s own integrative pressures, a causal arrow running from the racial order into the class order that any adequate theory must accommodate. The strongest Marxist writing has always accommodated it, as the segmentation literature showed; the strongest intersectional writing is unintelligible without it, since the matrix’s interpersonal and hegemonic domains are, in large part, the psychological wage described at the level of the system. Du Bois belongs to neither camp because he precedes the camps, and his presence at the head of the genealogy is the final embarrassment of the two-camps framing: the tradition that produced the compound analysis, the segmentation analysis, and the wages analysis is one long materialist conversation, and the quarrel is a dispute among its heirs about the estate.

The party framework Jones worked inside

Jones’s essay gains a further layer when its immediate theoretical setting is restored, since she wrote inside a movement that possessed an official theory of the race question, and her argument is partly an intervention in it. The American party, following the Communist International’s resolution of 1928, held that Black Americans in the Black Belt counties of the South constituted an oppressed nation with a right to self-determination, and the party’s work was organised around that thesis. Whatever the thesis’s merits as strategy, its theoretical form is instructive for this article: it treated the racial question as a national question, a matter of a people’s collective status, which is to say it processed race through a category borrowed from another axis of the movement’s map. Jones’s intervention, from within that framework and without repudiating it, was to insist on the position the framework’s categories rendered least visible: the Black woman worker, whose situation was not exhausted by the national question, nor by the woman question as the movement’s women’s work conceived it, nor by the class question as its trade union work conceived it. Her triple oppression argument is thus, among other things, an early demonstration of the single-axis failure operating simultaneously in three departments of one organisation, each department equipped with a theory that explained why its axis was the real one.

The episode generalises into the article’s most portable historical lesson. Single-axis thinking is not a product of bad categories alone; it is a product of organisation, of departments, journals, caucuses, and funding streams each constituted around one axis and rewarded for its defence. Every movement in this article’s history exhibits the structure, including the movements founded to correct it, which is why the compound analysis has had to be re-founded in each generation and each institution, by Jones against the party’s departments, by the Combahee Collective against the movements of its decade, by Crenshaw against the doctrine and the movements that mirrored it. A theory of the phenomenon that ignores its organisational basis will keep being surprised by its recurrence. A theory that includes the basis, as the compound tradition at its best does, can at least predict where the next dissolution of a compound claim will occur: wherever a remedial or political system is administered in departments, which is to say, everywhere such systems exist.

The mirror slogans

Every durable debate grows a pair of slogans that relieve their users of the need to read, and this one’s pair is worth naming in the article’s descriptive mode, since students will meet both in the wild and should know what each is doing. On one side, the slogan holds that all politics not organised around class is a diversion arranged, or at least welcomed, by the existing order: the analytical content available inside the slogan is the institutional career argument, that additive frameworks are cheap for institutions to adopt, inflated into a universal verdict that relieves the user of distinguishing versions. On the other side, the slogan holds that all class politics is a cover for the postponement of every other emancipation: its analytical content is the deferral archive, inflated into a verdict on a category rather than on a documented institutional tendency, and relieving the user of the same distinction in the opposite direction. Each slogan is the other’s recruiting agent, since nothing confirms the first like an encounter with the second, and a literature policed by the pair will reproduce the quarrel indefinitely without touching the questions this article has recovered: the version in use, the grammar of class, the mechanisms named, the remedies implied.

The descriptive point can be sharpened into a rule of reading. When a text in this debate reaches for the universal form, all such politics, every such framework, it has left the ground on which evidence operates, since the evidence in this field is obstinately particular: this case’s seniority rule, that movement’s resolution, this labour market’s segmentation, that synthesis’s working chapters. The founding writers are particularists almost to a fault. Crenshaw’s argument is built from three lawsuits; Jones’s from the wage structure of two industries; Collins’s matrix from the documented organisation of specific institutions; Marx’s own analyses, the tradition’s standard of rigour, from the factory inspector’s reports of a particular industry in a particular decade. A reader trained to ask for the case will find the debate’s literature sorting itself, the strong work exhibiting its mechanisms and the weak work exhibiting its allegiances, and the sorting is more valuable than any verdict on the camps, because it is a skill that transfers to every other controversy the reader will meet.

What follows if the constitutive version is taken seriously

It is worth spelling out, in closing the analytical part of the article, what concretely changes for three kinds of reader if the constitutive version and the relational account of class are both accepted, since a distinction that changes nothing is a distinction without a difference. For the researcher, what changes is the design of inquiry: the unit of analysis becomes the mechanism rather than the group, so that a study begins not from a population’s characteristics but from an institution’s operation, asking how a hiring system, a benefit rule, or a care market allocates positions jointly by class, race, and sex, and treating the groups the census enumerates as the mechanism’s products to be explained rather than as its premises. For the lawyer and the policy designer, what changes is the drafting: categories of remedy are audited in advance for the compound cases they will dissolve, comparators are chosen with the DeGraffenried lesson in view, and provision is tested against the position of those at the crossings, on the principle the founding article established, that a system is characterised by what it does to the claims its categories fit worst. For the organiser and the citizen, what changes is the audit of one’s own movement: the deferral pattern is treated as a standing temptation of organisation as such, the compound claims inside the constituency are put on the agenda by design rather than by insurgency, and the movement’s theory is required to pass the test its tradition’s own best writers, from Jones onward, set for it.

None of these consequences requires the priority dispute to be settled, and that is the article’s final analytical point. The constitutive version and the relational definition are compatible with both answers to the question of a centre, since a centred constitutive analysis, of the social reproduction kind, and a decentred one, of the matrix kind, share every discipline listed above. The disciplines are therefore available now, without waiting for the frontier dispute to close, and the literature’s failure to adopt them is not a theoretical failure but the ordinary inertia of camps. A reference article cannot dissolve camps. It can put the instruments on the table, in findable form, and record that they were there.

The founder against the schedule: the Irish precedent

The claim that the Marxist founder’s own practice refutes the scheduling logic deserves its evidence in full, because it is the strongest card the Marxist side holds in the historical dispute and it is usually played without its documentation. In the eighteen sixties Marx’s analysis of the English working class led him to a conclusion that reversed the movement’s instinctive ordering. English workers, he observed, regarded Irish immigrant workers with a hostility that the press, the pulpit, and the employers cultivated, and he identified this antagonism as the secret of the English working class’s impotence despite its organisation: the ordinary English worker saw in the Irish worker a competitor who lowered his standard of life, and in that attitude, Marx wrote, the English worker echoed the poor whites of the American slave states toward the enslaved. The antagonism was, for Marx, artificially sustained and was the mechanism by which the possessing classes kept both populations in check. His conclusion was organisational and unconditional: the English working class would never accomplish anything until it had separated its policy toward Ireland from the policy of the ruling classes, and the emancipation of Ireland was not a side question of justice but the precondition, the lever, of the English workers’ own movement, a judgement he carried into the International Workingmen’s Association against considerable resistance.

Set beside the deferral pattern, the episode is a controlled counterexample from the tradition’s source. Here a division along national and quasi-racial lines, operating inside the class through wages, status, and cultivated contempt, is treated not as a secondary matter to be resolved after the main conflict but as the main conflict’s present mechanism, whose dissolution is the first item on the agenda. The analysis is constitutive in everything but vocabulary: the class relation in England is shown to exist in a nationally divided form, and the division is shown to be functional for the relation’s reproduction. A Marxism that generalises its founder’s Irish method would have no need to borrow the compound insight from anyone. The intersectional objection’s historical force is precisely that the method was not generalised by the tradition’s institutions, which reached for the schedule as often as for the lever. Both facts are in the record, and the article’s verdict on the theory side of the charge, stated earlier, rests on holding them together: the texts contain the method and the temptation, and the institutions chose, repeatedly, in ways the texts did not compel.

The representative worker

A quieter form of the single-axis failure operates inside theory itself, in the choice of the representative figure through whom a structure is imagined. Marxist writing’s exemplary worker was, for most of the tradition’s history, a man at a factory bench, and the choice seemed innocent because the theory’s categories are indifferent to the sex of the labourer who occupies them. But the representative figure steers the analysis that the categories permit: a theory visualised through the bench will treat the wage as the worker’s whole relation to subsistence, and will correspondingly underweight the unwaged labour that the wage presupposes, the labour markets that never had a bench, the domestic and agricultural sectors where, as Jones documented, the most exploited workers were concentrated, and the legal dependencies, of wives, of migrants, of the colonised, that determined who could stand at a bench and on what terms. The domestic labour debate of the nineteen seventies was the tradition’s own discovery of the steering effect, and social reproduction theory is its systematic correction. Intersectional writing’s exemplary figure, the Black woman plaintiff or worker, was chosen with the opposite intent, to stand where the categories cross, and it steers its analysis correspondingly toward the crossings and away from the centres of the system, the boardrooms and treasuries where no crossing is suffered but every crossing is arranged.

The mature position, available to both traditions and stated here as the article’s own, is that a structural theory needs both figures and a rule for moving between them: the figure at the crossing, who reveals how the structures are joined in a life, and the figure at the centre, who reveals what the joining is for. The debate’s sterility is in part a war of representative figures conducted as though it were a war of theories, and the war ends, where it ends, when each side admits the other’s figure into its own analysis as a necessary station rather than a rival emblem. Jones, characteristically, had both figures in 1949: the Black woman domestic worker and the employer whose household her labour subsidised appear in her essay as the two ends of one relation, which is why the essay reads, across the decades, like a dispatch from the debate’s conclusion rather than from its prehistory.

The statement’s afterlife

The Combahee statement’s own career illustrates the recoveries this article has practised, since the document was for decades more cited than read and more honoured than situated. Written in 1977 by a collective whose members included the writer Barbara Smith, the statement circulated first in movement channels and reached print in the anthology Capitalist Patriarchy and the Case for Socialist Feminism in 1979, a venue whose title alone refutes the later assumption that the statement’s feminism and its socialism were separable layers. Its members were explicit that the document was a collective’s working position, revised in discussion, and its most quoted sentence, the declaration that the most profound and potentially most radical politics come directly out of their own identity, sits inside paragraphs specifying the material systems, capitalism and imperialism named outright, that the politics existed to destroy. The sentence’s extraction from those paragraphs, and its circulation as a freestanding motto, is the additive fate in miniature: a constitutive argument, that a politics beginning from a materially specified position can see the whole, converted into an attribute claim, that identity as such confers authority.

The statement’s return to the centre of the record in the twenty-first century, through the scholarly recovery of Black left feminism and through the volume How We Get Free, edited by Keeanga-Yamahtta Taylor in 2017, which republished the statement alongside interviews with its surviving authors, restored the document’s context at the moment of its widest circulation. The interviews are instructive for the debate because the authors, given the floor at last, describe a politics that fits neither camp’s portrait: socialist in economics, feminist and antiracist without remainder, and impatient in equal measure with the class politics that scheduled them and with the later usage that quoted them as sponsors of a politics of recognition. A reader who wants the compound tradition’s self-understanding, rather than its camps’ versions of it, will find no better source, and the recovery of the statement’s full text and context is the single development of the last decade most damaging to the slogans on both sides.

The floodgates argument

One doctrinal detail from the founding case generalises so well that it earns a section, because it names the standard institutional defence against every compound claim, in law and far beyond it. The DeGraffenried court, refusing the combined class, worried in substance that recognising Black women as a protected category would open a Pandora’s box of compound categories, each demanding its own remedy, until the framework collapsed under the multiplication. The fear is perennial and it wears many costumes: the administrator who warns that every exception will be claimed, the movement that warns that every specific demand will fragment the whole, the theorist who warns that every added axis dilutes the analysis. In each costume the argument has the same structure: it defends a framework’s manageability against the claims the framework was built to serve, and it treats the proliferation of the inconvenient as a refutation of the inconvenient.

The founding article’s implicit answer is the one this article has given throughout: the multiplication is a property of the additive imagination, which pictures each compound as a new box in the cabinet, while the constitutive view dissolves the fear by changing the object. A doctrine that asks how a mechanism treats those positioned at its crossings does not need a box for every crossing, any more than a medicine that asks how a treatment affects the multiply burdened needs a disease for every combination. The floodgates argument is therefore diagnostic as well as obstructive: its appearance in a debate signals that the additive version is the one being imagined by the institution under criticism, and that the criticism has not yet been heard in its strong form. Students of the debate should learn to hear it as a tell, in judgments, in manifestos, and in their own first reactions, since the fear of the proliferating list is the list’s own defence mechanism, and it operates as reliably in the reader as in the court.

The standpoint’s Marxist root

The epistemological strand of the intersectional tradition has a documented Marxist source that neither camp advertises, since it embarrasses both simplifications at once. In an essay of 1983 that became one of feminist theory’s founding documents, Nancy Hartsock proposed a feminist standpoint explicitly on the model of Marx’s account of the standpoint of the proletariat. Her argument proceeded by parallel: as the proletariat’s position in production affords, when worked up into consciousness through struggle, an understanding of the social order unavailable from the standpoint of capital, so women’s position in the sexual division of labour, in the work of bodies, households, and care, affords the materials of a standpoint from which the order’s patriarchal structure becomes visible. The essay is unambiguous about its method’s provenance, engaging Marx and Lukács directly, and it generalises the model’s logic rather than borrowing its slogans: a standpoint is not a perspective, it is achieved, it is grounded in a division of labour, and its claim is that some locations make some structures legible.

The intersectional development of the standpoint, in Collins above all, then appears in its true light: not as the importation of an alien epistemology into the neighbourhood of Marxism, but as the correction of the standpoint model from inside its own logic, by the observation that there is no single location of the oppressed from which the whole becomes visible, because the structures cross, and the view from any one crossing is partial in patterned ways. That correction is the same move, one level up, that Hartsock had made on the classical model, which had treated the proletariat’s location as undivided. The sequence, Marx’s proletarian standpoint, Hartsock’s feminist standpoint grounded in the sexual division of labour, Collins’s matrix of situated standpoints, is a single research programme refining its central instrument across a century, and its members cite one another accordingly. A debate that presents the two traditions as epistemological strangers is refuted by their footnotes. What survives the refutation is the genuine question the programme has never closed, whether the legibility of structures from below requires, or excludes, the claim that one structure organises the field of vision: the priority dispute, wearing its epistemological dress, and no nearer settlement for being better dressed.

The textbook form

Every framework that enters the academy acquires a textbook form, and the textbook form is where the additive conversion is most efficiently performed, a fact worth recording because students meet the framework there first. The widely used introduction by Collins and Sirma Bilge, Intersectionality, published in 2016 in a series of short guides to key concepts, presents the framework through its core ideas and a range of applications, and its authors are scrupulous: the book insists on the framework’s origins in critical inquiry and praxis, on its attention to power, and on the relationality of its categories. Yet the genre exerts its pressure even on scrupulous authors. A key concept must be rendered portable, its ideas listed, its applications surveyed field by field, and the rendering reintroduces, at the level of the book’s architecture, the enumeration the theory forbids at the level of its claims: chapters proceed by domain, the domains accumulate, and the student who reads for the list will find one, whatever the text says. The phenomenon is not hypocrisy and not carelessness; it is the additive fate operating through the most innocent of mechanisms, the table of contents, and it explains why the weak version is so often the version that graduates.

The lesson generalises to this article, which is itself a reference text and therefore subject to the same pressure: its table, its tests, its numbered disciplines are instruments for finding an argument, and a reader determined to extract from them a list of approved positions will succeed, at the price of missing the argument the instruments serve. The defence available to a reference text is the one attempted here, to build the instruments so that their use requires the distinction they encode: a table whose columns are versions rather than camps, tests that grade the reader’s own side as readily as the other, and a genealogy that gives every camp an ancestor it did not expect. Whether the defence works is decided by readers, in seminars and elsewhere, and the persona sections that follow are written for the three kinds of reader most likely to decide it.

A reading path through the debate

A reader who wants the debate at first hand, rather than at the remove of this article’s summary, should take the texts in an order that is not the order of their fame. Begin with the two legal articles, since they are short, exact, and habitually misdescribed by friend and foe alike; an hour with the argument of 1989 and its sequel of 1991 will inoculate against most of what is later said in their name. Read Jones’s essay next, in the edition the recovery scholarship has made available, and read it as what it is, an organiser’s memorandum to her own movement, before turning to the Combahee statement, which repays being read entire, its economic paragraphs included, rather than in the extracts through which it usually circulates. Collins’s book of 1990 then supplies the systematic statement, and McCall’s methodological survey of 2005 the map of what researchers have done with it.

Only after that foundation should the objections be approached, and they reward the order given here. Hartmann’s unhappy marriage essay and Vogel’s reply to dual systems thinking show the quarrel’s first staging inside Marxist feminism, and they make the later exchange legible as a reprise rather than a rupture. Gimenez and Mitchell then give the Marxist objection its contemporary force, and they should be read against the synthesis shelf, Vogel again, Arruzza, and Bohrer, so that the objection and the answers contend in the reader’s mind at full strength simultaneously. Nash’s internal criticism and Bilge’s warning about domestication close the circuit, by showing the framework’s own scholars policing its deformations with a severity no external critic has matched. A reader who finishes the circuit owns the debate in the only sense that matters: not a verdict to repeat, but the capacity to place any new contribution in it, to say which version it uses, which court its evidence belongs to, and which of the standing objections it has actually met. That capacity, and not fluency in the slogans, is the education the texts exist to give, and it transfers intact to every neighbouring controversy in this series, from racial capitalism to the theory of the state, wherever a compound reality meets a single-axis vocabulary.

For students and exam candidates

Begin where this article begins, with the legal origin, because examiners reward the candidate who can say what the term was coined to do before saying what it has come to mean. Your first move in any answer on this topic is the recovery: an article in a law journal, a compound claim dissolved between comparison classes, a doctrine repaired in intention. Your second move is the antecedent, Claudia Jones in 1949, which demonstrates that you know the debate’s chronology and not merely its slogans. Your third is the distinction between the additive and constitutive versions, applied to whichever objection the question puts to you, since most set questions in this area are answered by specifying the version the objection targets. For the relational definition of class on which the whole analysis depends, work through the linked guide to the Marxist theory of class and keep the definition in its structural form. The VaultBook study notes include a card on the two versions designed for exactly this revision sequence: origin, antecedent, versions, objections, residue. Candidates lose marks here in two predictable ways, by treating the term as a general theory it never claimed to be, and by answering the priority dispute as though it had been settled. State it as open, locate it precisely, and you will be writing the answer the debate’s own literature supports.

For teachers and seminar leaders

The two versions table is this article’s gift to your classroom, because it converts a topic that reliably degenerates into allegiance into one that can be taught analytically. Set the founding case first, the seniority system and the dissolved claim, and require students to state the additive and constitutive readings of it before any theorist is named; the distinction learned on a mechanism is retained when the names arrive. Then run the objections in their strongest attributed forms, Gimenez and Mitchell against the list, the deferral archive against the schedule, and grade contributions by the four tests: version, relation, mechanism, remedy. The seminar exercise that does the most work is the transposition, in which each side’s objection must be restated as a grading of the speaker’s own tradition before it may be aimed at the other. Students who have graded Marxist texts for scheduling and intersectional texts for listing do not easily return to slogans. Keep the contemporary polemics out of the room by the article’s method, not by prohibition: a debate routed through 1851, 1892, 1935, 1949, and 1989 on its way to the present arrives at the present with its evidence in hand and its temperature lowered. The misreadings section supplies your closing summary.

For researchers, historians, and journalists

Your usable asset in this article is the genealogy, and its discipline of dates. The compound analysis descends through named texts in named venues: a Boston collective’s statement of 1977, printed in a socialist feminist anthology in 1979; an organiser’s essay in a party’s theoretical journal in 1949; a legal article in 1989 and its sequel in 1991; a sociological synthesis in 1990; a methodological map in 2005; the recovery scholarship of the last two decades. Each link in that chain is checkable, and the chain as a whole dissolves the two framings you will be offered by partisans, in which the term was founded in 1989 as an external criticism of Marxism, or in which it was always present and nothing was ever neglected. For researchers, the open frontier is the priority dispute, and the productive sites are mechanisms: segmented labour markets, care economies, doctrinal comparison classes, where centred and decentred models generate different expectations about the same archive. For journalists, the origin section is the single most useful correction available on a term you will meet constantly: it started life as an argument about a seniority rule in an employment dispute, and any usage that cannot survive that fact is a usage to describe, attributed, rather than to adopt. Quote the debate’s participants from their texts and dates, and the story you file will be the rarest one in this territory, an accurate one.

Before the term: the critique of rights

The founding article did not arrive from nowhere in its author’s own bibliography, and its immediate predecessor explains a feature of the argument that readers often miss: its ambivalence toward the very legal framework it sought to repair. In 1988, in the Harvard Law Review, Crenshaw had published Race, Reform, and Retrenchment, a long examination of civil rights discourse written from within the critical legal studies milieu. Its argument held two thoughts together that simpler positions separate. Rights discourse, she argued, had genuinely served the Black freedom struggle as a language of legitimation and mobilisation; and the same discourse, by framing racial subordination as a matter of formal inequality and individual prejudice, legitimated the order that reproduced the subordination in material form, so that each doctrinal victory could function as a ratification of the arrangements it left standing. The 1989 article is that double argument applied to doctrine’s internal machinery: the demand is not for the abandonment of rights but for a doctrine that cannot be satisfied by victories which dissolve the compound claimants.

The milieu matters for the debate with Marxism because critical legal studies was itself in argument with a Marxist inheritance, the critique of legal form, and Crenshaw’s position in that argument was a mediating one: against the strand of critical scholarship that treated rights as mere mystification, she defended their historical utility to the subordinated; against liberalism, she insisted that the utility was compatible with the form’s service to the order. A reader who knows the 1988 article will not mistake the 1989 one for a plea for more categories in a liberal framework, since its author had already published the framework’s critique. The point disciplines a lazy version of the Marxist objection, in which intersectionality figures as liberalism’s house theory of disadvantage. Whatever the institutional usage became, the founding texts were written against liberalism’s self-understanding from a position formed in argument with Marxism, and the debate between the traditions is, at its source, a family argument conducted with shared suspicions about formal equality.

The parallel from decolonial feminism

No map of the debate is complete without the parallel framework that arose outside the North American academy, since it criticises intersectionality from a direction neither camp in the main quarrel occupies. The Argentine philosopher María Lugones, in work published from 2007 onward, developed the concept of the coloniality of gender, arguing that the very categories the intersectional framework intersects, race and gender as separable dimensions, are themselves products of the colonial order: on her account, the colonised were not positioned within a pre-existing gender system as subordinate men and women, but were placed outside the colonial gender binary’s human terms altogether, so that an analysis which begins by separating the axes and then studying their crossing has already accepted the colonial sorting it hopes to criticise. Her preferred image was not the intersection but the fragmentation of the colonised subject by a logic she called the colonial difference, and her proposal was a decolonial feminism that would begin from the resistance of colonised women rather than from the categories of the coloniser’s sociology.

The parallel is valuable to this article for three reasons. It demonstrates that the constitutive impulse, the refusal of the list, is not a local peculiarity of one debate but a general pressure within the field, arising independently wherever the additive picture is examined hard. It supplies the strongest available version of a criticism that the Marxist objection only gestures at, that the framework’s categories are historically produced rather than given, though Lugones draws from it a decolonial rather than a Marxist conclusion. And it returns the debate, once more, to the question of what an analysis takes as its primitives: for the Marxist, the relation of production; for Collins, the matrix; for Lugones, the colonial difference. A student who can state the three primitives and what each does to the compound cases has understood the field’s actual structure better than a student who can recite the camps. The parallel debates among these frameworks are beyond this article’s scope, but their existence is the final evidence for its method: the interesting units in this territory are arguments and their primitives, not allegiances and their histories of grievance, instructive as those histories are.

Oppression as structure

Underneath the debate’s exchanges sits a philosophical shift that both traditions underwent in the same decades, and naming it explains why the debate became possible when it did. The older vocabulary of oppression pictured an agent and an act: a tyrant, a boss, a husband, doing something to someone, and the remedy was correspondingly personal, the removal of the agent or the prohibition of the act. The newer vocabulary, which Iris Marion Young systematised in Justice and the Politics of Difference in 1990, treats oppression as structural: a condition produced by the accumulated operation of institutions, norms, and markets, which can constrain a group’s possibilities systematically without any single agent willing the constraint, and which Young analysed into five faces, exploitation, marginalisation, powerlessness, cultural imperialism, and violence, each a distinct mechanism rather than a synonym for the others. The founding legal article belongs to the same shift, since the seniority system’s harm required no villain, only a rule, a hiring history, and a doctrine. And classical Marxism had made the shift earlier than either, in its central case: exploitation in Marx’s account is not an act of cheating by an individual capitalist but a property of a relation that honest men administer, which is why the tradition’s polemic has always targeted systems in its theory even when its movements hunted persons in their practice.

The shared shift has an underappreciated consequence for the debate’s conduct. Once oppression is structural on both sides, the argument can no longer be settled by exhibits of prejudice or of goodwill; it becomes an argument about which structures exist, how they are joined, and what reproduces them, which is to say it becomes the argument this article has reconstructed. The persistence of the older vocabulary in the debate’s public form, in which positions are graded by the moral quality of their holders, is another of the travelling shadows: a pre-structural rhetoric applied to a post-structural dispute. Readers who make the shift explicit, in their own writing and in their demands on others’, will find the literature’s centre of gravity moving toward them, since the working scholarship of both traditions already writes in the structural register and reserves the other for its dust jackets.

On sources and silences

A word about the archive is owed, because the genealogy this article relies on has a material history that is itself evidence. Jones’s essay survived because a party journal was printed and kept; her wider body of work survived partially, through the newspaper she founded in exile and the recovery labour of scholars who reconstructed her career from fragments, surveillance files, and the memories of a shrinking circle. The Combahee statement survived because its authors kept it and because an anthology printed it. The pamphlets of the Black Women’s Alliance survived in small numbers in the collections their makers donated. Against this record of narrow escapes stands the ordinary archive of the movements that scheduled the compound claims, whose proceedings, journals, and memoirs fill the libraries. The asymmetry is not a conspiracy; it is the documentary form of the deferral pattern, since institutions preserve what they value, and it imposes a discipline on the historian of this debate: the absence of a compound argument from a movement’s record is evidence about the movement’s preservation as much as about its thought, and the recovery of the missing arguments from the margins of the archive, which the scholarship of the last two decades has performed, is a correction of the record and not a revision of it.

The same discipline applies to the article’s own construction. Its genealogy is drawn from the recovered record, and a generation ago several of its central exhibits could not have been cited from standard sources, because the standard sources had not yet absorbed the recovery. Reference works age by their archives. The author of a future edition of this article will work from a fuller record still, very likely including materials the present debate’s participants have not thought to preserve, and the article’s claims are framed to survive that improvement: they rest on the texts that exist, quoted at their dates, and on distinctions that do not depend on the archive’s completeness for their force.

The exclusion written into the welfare state

One historical mechanism deserves detailed statement, because it shows a compound position being manufactured by legislation with a clarity that theoretical exposition cannot match, and because it is the concrete referent of Jones’s analysis. The American social legislation of the nineteen thirties, the National Labor Relations Act of 1935, the Social Security Act of the same year, and the Fair Labor Standards Act of 1938, built the century’s floor of workers’ protection, and each statute excluded from its coverage the two occupations in which Black workers, and Black women above all, were concentrated: agricultural labour and domestic service. The exclusions were the price of the Southern congressional bloc’s support, exacted to preserve the labour regime of the Jim Crow South, and their effect was to write the racial and gendered segmentation of the labour market into the welfare state’s foundations: the workers inside the statutes acquired the rights to organise, to social insurance, and to minimum wages, while the workers outside them, disproportionately Black and female, were left to the unregulated market for a generation, until amendments of the nineteen seventies partially closed the gap. Ira Katznelson’s When Affirmative Action Was White, published in 2005, documented the exclusions’ consequences across the century, and the documentation is now standard.

Hold the mechanism up against the article’s distinctions and it rewards the holding. The exclusion is not intelligible as a race burden added to a class burden added to a sex burden. It is a single legislative act that constituted a class fraction, the unprotected workforce, by racial and occupational lines that were also lines of sex, since domestic service was women’s work in the statute’s imagination as in the market’s. The fraction’s members were not workers who also happened to be Black and female; their position as workers, unprotected, unwaged by law’s floor, unorganised by law’s leave, was manufactured by the same clauses that marked their race and their occupation. Jones wrote her essay fourteen years after the Social Security Act, about precisely this workforce, and her category of super-exploitation names the position the exclusions built. Crenshaw, forty years later, analysed a seniority system that reproduced an analogous position inside a single firm. The line from the statute book to the essay to the legal article is the compound tradition’s main line, and it runs through mechanisms the whole way: no attitudes are required to explain it, and no list of burdens describes it, which is the constitutive thesis demonstrated on the debate’s home ground.

The universal as premise and as project

One philosophical issue underlies the whole debate and should be surfaced before the article closes, since it explains the passions that the analytical questions, left to themselves, would not generate. Both traditions are universalist in ambition and particularist in evidence, and each suspects the other of betraying the combination. Marxism’s universal is the human emancipation that the working class’s emancipation is held to prefigure, and its critics charge that the universal arrives pre-filled with the particular experience of one segment of the class, so that its universality is a premise smuggled past inspection. Intersectionality’s universal is thinner, a methodological refusal to let any category stand for the whole, and its critics charge that a universal so thin cannot found a politics, since it specifies what an analysis must not do without specifying what a movement must. The exchange of charges has a literature of its own, of which the volume Contingency, Hegemony, Universality, published in 2000 from an exchange among Judith Butler, Ernesto Laclau, and Slavoj Žižek, is the most cited philosophical staging: its participants, starting from different traditions, converged on the thought that the universal is not a premise any politics is given but a project that competing particulars contest to fill, and that the contest is not a defect of universalism but its mode of existence.

The convergence is directly usable in the present debate, because it redescribes the two traditions’ quarrel as a contest over the filling of a shared universal, human emancipation, by particulars each claiming to correct the other’s false filling. On that description, the compound tradition’s contribution is a permanent audit of the filling: any proposed universal must be tested at the crossings, against the positions its categories serve worst, and a universal that fails the test has been filled by a part. The Marxist contribution is the insistence that the universal have a mechanism and a bearer, that emancipation name the relation to be abolished and the force positioned to abolish it, without which the audit has nothing to audit. The two contributions are the article’s two disciplines once more, stated philosophically: the test and the theory, each incomplete alone, each the other’s standing correction. A reader who carries the redescription away will find the debate’s remaining heat easier to discount, since heat in a contest over a shared inheritance is the normal temperature of the proceedings, and is evidence of the inheritance’s value rather than of either heir’s bad faith.

A union that practised the compound claim

The historical record contains, alongside the deferrals, at least one well-documented case of a labour organisation that built the compound claim into its ordinary practice, and the case is worth stating because it converts the article’s conditional, what a movement that refused the schedule would look like, into an exhibit. The United Packinghouse Workers of America, the industrial union of the meatpacking industry from the nineteen forties, made anti-discrimination a department of its work rather than a resolution of its conventions: it maintained an anti-discrimination programme with staff and budget, pressed grievances over the racial and sexual allocation of jobs inside the plants, campaigned against discrimination in the towns where its locals sat, and understood, as the historian Rick Halpern documented in Down on the Killing Floor and Roger Horowitz in his study of the union, that the industry’s job structure was a racial and sexual structure, allocating the killing floors, the trimming lines, and the canning rooms by race and sex through rules that had to be fought as class rules because they were the form class took in the plant. The union was not spotless, and its record has its own internal critics, but its programme demonstrates the practical content of the constitutive claim inside a classical class organisation: the fight over the allocation of jobs by race and sex was the fight over the wage structure, conducted at the point where the structures were joined.

The case also refines the intersectional objection into its most defensible form. If a packinghouse union of the nineteen forties could practise the compound claim with the theoretical instruments then available, then the deferrals elsewhere cannot be excused as the period’s necessity; they were choices, made by organisations that had the alternative in view, often in the same federation and the same decade. The objection, restated with the exhibit in hand, is no longer that Marxist movements were children of their time. It is that the tradition contained both practices, the compound and the scheduling, and that its institutions rewarded the scheduling often enough for the pattern to become the charge. That restatement is harder to answer than the slogan it replaces, and it is the form in which the objection should be taught, since it gives the Marxist side something specific to answer: not a definition, but a comparison of two practices inside its own house, and an explanation owed for the distribution of rewards between them.

The concept in health and policy research

The framework’s migration into public health and policy research, the last of its major travels, is worth recording because that field rebuilt the concept’s instruments for its own purposes and the rebuilding is the additive and constitutive distinction fought over in methodological form once more. Health research took the concept up to explain a stubborn fact of its own: that health outcomes are patterned jointly by the structures this article has tracked, and that interventions addressed to a population defined by one axis routinely fail the subpopulations at the crossings, as screening programmes miss the women whose working hours, languages, and legal statuses the programme’s design assumed away. Olena Hankivsky and her collaborators formalised the uptake in an intersectionality-based policy analysis framework, published in 2012 and elaborated in a 2014 collection, which instructs analysts to begin from the positions policy reaches worst, to treat categories as products of the policy process itself, and to interrogate the problem representations built into a policy’s statement of its aims. The framework’s questions are the founding article’s questions translated into the grammar of ministries, and its existence in a field manual, used by governments’ analysts, marks the concept’s furthest institutional reach.

The same field also exhibits the additive fate at industrial scale, since a framework that enters a ministry as a checklist of populations to be consulted has been converted, by the conversion this article has tracked throughout, into the enumeration of attributes with a consultation attached. The health literature is unusually candid about the tension, because its failures are measurable in unmet need, and its methodological debates, between intercategorical measurement and intracategorical depth, replay McCall’s map with budgets attached. For the student of the concept, the field is therefore the best observatory available: a domain in which the two versions are not merely argued about but funded, implemented, and audited, and in which the constitutive version’s superiority, where it is superior, shows up in the unglamorous currency of reached and unreached populations. The debate with Marxism rarely visits this observatory, since neither camp’s manifestos are written for it. The article commends it to both, as the place where their shared instrument is tested by people who cannot afford the slogans.

The question to keep

This article set out to recover three things that the public quarrel had buried, and its findings can be stated as compactly as the evidence allows. The term intersectionality began as a legal argument about compound claims, made in a law journal in 1989 against a doctrine that dissolved Black women’s discrimination claims between its comparison classes; it was remedial, structural, and innocent of the general theory later attributed to it. The argument had antecedents inside the Marxist tradition itself, most exactly in Claudia Jones’s essay of 1949, which built the case for Black women’s compounded position out of the labour market’s own arrangements and directed the indictment at the movement that employed her, so that the later confrontation between the traditions is a family quarrel misdescribed as a meeting of strangers. And the substantive dispute inside the quarrel is the relation and identity distinction: whether class names a relation of exploitation or a category of disadvantage, a genuine analytical question on which the Marxist objection is decisive against the additive version of intersectionality, the version of the list, and largely powerless against the constitutive version, the version of the founding texts, in which structures produce one another’s forms.

What remains open, the article has insisted, is the question of explanatory priority, whether the configuration of structures has a centre in the organisation of production or is a matrix without one, and the article records that question as a frontier rather than a verdict because the evidence, honestly marshalled, supports research programmes on both sides of it. Everything else the quarrel contains, the article has found to be shadow: the additive usage and the scheduling practice, each richly deserving the other’s criticism, each disowned by the founding texts of the side that practises it. The reader who keeps the question, class as relation or as attribute, and the instrument, the two versions table, is equipped to grade any contribution to the debate, from whichever camp, including the contributions of the traditions’ founders and including this article. That equipment, and not a allegiance, is what a reference work on a contested term owes its readers, and it is what the traditions’ own best writers, from Jones’s journal to Crenshaw’s, were in their different idioms trying to supply.

Reading the founding article closely

Close reading is this series’ habit with founding texts, and the 1989 article rewards it, because its architecture is its argument in a form the summaries flatten. The article proceeds against three targets in sequence, and the sequence is a demonstration that the single-axis failure is one failure with three addresses. The first target is anti-discrimination doctrine itself, examined through the three employment cases: the law’s categories dissolve the compound claimant. The second target is feminist theory, which the article charges with building its subject, woman, from the experience of women privileged by race and class, and with treating the resulting theory’s gaps as the marginal cases’ problem rather than the theory’s. The third target is antiracist politics, charged symmetrically with building its subject from men’s experience and treating women’s subordination within the community as a divisive import. The three charges are the same charge: a framework generalised from the least burdened segment of its constituency will misdescribe the rest, and will experience the misdescription, when challenged, as an attack on the framework rather than as evidence about it. Readers who grasp the architecture see why the article could not be answered by improving any one of its targets. The doctrine’s defenders who reformed a comparison class, the feminists who added a chapter, the movements who added a caucus, were performing the additive repair on a constitutive diagnosis, and the article’s subsequent career, in which it was honoured by exactly those repairs in field after field, was predicted by its structure.

The article’s constructive proposal is correspondingly easy to miss, because it is negative in form. Crenshaw does not supply a replacement doctrine; she supplies a discipline, that the claimants at the crossings be treated as the framework’s test rather than its embarrassment, and that a theory or a movement be required to show what it does with the cases its categories fit worst before it is believed about the cases they fit best. The discipline is the article’s bequest to every field that took the term up, and it is the standard by which this article has graded the debate’s participants, the traditions’ institutions, and the concept’s own custodians. It is also, not incidentally, a standard the Marxist tradition states in its own idiom whenever it is being careful: a theory of a social order is to be judged by its handling of the phenomena at its margins, the lumpenproletariat, the peasantry, the colony, the household, the cases the categories were not built for, and the tradition’s canonical embarrassments are precisely the cases where the margins were scheduled instead. The founding article and the founding tradition converge on the discipline from opposite directions, and the convergence, rather than either camp’s platform, is the debate’s usable inheritance.

Why the argument arrived when it did

The date is not incidental, and a historical materialist reading of the concept’s own emergence is owed to a series of this kind. The founding article appeared at the end of a decade in which the enforcement of civil rights law in the United States had been retrenched, the labour movement had contracted sharply, and the socialist tradition that had carried the compound analysis in its organisational form had been reduced, by persecution and by history, to a fraction of its former presence. The compound analysis survived that decade by changing its carrier: out of the party journal and the movement pamphlet, into the law review and the university, institutions with their own protections and their own deformations. The change of carrier explains the concept’s double aspect ever since. The university preserved the argument through a period when its organisational carriers were weak, and the preservation was a genuine service, without which Jones’s line might have ended in the archive. The university also subjected the argument to its own selection pressures, rewarding the portable, the citable, and the administratively legible over the organisational and the remedial, and the additive version is the selection’s predictable product. To say this is not to indict the academy, which is a set of institutions behaving as institutions do. It is to supply the debate with the materialist premise both traditions claim to share: that ideas are carried by institutions, that carriers shape what they carry, and that a quarrel about a concept which ignores its carriers will mistake the history of its custody for the logic of its content. Much of the published quarrel is exactly that mistake, and it dissolves, like the others catalogued here, when the custody and the content are distinguished.

Frequently Asked Questions

Q: What is intersectionality?

It is the claim that systems of subordination operate jointly rather than in isolation, so that the situation of a person standing where several of them cross cannot be recovered by examining each system separately and summing the results. The term was coined for a specific failure of anti-discrimination law, which forced claimants to proceed on a single ground although their treatment had been compound. In its stronger form the claim is constitutive: the systems shape one another’s concrete operation, so that class in any actual society is organised through race and gender, and they through it. In its weaker form it is a reminder that disadvantages accumulate, which is true and much less than the founding argument said.

Q: Where does the term intersectionality come from?

Kimberlé Crenshaw introduced it in a Chicago law journal in 1989, in an article on how discrimination doctrine, and the movements built around that doctrine, pushed Black women to its margins. Her central exhibit was an employment dispute in which a last-in-first-out seniority rule wiped out every job held by the plant’s Black women, while the court declined to treat Black women as the class the claim concerned, the employer having taken on white women and Black men in earlier years. A sequel in 1991 carried the analysis into violence, service provision, and representation. The word names the crossing point where single-axis categories stopped describing the claimant in front of the court.

Q: Was intersectionality originally a legal argument?

Yes, entirely. The founding texts appeared in law reviews and speak to the doctrine governing discrimination claims, and what they propose is doctrinal repair: a claim compounded of race and sex should be hearable in its compounded form, the comparison classes of the existing case law should not be allowed to make it vanish, and the treatment of claimants at the crossings should be read as evidence about the categories themselves. Nothing in the original argument offers a general theory of society, a ranking of oppressions, or a political programme. Its later expansion into social theory was the work of sociologists and others over the following decades. Remembering the legal origin matters because the argument’s precision, its attachment to mechanisms and remedies, belongs to that origin and is what its generalisations most often lose.

Q: Who were the antecedents of intersectionality?

The decisive antecedent is Claudia Jones, who argued in a party journal in 1949 that the oppression bearing on Black women was compounded three ways at once: through the unprotected labour markets they were crowded into, through race, and through sex, in a combination that the unions and the feminism of the period each failed to take up. Behind the term’s other flank stands the Combahee River Collective’s statement of 1977, with its picture of mutually interlocking systems of oppression. Further back, Anna Julia Cooper’s book of 1892 stated the standpoint of the compound position philosophically, and W. E. B. Du Bois in 1935 analysed how a racial dividend held a divided working class together. The line is one of successive formalisations, not of a single invention.

Q: Is class an identity or a relation?

In Marxist theory, a relation. To be of a class is to stand in a determinate place inside the production process, a place that exists only against its counterpart: the seller of labour power is defined by the buyer who confronts her. An identity, by contrast, is an attribute a person carries between situations. The distinction governs the debate because a framework that lists class among a person’s attributes has converted the relation into a characteristic, and with the conversion goes the mechanism, exploitation, that made the category explanatory. Writers in the intersectional tradition who work constitutively do not need class to be an identity; their claim concerns how structures interlock, and a relation can interlock without becoming a trait.

Q: What is the Marxist criticism of intersectionality?

In its strongest attributed form, in Martha Gimenez’s work of 2018 and Eve Mitchell’s essay of 2013, the criticism is architectural. A framework that places class on a list of axes alongside race and gender converts a relation of exploitation into one disadvantage among others, and thereby loses the mechanism that connects the analysis to a theory of how the society reproduces itself. Such a framework can catalogue the coincidence of burdens but cannot explain the system producing them, and its natural political output is a demand for recognition and fairer representation within existing institutions rather than for changes in ownership. The criticism succeeds against the additive version of the framework and against its administrative usage. It does not reach the constitutive version of the founding texts.

Q: What is the intersectional criticism of Marxism?

It is historical before it is theoretical. Marxist movements and theory, the criticism runs, have a documented tendency to subsume race and gender under class and to schedule their claims for resolution after the primary conflict, a pattern visible from the socialist parties’ handling of women’s suffrage, through the labour movement’s internal racial hierarchies, to the neglect that Claudia Jones protested from inside a communist party in 1949. Angela Davis assembled the documentary record for feminism in her book of 1981, and the Combahee statement refused the schedule from inside two movements at once. The charge of class reductionism names this archive. Its strongest form concedes that the tradition also produced the instruments against the pattern, and asks why its institutions so reliably chose the schedule.

Q: What is the difference between additive and constitutive versions?

The additive version treats oppressions as distinct burdens piling up across a life, each complete in itself before it meets the others, so that a person’s situation can be described by listing and summing them. The constitutive version holds that the systems produce one another’s concrete forms: a labour market’s racial segmentation is not something added to its class structure but one of the ways that structure exists, and a seniority rule can fall on a position that hiring history and job segregation have already built for it. The difference decides the debate’s main questions, because the Marxist objection from the grammar of the list is decisive against the additive version and much weaker against the constitutive one, which is the version the founding texts defend.

Q: Can intersectionality and class analysis be combined?

Yes, and the combination exists in print. It requires two conditions: class enters the analysis as a structural relation and not as one more entry in an enumeration, and the intersectional contribution is kept in its constitutive form, a claim about how structures make one another, instead of a tally of burdens. Social reproduction theory, descending from Lise Vogel’s work of 1983 and renewed in the collection of 2017, performs the combination by treating the renewal of labour power as internal to the capital relation. Cinzia Arruzza argued for a unitary account in 2013, and Ashley Bohrer’s book of 2019 is a book-length reconciliation. Whether partisans of either tradition accept the result is a separate question from whether the combination is coherent, and the literature increasingly treats it as coherent.

Q: Is class reductionism a real tendency?

Yes. The phrase names a documented pattern rather than a debating insult: movements, prominently including Marxist ones, made a habit of conceding race and gender claims in principle while postponing their settlement until the main struggle had been won, a sequence whose convenience for the organisation’s cohesion, and for the safety of its doctrine, is visible in the record. The evidence is archival, in resolutions, memoirs, and the protests of those who were scheduled, from Jones’s essay of 1949 to the Combahee statement of 1977. The tendency is not unique to Marxism, since the women’s and civil rights movements deferred each other’s claims in the same way, but the Marxist archive is extensive and is not answered by denying the pattern’s existence.

Q: How does the relation and identity distinction grade a use of the term?

Ask what the word class is doing in the use before you. If class appears as a level of disadvantage, a quantity of hardship that a person or group can be assigned alongside others, the use has converted a relation into an attribute, and the analysis that follows will treat redistribution toward parity as the remedy, whether or not its author notices. If class appears as a position in a structure, defined by its opposite pole and carrying the mechanism of exploitation with it, the use passes the test whatever its author’s allegiance. The test also grades Marxist writing, since a Marxism for which the working class figures as a badge of origin and loyalty, rather than as a position in a process, has performed the identical conversion in mirror image and is vulnerable at the same joint.

Q: What should a student separate between the original legal argument and its later generalisation?

Three things. First, the object: the original argument concerns the failure of a body of doctrine to process compound claims, while the generalisation concerns the structure of social stratification as such, and arguments valid about one do not transfer automatically to the other. Second, the evidence: the legal argument rests on cases and on the behaviour of institutions under rules, while the general theory rests on sociological claims that need sociological support. Third, the remedy: doctrinal repair is a specific, achievable proposal, whereas the generalisation’s remedies are as large as its claims and correspondingly harder to test. A student who keeps the three separations can use the founding texts precisely and can identify the point in any later work where the legal argument’s authority is being borrowed for claims it never underwrote.

Q: Why did a term from law travel into social theory?

Because the failure it named was not peculiar to law. Any field whose categories are administered in departments, whose comparisons are built on single axes, and whose universal subjects are generalised from their least burdened members will dissolve its compound cases in the same way, and sociology, political science, education, and public health each discovered the dissolution in their own materials. The term travelled because the problem travelled. Its deformation in transit has the same explanation: a concept that began attached to mechanisms and cases was compressed, for portability, into a nod toward the thought that everything matters together, and the compression is what the additive version is. The travel and the deformation together explain the debate, since the critics mostly meet the compressed version and the defenders mostly cite the original.

Q: How does the two versions table order the debate?

By relocating every standard move in it. The table contrasts the additive and constitutive versions on how oppressions relate, how each version treats class, what the Marxist objection shows against each, and what each implies for analysis. Once a contribution to the debate is assigned its column, the applicable criticisms sort themselves: the objection built on the list lands on the additive column and misses the constitutive one, while the objection from the historical record of deferral lands on scheduling practice wherever it occurs. Most published exchanges in the debate consist of a criticism aimed at one column being answered from the other, and the table exposes the exchange as a mismatch rather than a refutation. It is an instrument for reading, not a verdict, and it grades this article’s own claims by the same columns.

Q: What separates a compound claim in law from a general theory of identity?

Everything that separates a remedy from a worldview. A compound claim in law is a pleading about a specific treatment: a claimant shows that a rule or practice bore on her as a member of a combined class in a way it bore on neither comparison class, and asks the court to register the treatment and repair it. The claim is about what was done, by an institution, through a mechanism, and its proof is documentary. A general theory of identity is a claim about who persons are and how their self-understanding is formed, and its proof, such as it is, is of an entirely different kind. The founding argument contains the first and not the second, and the confusion of the two, in which a doctrine about structural position is received as a doctrine about the self, is the single most consequential misreading in the term’s career.