On the Charge of "Class Abstractionism"

Ben Burgis

30 April 2025

This is a transcript of a talk I gave on February 24th at the Caucus for a Critical Political Science (CCPS) conference in South Padre Island, Texas, based on a paper I’m co-writing with Michael Sechman.

Socialists who are critical of identity politics are often accused, even by other socialists, of something called “class reductionism.” But, what exactly does that mean? Taken literally, “reductionism” about some phenomenon means thinking it can be completely analyzed in terms of some other underlying phenomenon. To be finicky and precise about this, an A-reductionist believes that instances of A are “nothing above and beyond” instances of the underlying B. So, for example:

• In the philosophy of science, “reductionism” about the special sciences is the position that biology can be “reduced” to chemistry and chemistry can be “reduced” to physics in the sense that ultimately facts about biology and chemistry are nothing above and beyond certain kinds of extremely complicated facts about basic physical particles.

• In the philosophy of mind, “reductive” physicalism is the view that mental states are nothing above and beyond certain bits and pieces of physical reality, perhaps involving brains or behavioral dispositions.

Taken literally, “class reductionism” about things like race or gender would be the view that facts about race and gender are nothing above and beyond facts about class. And, sure, some trends in Marxist theory do point in that direction, at least to a degree. Socialist-feminist “social reproduction theory,” for example, seeks to use the categories of Marxist political economy to understand gender roles, and “caste” theories often feel like a way of trying to turn race into a sort of semi-class category. But, first, these aren’t the kinds of views that anyone is railing against when they throw around that charge of “class reductionism.” Second, it’s deeply unclear what consequences, if any, even a literal full-on “reductionism” about these questions would have for debates around socialist political strategy. By analogy, an epidemiologist who’s studied a little philosophy of science and come away a convinced “reductionist” about the special sciences, fully committed to the claim that the viruses she studies have no metaphysical status above and beyond the protons and electrons that make them up, would be neither more nor less likely as a result to recommend masking in response to a bird flu outbreak. It’s just an unrelated issue.

Writing in the journal Sociological Theory, Michael McCarthy and Matthiue Desan criticize the same socialists who are often accused of “class reductionism,” but they do so in a new way. Abandoning the “reductionism” charge as a conceptual mess for more or less the reasons we’ve just gestured at, they instead accuse these socialists of “class abstractionism.” This is supposed to be the misguided attempt to infer from the premise that class is “structurally primary” in terms of a correct analysis of the workings of capitalist society to the conclusion that class is “politically primary” in terms of socialist strategy.[1]

McCarthy and Desan have no quarrel with structural primacy. In fact, something I’ve seen McCarthy say a couple of times is that in order to be a Marxist you have to believe in some version of structural primacy, however exactly you cash out that concept. But they reject political primacy, for reasons suggested by a quote they give in a footnote from David Roediger’s book The Wages of Whiteness. “If,” Roediger writes, “racism is a large, low-hanging branch of a tree that is rooted in class relations, we must constantly remind ourselves that the branch is not the same as the roots, and that the best way to shake the roots may at times be by grabbing the branches.” Commenting approvingly on this passage, McCarthy and Desan say it suggests that class is “more fundamental, more basic” than race in an explanatory sense but that it’s not “more important” in a political sense.

Everyone in this debate has been massively influenced by the late great Erik Olin Wright, who famously loved to illustrate his concepts and distinctions with 2x2 diagrams, and McCarthy and Desan duly throw one of those into this essay.

If you believe that class is neither structurally nor politically primary, then you’re a class “relativist.” If you believe that it’s not structurally primary but it is politically primary, you’re a class “constructivist,” which is a position they vaguely associate with non-Marxist left populisms that lean on language about “the 99%.” They don’t say much about either of those possibilities, so we won’t either. Instead, their point is to critique class “abstractionism” in the name of the approach they take to be superior, which is class “dynamism.” Class abstractionism, according to the 2x2, is the view that class is both structurally and politically primary, while class dynamism affirms structural but not political primacy.

And the first thing we want to call your attention to is that, without quite seeming to realize it, McCarthy and Desan have two very non-equivalent definitions of “class abstractionism” in play here. If we take the 2x2 literally, you can be a class abstractionist without trying to derive political primacy from structural primacy. On the other hand, if you take the entailment definition seriously enough to ignore the way the 2x2 is set up and assume that the bottom-right cell is supposed to represent abstractionism in the political-primacy-derived-from-structural-primacy sense after all, then the 2x2 is no longer exhaustive of the possibilities. And this isn’t just nitpicking. It never seems to occur to McCarthy and Desan that someone working with Marxist class categories might think class is politically primary for any reason other than a mistaken belief that this is entailed by the causal/explanatory primacy of class in our analysis of capitalist society. We’ll return to that point in a moment.

While there are some concerns about what they mean by “structural primacy”[2] (perhaps we can get into that in the Q & A), what I want to focus on now is an important ambiguity in what they mean by “political primacy.”

To help see the problem, let’s go ahead and bring in a second essay, this one written by William Clare Roberts for Crisis & Critique. In it, Roberts endorses and further develops McCarthy and Desan’s critique of “class abstractionism.” McCarthy in turn has effusively praised Roberts’s article.

The only alleged “abstractionist” named and shamed in the original McCarthy and Desan paper is Vivek Chibber. Roberts widens the net, applying the label not just to Chibber, but to Adolph Reed, as well as the authors of various articles published in Jacobin magazine, including, I’ll just pause to awkwardly note, this one.

Roberts ignores the 2x2 definition of abstractionism and endorses the entailment definition. He identifies the general target of the critique as “a new discourse of class” that’s been “ascendant since 2016,” of which the foremost representative is Reed. A couple paragraphs later, he says that, apart from Reed, “the most prominent” class abstractionist is Vivek Chibber. At this point, he gives us the only textual citation in either paper that even looks like an instance of an “abstractionist” actually trying to deduce political primacy from structural primacy. So, we should pay close attention to it. Here's the quote:

The working class has this power [to overcome the resistance of the capitalist class and its political functionaries], for a simple reason – capitalists can only make their profits if workers show up to work every day, and if they refuse to play along, the profits dry up overnight. […] It is this power to extract real concessions from capital that makes the working class so important for political strategy.

Now, this does indeed look like a smoking gun of entailment-definition abstractionism. Chibber is arguing for the importance of the working class for socialist political strategy on the basis of the structural role of the working class within the capitalist machine. But notice Roberts’s bait-and-switch in the next two paragraphs.He writes, at the end of the paragraph immediately following the quote:

The intuition…is that organizing around racial, gendered, or other identities, and around injustices of status and standing, is a distraction from what is really going on, and, for that reason, is also a distraction from what could actually work.

Then, in the very next paragraph, he says:

The structural primacy of class can be fleshed out as the claim that the fundamental social processes that drive, undergird, and explain conflicts over status and standing and identity are the class processes of capitalism in its current neoliberal form. The political primacy of class can be fleshed out as the claims that naming and appealing to people’s material class interests is both more motivating and more inclusive than naming and appealing to people’s status, standing, and identity, which are particularistic and divisive, rather than universalistic and unifying.

But hold on now. If this is what Roberts means by structural and political primacy, then no smoke is coming from the smoking-gun quote after all. Nothing Chibber says in that passage has anything to do with class processes driving conflicts over identity, nor do we find Chibber even broaching the topic of how best to organize the working class against capitalism.This brings us to the giant ambiguity at the heart of all the talk about “political primacy” in both the original McCarthy and Desan paper and the Roberts follow-up.

Are they talking about class being politically primary in the sense that the working class is the agent of historical change that has the capacity to bring about socialism (or, even short of that, the capacity to bring about important reforms)? Or are they talking about “naming and appealing to people’s material class interests” as a strategy for mobilizing them?

The same ambiguity shows up in the McCarthy and Desan paper. When explaining how abstractionists do the abstracting, they accurately summarize arguments like Chibber’s from structural premises about the location of the working class within capitalism to the conclusion that the working class is the necessary agent of historical change. But then when they define their terms, in order to explain why class abstractionism is misguided, suddenly we find them defining political primacy in terms of “the priority of class subjectivity in political mobilization.”

The difference is crucial because McCarthy, Desan, and Roberts very explicitly agree that the working class is the necessary agent of historical change. If believing that, and arriving at it from standard Marxist structural premises about capitalism, makes one an abstractionist, than McCarthy, Desan, and Roberts are “class abstractionists” no less than Reed or Chibber or, God help us, brocialist Jacobin writers such as myself. When McCarthy and Desan write that they’re making a contribution “within” Marxism, part of what they mean is that they still believe that the proletariat will necessarily be the agent of its own emancipation. The sense in which they deny “political primacy” is that they deny that “political mobilization” is best served “in the first instance” by the kind of political strategy endorsed by Reed, Chibber, those assorted Jacobin writers, and so on. Indeed, they’re often at pains to insist that political struggles fought on the terrain of race, gender, and so on count as a kind of class struggle no less than workplace-based organizing or political movements to achieve universalistic social-democratic demands like Medicare for All.

For example, in a slightly hysterical social media thread a week ago accusing me and Vivek Chibber of “smearing” him on my podcast, McCarthy characterized the core difference between dynamists and abstractionists as being about what “form” the political mobilization of the working class will take and his belief that contemporary identity-based struggles are themselves a potentially promising form of class struggle.

And in the original paper, he and Desan characterized these “other solidarities that make up the texture of workers’ lived experience” as “the raw materials upon which class formation must work…”

That’s the sense in which they’re denying the political primacy of class. And they simply don’t have a lick of textual evidence that Reed or Chibber or the Jacobin-bros believe that political primacy in this sense is abstractly entailed by any sort of structural primacy. That’s just a straw man.

And, since we’re nearing the end of our time, let me just wrap up with several general points about what Michael and I think about all of this.

First, we agree that, if we’re operating at the level of theoretical abstraction where we’re just looking at these two propositions:

• Class-based material explanations have ‘primacy’ in the correct explanation of the dynamics of capitalism (or even in the correct explanation of various particular forms of social injustice disproportionately experienced by marginalized groups).

• In order to achieve socialism (or even make the biggest possible strides in overcoming those other injustices) we should pursue a political strategy in which we primarily appeal to people on the basis of broadly-shared material interests.

…the first proposition doesn’t entail the second one.

Second, though, we do think the first proposition gives us a reason to believe the second. If B is a problem we’re trying to solve and A is the thing we think primarily explains how B arises, then these facts do give us at least some reason to look for possible A-based solutions to B. That much is true. But it’s a defeasible reason. Sometimes the cure for a disease has nothing to do with what caused it in the first place. If you chain-smoked for twenty years and you get cancer, then quitting tomorrow won’t obviate the need for surgery or chemo. Hell, sometimes cures aren’t on the table at all, and the best you can do is suppress the symptoms. That seems to be roughly the “Afropessimist” perspective on race-based oppression, for example. So, again, if we’re operating at the level of abstraction that McCarthy, Desan, and Roberts accuse abstractionists of operating on, the first of the two propositions I just mentioned doesn’t come close to entailing the second. It does, however, point us in the general direction of the second as a place to at least start looking for possible cures.

Michael and I also agree that whether the best socialist strategy revolves around a heavy emphasis on workplace-level organizing, and at least primarily trying to appeal to people on the basis of broadly shared material interests even in electoral contexts, is something that has to be hashed out on a more concrete level. But we’d point out that this is exactly what Reed, Chibber, and the rest of the alleged “abstractionists” actually do. None of them are content to derive their preferred political strategy from nothing more than big structural facts about class. They all make concrete strategic arguments. That’s certainly what I do, and in fact it’s all I do, in the article of mine that made it into Roberts’s citations. And Adolph Reed’s popular-level political writing for decades has conspicuously centered around strategic arguments for his preferred strategy of relying, as Roberts correctly characterizes the view in the Crisis & Critique article, on “appealing to universal ideals and material interests.” Reed’s argument against raising the demand for racially targeted reparations, for example, is rarely made in any way that uses any sort of abstract Marxist terminology. Instead, he’s doggedly pointed out, over and over again, that if our goal is to redistribute resources to poor and working-class black people, allying with poor and working-class white people by demanding universalist solutions is a far more promising strategy than trying to go it alone. Any attempt at achieving large-scale redistribution, whether universalistic or racially means-tested, would run up against the resistance of deeply entrenched interests. As Reed has argued over and over and over again, deploying his characteristic razor-sharp acidic wit, there’s simply no realistic scenario where we could mobilize a majority of the population to overcome that resistance in the name of a program that would only benefit a minority. That’s an independent strategic argument.

And in fact, the bit of my conversation with Vivek Chibber that McCarthy complains about in that thread cuts off just before Chibber’s answer to my question. In that answer, Chibber infers the correctness of his preferred strategy not from any sort of high-level theoretical abstraction but from a straightforward appeal to the historical evidence.

Chibber says McCarthy is saying, as “a thousand other” left intellectuals have, that “non-class factors can be just as important.” And sure, Chibber grants, “can’t is a strong word,” there could be “certain circumstances” where grabbing the branches could be the best way of shaking the tree, “but as a rule, typically, no.” As “a rule,” he thinks it’s generally a better idea to organize people “as workers not people of a certain color or people of a certain gender.” And he throws down the gauntlet that “if you doubt” that organizing people as workers is a more effective way of advancing working-class interests than organizing on the basis of various identity categories, you need be able to identify “one instance historically” where it’s worked out the way that the so-called dynamists think it will.

Matt Huber makes similar points about effectiveness-in-practice in his response to McCarthy and Desan in Damage magazine, where he points out that the conception of political “mobilization” they’re talking about is oddly disconnected from the logistics of actually building or exercising power in order to achieve concrete political ends. If it’s a question of whether identity-based messages can succeed in getting large numbers of people to show up at protests, for example, then who denies that? But this is a very different thing from achieving socialism, or even Medicare for All, or even tackling the problem of police violence, for which, as Cedric Johnson argues in his book After Black Lives Matter, the primary thing we’d need to do is to “abolish the conditions” that lead to high crime rates and the blunt-force response of militarized policing in the first place.

Perhaps McCarthy, Desan, and Roberts can give compelling responses to all these concrete strategic arguments. But in order to even be in the right discussion, they’d need to engage with the actual points that living breathing people like Adolph Reed and Vivek Chibber are making and stop fighting with straw “abstractionists.”

---

Ben Burgis is a Jacobin columnist, an adjunct philosophy professor at Rutgers University, and the host of the YouTube show and podcast Give Them An Argument. He’s the author of several books, most recently Christopher Hitchens: What He Got Right, How He Went Wrong, and Why He Still Matters. 

---

Notes

[1] They seem to take it so thoroughly for granted that this is what “abstractionists” think that they’re doing that they repeatedly take it as a damning criticism of Vivek Chibber that they can show that, “in effect,” he hasn’t really inferred political primacy from structural primacy after all.

[2] Since I’m freed here from the time constraints of a 20-minute conference presentation, here’s what we wanted to point out about structural primacy:At one point, McCarthy & Desan seem to endorse a definition pulled from Erik Olin Wright, Andrew Levine & Elliott Sober (henceforth WLS)’s excellent book Reconstructing Marxism. After mentioning Althusser’s claims about the economic being “determinant in the last instance,’ they say that "the coherence of Althusser's oft-maligned formulation need not detain us here" because there are other, more plausible ways of cashing out “the basic claim that class structure is unique and fundamental relative to the social structure as a whole.” But they only actually mention one other way:”For example, Wright, Levine, and Sober (1992) concede that variation in noneconomic phenomena is irreducible to class but nevertheless posit a ‘causal asymmetry’ in which class structure determines the limits within which noneconomic forces exercise their determination.”And in the follow-up article by Roberts discussed below, he pretty much restates this definition of structural primacy (although he omits the attribution to WLS):

“Class is structurally primary, for Marxists, because production is primary. Class relations organize production, and, since production is fundamental to the existence of human society, solving the class relation problem is a limiting constraint on everything else that goes on in society.”

There are three problems with all of this. The first is that McCarthy and Desan fundamentally misread WLS. In the long chapter in which they discuss possible senses of class primacy (see here for my own thoughts on that chapter!) they argue explicitly that merely establishing this kind of “qualitative asymmetry” between the class structure and other factors isn’t enough to establish the structural primacy of class.The second is that McCarthy and Desan (and Roberts) also seem to endorse a sense of structural primacy that goes far beyond this, but none of them seem to realize the disconnect between the “definition” telephone-gamed from WLS and these vastly more ambitious notions of “structural primacy.” I give some textual evidence for the more ambitious notions below in the transcript, but I’d add here that (a) McCarthy and Desan’s own positive case for their “dynamist” view relies on the claim that various identity-based political struggles are downstream of the dynamics of capitalist economic structures, and Roberts goes so far as to make the frankly bizarre claim that “culture war” is nothing but “the lived experience of class struggle,” a claim that only makes sense if the dial of base/superstructure determinism is turned to its maximum setting.This brings us to the third problem, which is that Vivek Chibber (the only named “abstractionist” in McCarthy & Desan’s paper and one of Roberts’s main targets as well) very explicitly doesn’t believe in anything even close to such an explanatorily ambitious sense of the “structural primacy” of class. In The Class Matrix he only argues that the class structure is structurally primary relative to those bits of culture that are necessary for the continued existence of the class structure itself. (Think: Workers’ internalization of norms about how to behave and dress at work, how to interact with managers, etc.) Chibber is very explicit that, while the capitalist class structure is self-perpetuating in this sense (mute compulsion powerfully prods people to internalize the necessary cultural scripts for integrating themselves into the class structure, and hence the class structure is not hostage to more culturally contingent facts about any particular society) it’s also fully consistent with a dizzying variety of cultural forms that can develop in ways that are quite autonomous from it. Adolph Reed and Walter Benn Michaels seem similarly skeptical of the value of turning the dial of base/superstructure analysis all the way up. All of which is just to say that, from the evidence, the alleged “abstractionists” are (to their credit) far more measured and cautious about what class and economics can explain.

---