The class-relation does not appear immediately in the categories utilized by the sociological understanding to group the population into various strata based on various criteria; when such groups have in the past affirmed themselves in the language of class, the negative dynamic of the class-relation has disappeared in the appearance of a particular positively-existing (quasi-)caste of the type which constituted the social structure of precapitalist modes of production based on direct domination and stable, material forms of wealth. Both ‘bourgeoisie’ and ‘proletariat,’ as they were traditionally understood, were caste-like political forms of appearance of the class-relation, with their own distinctive cultures, styles, institutions, etc.
Moments of the class-relation appear in caste-like shape predominantly when they are struggling against, and/or in the context of, a caste society: the great revolutionary upheavals of the modern period, whether the revolutionary forces had a predominantly bourgeois or proletarian composition, have been revolts of the third estate (in a broad sense) against the various ‘unproductive’ ruling estates or castes of premodern agrarian modes of production (and the historical and geographic peculiarities of these different caste systems and the struggles against them have determined the character of the capitalist state forms which subsequently emerged: versions of bourgeois-liberal, bureaucratic-stalinist, postcolonial-developmentalist, or something else entirely (see Japan, for example, in which capital accumulation was overseen by old-regime elites themselves).
The class-relation, the dynamic of economy immanent to the third estate, appeared in political shapes determined by the contest of exploiter and exploited to be the authentic representative of this estate as a whole in its struggle against the fixed and frozen castes (landholders, peasantry, clergy, military, sovereigns and their retainers), hence their remarkable confluence of rhetoric, values, aims, methods. The claim is one of productivity against parasitism, equality against privilege, novelty against tradition, enlightenment against superstition, a morally superior ‘people’ against a decadent clique. Ideologues and parties are sufficient to make such a political claim, their (rarely conscious) aim the reconstitution of the state in a form adequate to the requirements of the development of capitalist production relations in their particular historico-geographic context. The parties representing the propertyless of the third estate turned these claims around on its property owners, while subsuming to themselves the struggles of proletarians to survive.
Through this process the class-relation, the relation of capital accumulation, was driven on, and by the time the class-riven third estate as a whole had destroyed its hostile context of estates, and was able at mid-20th century to present itself as the whole of society, the class relation could no longer appear politically constituted as ‘two great camps’ within it. The dominant appearance was of a class society without classes, or rather, without castes: in Adorno’s words, “a parody of classless society.” In the era of the casteless class society, efforts to politically recompose ‘the proletariat’ or ‘the working class’ in estate-like shapes have had something comic about them, and always required personifying the opposite pole of the class-relation in pseudo-concretion as a conspiratorial, parasitic billionaire caste of fatcats (or Jews) which could be hated, fought, pleaded with, or replaced with the just representatives of the deserving.
While it is certainly true, for instance, that trade union apparatuses have long been an obstacle to the effective prosecution of system-immanent wage and conditions struggles, and workers must often fight without and against them, to hypostatize this situation as the demand for a politically-constituted class autonomy within this society has something inhuman about it, as it inevitably drags along an ideology that poses dehumanization, misery, and the meager means of coping with them as somehow more pure and noble than the easy living of the lazybones bossman. It also tends to obscure the fact that such struggles are less and less effective no matter how combative the workers, and are less and less possible even to undertake as profitable accumulation has ground to a crawl, and the class-relation is spinning apart: surplus capital blows its own bubbles at one end, and the devalorized surplus proletariat is lucky to be exploited (or more often, to find some unproductive wage slavery funded out of the total social capital) at the other. Class society appears as classless (which does not mean absent maldistributed misery) when it has become an anachronism.
Contained in the concept of the class-relation are the open antagonisms of interest obtaining within it, across (horizontally, as it were, to use a problematically spatial figure of speech) all the heterogeneous lives structured by it, from one pole to the other, not simply the hidden antagonistic relation of exploitation between the poles (so to speak, vertically), which plays out largely behind the backs of individuals, while leaving its scars. The only real unity of proletariat or capital respectively is in their respective internal separations, the complementary processes of competition and the multiplication of negative identifications through which the capitalist totality is reproduced. Caste-like political compositions conceal this nonidentity by forcing it under the aspect of identity with a positive image of ‘the class,’ or some substitute for it.
But this is not merely the result of the canny manipulation of class-affirmation-obsessed leaders cooking up ‘hegemonic figures,’ but once had a relatively organic basis in the historical situation of the struggles: a world of estates beats the class-relation into estate-like shapes, but a whole world constituted through the exchange abstraction silently identifies its poles, and they appear as bearers of so many ‘identities.’ Those who still insist on talking about class must talk about it like it’s an ‘identity’ too. But it is merely a negative dynamic which structures in relation to itself other negative dynamics (the hierarchical racializing constitution of abstract legal personhood, the dimorphic sexualizing scission of reproductive labor). The totalizing social form of capital is now their axis, and it is through its reproduction that these dynamics throw off reified images apparently identical with themselves.
If ‘class’ is bad at taking the shape of an ‘identity,’ that is not because its proletarian pole has failed to fulfill predictions and become so vast and essentially homogeneous it embraces nearly the whole globe, but because it has ‘succeeded’ in doing precisely that. Its ‘external enemies,’ the ancien caste-structured regimes, have been vanquished, and the capital to which it is subsumed has largely divested itself of private proprietors who endeavor to ‘direct’ or ‘manage’ their means of production. Capital accumulation itself has “expropriated the expropriators” (for those who care, this is entirely consistent with Marx’s passive-voice formulation, not to mention Engels’s startlingly accurate prediction of a motley lot of functionless beneficiaries of dividends with nothing to do but despoil each other on the stock market), liquidating an empirical group that could take the socio-cultural-cum-political form of ‘the bourgeoisie’ (a transitional moment in the emergence of world class society from a collection of caste-structured regimes).
Anyone bound to the positivist method of merely sorting individuals into the traditional class categories based on empirical attributes, including the somewhat ambiguous notion of ‘position,’ would indeed be confounded by wage laborers with a bit of money invested in the stock market, executives who do not own the firms they manage, desperate individuals who work for a wage by day, but by night carry on various ‘side-hustles’ in which they play the part of ‘entrepreneurs.’ But the class-relation stands behind all these roles which slice and dice up the determinate individuals through which it appears, but also vanishes. This situation is not some deviation from ‘the way capitalism’s supposed to be,’ caused by neoliberal meddling or whatever, but the outcome of its own immanent tendency — the objective polarization of the class-relation, accumulating wealth at one pole, and misery at the other, striating the mass in between with a staggering multiplicity of very real antagonisms, while at the same time it becomes essentially homogenized in its dependence on the exchange nexus and the vicissitudes of the accumulation process.
To set about recomposing for the sake of system-immanent politicking some kind of unified representation of ‘the proletariat’ or ‘the working class’ out of this sea of separations would be just another version of the project of ‘making a people,’ and it’s no surprise that such projects these days are virtually indistinguishable from populisms and nationalisms of various stripes. These ‘working peoples’ always become fixed as their own sectional interests ranged against the interests and needs of determinate proletarians, as the real antagonisms therebetween cannot simply be fused in a political semblance. But to seize on other sectional interests, building up a political semblance around them, or to try to quantify and aggregate every claim to suffering in some kind of intersectional frankenstein’s monster more quickly than it can tear itself apart, are only other ways of aggravating the problem. Its internally antagonistic unity-in-separation belongs to the concept of the proletariat, and the antagonisms are all the more heightened as proletarianization itself increasingly cuts unevenly across particular individuals in their world context.
This is not a new problem, though it has come newly to define class struggle in the past 50 to 100 years (depending on one’s reckoning) since the estate-like political forms (the strength and relative cohesion of which having been facilitated by, and often based upon, preestablished bonds and factors not directly related to the condition of proletarianization as such, though the then-dominant technical composition of the labor processes was also important) began to break down. It is actually the early stages of this breakdown process that threw up workers’ councils, whatever their limits and weaknesses, as proletarians attempted to confront their differing needs and problems that were not included in the ‘consciousness’ held and imputed to them by the Party representatives of the proletariat as an estate. It was even seen by a few of the farsighted communists like Pannekoek then that struggling proletarians “whose interests are still far from being homogeneous” were strengthened by confronting their real disunity, rather than having their unity presented to them as already given, over and against themselves: “it is within the proletariat itself that the resistances develop which it must overcome; and in overcoming them, the proletariat overcomes its own limitations and matures towards communism;” “its task will not become clear as the result of enlightenment from above; it must discover its tasks through hard work, through thought and conflict of opinions. It must find its own way; therefore, the internal struggle.”
Today, the system has become so total, yet so decrepit and irrational, that the old opposition of ‘sectional needs’ to ‘class needs,’ to whatever extent it once had some strategic meaning, has become a kind of aporia that shatters thought. Every attempt to pose a class need is shown up as a sham as proletarians somewhere else, or a year down the line, suffer from it. The abolition of the class-relation, that is, the capitalist totality, is the only possible need that could now embrace the proletariat as a whole. But the totality cannot be confronted immediately, as it, the all-penetrating ether of society, mediates every confrontation. A radical ideology set immediately against everything fizzles out in its own abstract, gestural extremity. The old process of relating sectional to class needs when it appeared possible to pose the latter within the bounds of the capitalist system was that of a class-constituting consciousness, a consciousness of the class as a quasi-estate. It seemed that the class had its own positive existence in-itself, which could outlive the sloughing off of the parasitical exploiters (who were thought of as simply another caste to dispose of, like the feudal lords, to expose the healthy, productive ‘people’ beneath), thus a kind of aggregate or typical consciousness could be devised which would affirm this class and promote its ‘hegemony.’
But just as wherever there is capital, it accumulates a proletariat, wherever there is a proletariat, it accumulates capital. This double movement has swallowed the whole world, and in so doing produced its own obsolescence. The mutual internal dependence of capital and labor is now asserting itself as an insoluble problem for both poles. When proletarians run aground on the world class society itself as a universal block, as well as grope after whatever is not simply delusion in its ‘classless’ appearance — the objective anachronism of the value-relations, of the reified exchange of living for dead labor — there results a multiplicity of class-transcending consciousnesses, which remain inchoate and subordinated to their own real conditions’ semblance of social necessity: we are not ‘deceived’ by ideology but are ideology materialized.
Thus class-transcending consciousness becomes false and apologetic when it is taken as referring to an existent positivity; rather its truth is the content of the immanent critique of the class-constituting consciousness and its practices. Possibilities of that critique becoming practical are determined by the antagonistic context of the historically developing crisis: the only way out of the dialectical context of immanence is through that context itself.
 Some, most notably Théorie Communiste, have attempted to theorize this historical situation as one in which “class belonging now appears as an external constraint,” while it used to be something affirmable. But actually, ‘class belonging’ used to appear as belonging to a discrete, potentially self-sufficient estate whose class-relation with capital appeared as an external constraint, as merely oppression by an unnecessary ‘high-born’ caste of snooty parasitic loafers, squatting atop the backs of the working people, which simply had to be deposed. This is what G.M. Tamás calls ‘rousseauist socialism,’ which sees only impositions that need to be removed, whereas, when it comes to the class-relation, the capital-labor relation, proletarians will have to “liberate themselves from themselves,” as Marcuse put it. Now, with the disappearance of the proletariat’s caste-like shape, ‘class belonging’ not only does not appear as ‘an external constraint,’ it barely appears at all, or it appears as generalized separation (which it is), because, properly speaking, one does not ‘belong’ to a class (like one belongs to an estate). Class is a lived abstraction. Class-affirmation (which, contra TC’s optimistic claims, certainly continues, as witnessed by the recent wave of neo-social-democratic labor nationalism) can only ever be a misrecognition of the class-relation as a caste or estate stratification — its critique would entail demonstrating how this misrecognition had a material basis in an earlier era of capital accumulation, but doesn’t any longer. TC’s theory can’t do this, nor can any which remains hobbled by their framework. This is in part because, beneath all their ‘anti-humanist’ fulminations, they retain something of the ‘rousseauist’ approach. Despite seemingly stressing the mutual imbrication of capital-labor, they do so only to return, via their historical schematic, to the problematic of ‘external constraints,’ whereas in truth, there is no longer anything external in this totally socialized society. And in TC's vision, the throwing off of ‘class belonging as an external constraint’ takes the form of another machine-wrecking fantasy (the proletarians simply destroy all that ‘external’ junk that ‘defines them as proletarians’), which issues in immediate relations and the ‘de-objectivation of the world’ — none of that vexingly complex, richly differentiated human social life, just an undivided state of nature with the unreflected general will springing up organically. But the class-relation, the capital-labor relation, the relation between dead and living labor, is a self-relation — the self-development of not-yet-humanity turned upside down and against itself, and its determinate negation, communism, would be, in the words of C.L.R. James, “self-developing humanity taking the place formerly held by self-developing value.”