This essay began as a guest talk at the UIC Department of English in October 2023. It was intended as an expression of (at points highly critical) support from an intentionalist position for Vivek Chibber’s then-recent attempt at reviving something like a base-superstructure model as a valid tool of Marxist analysis. Over time,1 it grew into a slightly larger review of the possibilities, and the implications, of aligning certain elements of Marxist theory with the good picture of intention. Because of this, although the notion of base and superstructure remains central throughout the argument, in the end the piece is composed of two somewhat distinct parts. Sections 1–6 present the main body of the argument on the relationship between the base/superstructure and the good picture of intention (with short detours into Voloshinov in section 3 and Althusser in section 5). Sections 7–8 provide an extended coda of sorts, where the focus is on the logical entailments of intentions rather than intentions as such.

1. Base and superstructure I

The proposition of “base and superstructure” might just be the single most genuinely controversial concept in Marxism. Unlike with other ill-reputed Marxist concepts and metaphors—think the “dictatorship of the proletariat”—the controversies surrounding it are not rooted in a simple misunderstanding, a wilful misreading, or a case of particularly poor phrasing or bad translation. Conversely, it doesn’t seem to turn less controversial once placed within a broader theoretical framework or once its actual technical meaning becomes clear. For one, it is famously difficult to work out what the technical meaning of “base” and “superstructure” is and whether Marx himself even thought about the two as analytically useful notions. As such, the concept has been highly controversial from the very nascence of Marxist theory, with Engels himself quickly retreating to a safe theoretical position of base and superstructure mutually affecting each other.2 For crucial thinkers of Western Marxism, from Althusser3 to Adorno4 to Williams,5 the readiness to cushion the base-superstructure model with a series of philosophical caveats was the measure of one’s opposition to “vulgar” Marxism; a nuanced understanding of the relationship between base and superstructure, or perhaps a relative weakening of its ominously central role, was to safeguard Marxism as a progressive nexus of theory and practice from the monstrous errors of Stalinism.

But the notion of “base” and “superstructure” is also particularly resistant to any theoretical attempts at pushing it out into the margins of a broader Marxist framework. Unlike many other controversial concepts within the same tradition, it is not the kind of a niche technical notion that only poses potential problems to those already highly invested in the complexities of the theory. In fact, for all its simplicity, the idea of a cultural-ideological “superstructure” being causally (and at least in some sense unilaterally) influenced by a certain economic “base” resonates for many—both within and without Marxism—with their intuitions about what forms the necessary core of any coherent Marxist anthropology, theory of history, or indeed politics. And although it is undoubtedly true that Marx himself did not leave us with a clear definition of what constitutes either the “base” or the “superstructure,” it is equally true that whenever he came anywhere close—most famously, in the canonical formulation in A Contribution to the Critique of Political Economy6—he clearly implied he was laying the foundation for a new account of social change, where various cultural and ideological forms adapt more or less by default to the more fundamental shifts in the mode of production.

Equally crucially, the validity of the base-superstructure model seems today a matter of some political urgency, as showcased by Vivek Chibber’s 2022 book The Class Matrix7—or, more precisely, as much by the book itself as by certain polemic reactions to it. Although Chibber avoids using the notions of base and superstructure other than in his account of the historical development in Marxist thought,8 his own theory is clearly a variation of the model rather than an alternative to it. Proposing a strong distinction between the “structural” and the “cultural,” Chibber defends the causal primacy of the first as the only safeguard against the Gramscian “culturalism” of the post-war Left: its tendency to locate the entirety of politically important conflicts and contradictions within the sphere of ideology, its choice to focus on culture as the locus of political agency, and its willingness to essentially dismiss the political importance of class struggle while nominally acknowledging its crucial structural role. Chibber reminds his readers that, although the class struggle is obviously never contained just to the “base,” its source—the class antagonism, as it were—lays at a level distinctly deeper than any cultural issue of the moment. His point is not that one needs to uphold some abstract materialist dogma in order not to give any ground to idealists and culturalists but that for Marxists specifically culturalism is a kind of a slippery slope: even acknowledging the kind of feedback loop between the base and the superstructure that Engels famously imagined as the solution to the model’s potential issues can lead very quickly to abandoning our commitment to the political primacy of class struggle—in favour of a politics that sees itself as confined entirely to culture and a vague fight for ideological hegemony.

Curiously, at least some of Chibber’s more vocal polemicists chose to explicitly defend the latter approach. Rather than argue, for instance, that a certain version of culturalism does not necessarily lead to abandoning the centrality of class struggle and class analysis, theorists such as Michael McCarthy and Mathieu Desan juxtapose Chibber’s alleged “class abstractionism” (a vaguely rebranded version of “class reductionism”) with their own model of “class dynamism,” whose entire point is “to suggest that the political primacy of class does not necessarily follow from an account of its structural primacy.”9 Crucially, their argument is not that certain cultural identities may be, in any given historical context, good correlates of or approximation for class—a claim that would be perfectly compatible with Chibber’s model. Rather, they seem to believe that whereas the mode of production as such is indeed structured primarily by the class antagonism, this doesn’t necessarily translate to any theory of crisis or social change and hence doesn’t imply the practical-political importance of class-based organising; apparently, although the world we live in is indeed structured by an inevitable and ongoing conflict between distinct classes, a fundamental change in this state of things may come about through forces that have nothing to do with that conflict at all. Skipping past the obvious question about the usefulness of class analysis in such a scenario, McCarthy and Desan use the common-sense observation that the working class is often if not always structured by various internal conflicts to propose that categories of ascriptive difference, such as race, are somehow parallel to, rather than a part of, the political framework defined by class struggle. (In fact, as Chibber rightly points out, the fact that members of the working class have inevitably contradictory interests is implied by the very nature of that class—i.e., its members’ inevitable reliance on the labour market—and is precisely the reason why socialist organisations should help the workers mitigate the individual cost of acting in the collective, class-based interest).

This rhetorical sleight of hand—the working class is forced to compete against one another, and so socialists can focus on intra-class conflicts instead of class struggle—might be par for the course among the self-proclaimed opponents of class reductionism/abstractionism and would be easily exposed by the introduction of value into the mix (a notion that is curiously missing from Desan and McCarthy’s analysis), as it would highlight the fact that the only reason capitalist labour market exists is the extraction of the surplus value itself, and thus intra-class competition among the workers is a function or a tool of class struggle, not somehow parallel to it. But ultimately what makes McCarthy and Desan’s polemic interesting is that rather than offer a convincing alternative to Chibber’s model, they more or less prove his point; they show that the forceful decoupling of the “cultural” from the “structural” serves, first and foremost, to push class and class analysis aside into the margins of theory.

2. The good picture of intention

All of the above means that those of us who find themselves in a somewhat unenviable—since theoretically challenging—position of believing both in the essential validity of the base-superstructure model (perhaps in a modified version, such as Chibber’s) and in the so-called intentionalist10 theory of action face a potentially troublesome contradiction. After all, at first glance the two may seem basically incompatible.

What I mean by the intentionalist theory of action is, in short, a theory (or simply an argument) derived from the fundamental philosophical claim that intent, instead of being either a mental state accompanying an action or its external cause (what Stanley Cavell calls the “bad picture” of intention11), is in fact immanent to what is intended; more than that, it is not just a part of an action but specifically the part that makes the action into what it is as a certain legible totality. Supported by a certain reading of Elizabeth Anscombe’s Intention,12 the theory was largely developed out of a set of claims about the nature of meaning and authorship found in a series of articles by Walter Benn Michaels and Steven Knapp published in the 1980s,13 and it has far-reaching implications for one’s account of—if not necessarily one’s practice of—interpretation, as its main tenets apply not just to intentional actions but equally to intentional works. The radical discontinuity between intents and causes—what Jennifer Ashton called the “categorically distinct” nature of causes and meanings14—indicates that a causal account of how any meaningful work came to be can never replace the interpretation of said work in the sense that the discovery of meaning is by definition done in the course of interpretation and can never be achieved through (or reduced to) the analysis of various processes which contributed to the work’s emergence. Or, to put it slightly differently, the only way to distinguish something like meaning from the infinite pool of factors that causally led to its creation is by invoking what is actually expressed in the work as the intention of its author (which ultimately means abandoning the criteria of causality altogether). (Which also points, as we will soon see, to the fundamental asymmetry of the problem at hand: the challenge posed to a causal account by the presence of meaning simply does not arise from the standpoint of interpretation, where what is relevant from among the plethora of causal factors that contributed to the existence of the work is only made so by the meaning itself.)

Crucially, the fact that interpretation is never reducible to a causal account doesn’t stem from any practical problem with doing interpretation or even any epistemological one; rather, it is an ontological issue, with the non-interchangeability of the two rooted directly in the nature of intention itself. If intention remains always immanent to what is intended—and not just as one of its many parts but as its structuring principle—then it is only recoverable through and as a part of the structure itself, and such recovery, by definition, already constitutes interpretation. In other words, if there is an epistemological issue at play here, it is an incredibly trivial one: in order to look at the thing, one has to look at the thing.

Now, it has already probably become apparent why reconciling such an account of intention with the base-superstructure model may potentially pose a problem. Seen in a certain way—perhaps the default, most intuitive way—the base-superstructure metaphor seems to encourage exactly the kind of “reduction to causes” that an Anscombean account of action dismisses as impossible (and that Ashton calls a “causal fallacy”), and vice versa, a reader of Anscombe may perhaps approach culture and ideology in a way that Marxism would see as crudely idealistic. Again, this is not just a practical issue to do with the institutional reality of doing criticism or research—what is at stake is the validity of both sides’ claims about the relationship between the meaningful (or the intended) and the causal.

However, as we will shortly see, whereas a certain reading of Anscombe may pose a challenge to the base-superstructure model, it also ultimately provides a solution to the problems otherwise inherent in it. In fact, if historically one of the arguments against the model was that it implicitly abolished any meaningful autonomy of culture, literature, and art, what the intentionalist theory of action shows is that it is precisely the asymmetrical (and, yes, dialectical) nature of the base-superstructure relationship that safeguards the only kind of autonomy that doesn’t ultimately collapse in on itself as a purely theoretical construct. In addition to that, a reconciling of the Anscombean model with the notion of base and superstructure may help us understand the limit of our own critical practice—in the sense of natural limits to what might be considered a “Marxist interpretation” of art, literature, or ideology.

3. Voloshinov’s “superfluous men”

Although the scepticism about the base-superstructure model—rooted in its supposedly “vulgar” nature—may be seen as one of the hallmarks of Marxist high theory in the West, it hasn’t necessarily always been the case elsewhere. Valentin Voloshinov’s 1929 book, Marxism and the Philosophy of Language15—famously one of the first coherent theories aimed at putting linguistics and philosophy of language firmly on a Marxist foundation—touches on the issue in a distinctly different way. It’s important not to overstate its insights: today, Voloshinov’s theory of sign—in a way his main accomplishment—is certainly outdated and clearly incompatible with an Anscombean account of meaning and action. The main reason it may still be of interest is its underlying attempt to resolve or overcome a contradiction between two opposed modes or levels of criticism. On the one hand, Voloshinov says, every sign being “not only a reflection … of reality” but also “itself a material segment,” “a phenomenon of the external world,” has to lend itself by definition to “a unitary, monistic, objective method of study” (MPL, 11). On the other hand:

Any ideological product is not only itself a part of a reality (natural or social) … it also, in contradistinction to these other phenomena, reflects and refracts another reality outside itself. … The physical object is converted into a sign. Without ceasing to be a part of material reality, such an object, to some degree, reflects and refracts another reality. (MPL, 9)

The dual nature of the sign means that meaning, although in itself “not a thing,” only exists through certain material objects and as such is inconceivable independent of them (MPL, 28). What’s at stake in getting both sides of the relationship right, for Voloshinov, is as much the defence of the historical-materialist project—which relies on the possibility of a “unitary, monistic” method of study of the material world, essentially a causal account of what just is—as the preservation of something like interpretation or criticism, which requires an establishment of a normative field and a separate mode of study. This is because various elements of what is meaningful—a novel, a painting, or even ideology itself—cannot be derived directly from some corresponding processes in the world (this would amount to an embrace of the vulgar, “mechanical” causality, Voloshinov’s rejection of which already foreshadows Althusser’s much later criticism of “linear” and “expressive”causalities):

No cognitive value whatever adheres to the establishment of a connection between the basis and some isolated fact torn from the unity and integrity of its ideological context. … Even if the correspondence established [between an element of a novel and a particular social fact] is correct, even if it is true that [the nineteenth-century trope of] “superfluous men” did appear in literature in connection with the breakdown of the economic structure of the gentry … the correspondence established itself remains without any cognitive value until both the specific role of the “superfluous man” in the artistic structure of the novel and the specific role of the novel in social life as a whole are elucidated. (MPL, 18)

In other words, the elements of a work have to be mediated through the notion of a meaningful, bounded, essentially formal totality in order to appear as meaningful themselves—or really to appear as anything at all. We only know that the novel touches on the “superflous men” once we’ve read the novel. And once we start reading the novel, we’re no longer in the domain of the purely causal: the moment aesthetic or ideological totalities become subject to analysis, they are already posited under normative categories, which a simple causal nexus would always logically undo.

This is where, in Voloshinov, the notion of the base and the superstructure comes in. Any causal analysis of something like ideology or a work of art—that is, any attempt to grasp the superstructure from the standpoint of the base—is doomed to fail unless it understands the need for such mediation through a formal totality. In this way, a consistent account of the base implies the existence of a superstructure. This is certainly an undertheorised moment in Voloshinov’s thought, and he might not have put it exactly this way, but we could perhaps suggest that under his account, the notion of the superstructure is a way in which the causal mode of analysis acknowledges its limits—points to its blind spot, as it were, or brackets itself—while preserving its claim to coherence (as the causal primacy of the base over the superstructure is never subverted). This is why, while for some Marxists the base-superstructure model implies an abolishment of art’s autonomy, for Voloshinov the very same model explains why (or perhaps in what way) superstructures are never reducible to their base—and thus makes autonomy possible.

4. Chibber and the backdoor culturalism

With this insight in mind, it might prove interesting to return to Chibber and try to show why such a problematic matters today—and why even those who (like myself) consider The Class Matrix one of the more convincing critiques of culturalism from a Marxist standpoint published in recent years should still try and reconcile it with the kind of radical discontinuity of cause and intent posed by the intentionalist theory of action.

Going beyond the narrow scope of McCarthy and Desan’s polemic, it bears reminding that The Class Matrix’s main argument consists of two essential claims, one historical and one theoretical. The historical argument considers various developments on the Marxist left since, roughly speaking, the turn of the twentieth century—its increasing focus, or indeed reliance, on cultural politics—which Chibber ties explicitly to “classical” Marxism’s inability to account for the fact that the successive economic crises of the nineteenth and twentieth century never brought about a socialist revolution in developed Western economies. Looking for an explanation of capitalism’s surprising stability, Chibber’s narrative goes, Marxists turned to culture as an external stabilising force—a political deus ex machina of sorts—whose influence on workers and voters could allegedly override the structural contradictions that were originally supposed to bring about capitalism’s downfall (chief among them the class antagonism itself). This is where Chibber’s theoretical argument picks up: rather than turn to culture, the Marxist left should appreciate the role of individual market competition in stabilising capitalist economies; that is, it should accept that by default, in a typical labour market, the worker is more incentivised to act in their individual rather than collective (and potentially revolutionary) interest, as normally prioritising the latter implies too much risk and too many costs. This means that structural contradictions of capitalism do not have to be “overridden” or camouflaged by a reactionary culture in order for the market to stabilise itself; rather, what acts as the stabilising factor is of the same structural nature, or exists at the same level, as the contradictions themselves. In other words, both the sources of capitalism’s instability and the forces that ultimately assure its survival stem directly from the capitalist labour relation itself.

Chibber’s argument opens up a path for an account of class politics that sees short-term cultural conflicts as both symptomatic of underlying economic contradictions and unambiguously subordinate to them. And while his model certainly emphasises the internal tensions within the working class itself, it avoids rooting those tensions in some form of identity politics, focusing rather on the general structural contradictions generated by the market competition itself—and the kind of self-interested individualism that it ultimately encourages. This allows Chibber to offer political solutions that are decidedly anti-identitarian and universalist in nature: the task of the Left would be to mitigate the material risks and costs incurred by the workers whenever they decide to act in the collective rather than individual interest (with strike funds being a model example of such mitigation measure).

Chibber’s account is powerful insofar as it asserts the political primacy of class but also because it draws our attention precisely to those structural contradictions and inequalities that remain independent of, or indifferent to, any particular subject-position: “I am not aware of any instance in modern history where a transition to capitalism was derailed, or even significantly delayed, by the inability of social actors to understand what wage labor meant.”16 Throughout the book, he consistently and convincingly articulates the political advantages of a methodical study of “the structural.” However, Chibber’s book itself betrays a contradiction that its author is ultimately unable to solve—and this contradiction is precisely where we can see the importance of reconciling the base-superstructure model with the Anscombean emphasis on the discontinuity between causes and intentions, or something like Voloshinov’s approach to ideology.

In chapter four, devoted to issues of agency and contingency, Chibber specifically confronts the fact that the contents of at least some cultural practices seemingly cannot be derived purely or immediately from the class antagonism itself. This poses an obvious problem for the very foundation of his explicitly anti-“culturalist” approach: if, under capitalism, culture is restricted by the underlying socio-economic structures—that is, first and foremost, the class antagonism—and if it is indeed, as Chibber says, a strictly one-way relationship (with culture incapable of ultimately overriding “the structural”), then why is it that specific cultural forms cannot seem to be derived from the class struggle itself or cannot be reduced to an account of it? Chibber’s answer is to allow these forms to develop independently of “the structure” insofar as they are not in an open conflict with it—that is, he says, the structure forces changes in the culture when they are necessary for the former to survive, but otherwise the latter remains largely autonomous. What this odd kind of autonomy means in practice, however, is only that there must be forces other than class that determine the cultural forms of any given historical moment.

The scope of influence of economic structures is thus limited, and part of the research program of class analysis is to investigate the actual boundaries of that influence. How far does it extend into the political or ideological institutions of society? Which components does it subordinate to its logic? Those that remain outside its scope will have causal histories of their own. They are not undetermined—they are just not determined by the class structure. Or, to put it differently, they are not purely contingent, they are just contingent with respect to the class structure.17

This might seem like a satisfying, indeed common-sense answer to the problem of agency and cultural autonomy. However, what it does in the context of Chibber’s more fundamental points is that it essentially creates a backdoor for the return of “culturalism.” For if actions are indeed always culturally mediated (which Chibber allows and even insists on) and cultural forms may be determined by external forces of the same general quality (if not necessarily importance) as class, then it is hard to see what prevents these other forces from becoming, under certain circumstances, the driving force behind collective action of political or even revolutionary nature. In other words, it suddenly seems perfectly possible for culture to at least sporadically override the allegedly fundamental “structural” conditions posed by Chibber. And insofar as the relationship between the two is politically contingent and the difference quantitative rather than qualitative, “at least sporadically” can easily become “possibly always”—it is not hard to imagine that within such framework, at any particular time, at some level of granularity, some factors other than class will appear as more determinative of the overall social structure. Chibber’s otherwise excellent account collapses in on itself as the primacy of class becomes—in practice—an inconsequential abstraction, a theoretical fiat, precisely contrary to The Class Matrix’s stated goals.18

Seen this way, it is an issue that cannot be easily fixed. No class analysis, no matter how precise or exhaustive or “dense” in its description, will provide the answers of the kind Chibber requires, and to invoke causal factors other than class only delays the issue, so to speak—no collection of causal accounts will ever account for all of the existing meanings because, to go back to the intentionalist theory of action, on a fundamental level interpretation cannot be reduced to a causal account, and meaning cannot be reduced to a sum of the work’s causes. Even the admission of something like intentions (agency? free will?) to a list of causal “agents” would not bypass the need for interpretation. Which is just to say that insofar as intention is not simply contained in the mind but rather remains immanent to what is intentional (e.g., a work of art), no account—psychological or sociological—of how it came to be may replace the interpretation of a meaningful totality of which it is part.

5. Althusser on autonomy

In fact, what Chibber proposes at this specific point is surprisingly close to the essential Althusserian concept of relative autonomy with all its analytical baggage—curiously, however, he gets to this point from an angle almost opposite to Althusser’s. For the latter, the notion of relative autonomy was famously born out of his attempt to preserve the idea of economic determination “in the last instance” within a model whose stated goal was to offer an alternative to Stalinism and vulgar economic determinism more broadly. In other words, while Chibber settles on relative autonomy as a concession to culturalism within a broadly economist framework, Althusser accepts it as a concession to economic determinism within a broadly culturalist framework.

At its core, Althusser’s model relies on a differentiation between determination (in the last instance) and domination (within any given social structure or mode of production):

It is economism that identifies eternally in advance the determinant contradiction-in-the-last-instance with the role of the dominant contradiction, which for ever assimilates such and such an ‘aspect’ (forces of production, economy, practice) to the principal role, and such and such another ‘aspect’ (relations of production, politics, ideology, theory) to the secondary role—whereas in real history determination in the last instance by the economy is exercised precisely in the permutations of the principal role between the economy, politics, theory, etc.19

Norman Geras explains the difference in the following terms:

[T]he economy determines for the non-economic elements their respective degrees of autonomy/dependence in relation to itself and to one another, thus their differential degrees of specific effectivity. It can determine itself as dominant or not-dominant at any particular time, and in the latter case it determines which of the elements is to be dominant.20

Gregory Elliott puts the point even more forcefully:

Every social formation is a ‘structure in dominance’ insofar as each contains a dominant element which organizes the hierarchy and interrelations of the various social practices. This dominant element, however, does not override the causal primacy of the economic. The economic is not always dominant, but it is always determinant in the last instance, responsible for the coordination of the instances, allotting the dominant role within the totality to one of them and subordinate roles to the others, fixing their degree of relative autonomy and efficacy.21

“Relative autonomy” thus denotes the kind of autonomy “the economic” gave its assent to; it is very explicitly not the kind of “independence”22 one would associate with the breaking of a causal chain.

Of course, such an approach leads to several issues. For one, for all of Althusser’s explicit rejection of pluralism and assertions that within his model of “structural causality” everything does not cause everything else, it may often seem that this is in fact very much the case. Hence the accusations of “fatal circularity”23 or, in more straightforward terms, wishful thinking supported with tautologies: “Though the provision of concepts with which to think the effectivity of a ‘structure’ on its elements and subordinate structures and all their effects is presented by Althusser as a problem, it is hard to see in his attempt to resolve it any more than a restatement of the question.”24 But even vocal proponents of Althusser sometimes inadvertently admit this point: when Robert Resch, in his extensive recapitulation of Althusser’s theory and its impact on Marxism, finally offers a single graph representing the unified model of the “modes of determination in contemporary capitalism,”25 the first thing to draw the reader’s attention is inevitably the fact that various arrows connect every node of the network to every other one, usually in a bilaterally symmetrical way. We’re left, once more, with a feedback loop (or perhaps now a series of loops).

More broadly speaking, the model’s original sin might be its abstract nature. For all the effort by Althusser and his followers to give the theory some empirical weight, it remains painfully unclear what exactly differentiates autonomy (from domination) from full independence (from determination). The role played by “the economy,” on the other hand, might almost be compared—at the risk of anthropomorphising it—to some sort of a legal-political trick, a tactical retreat by an unpopular ruler: the productive forces and the relations of production do not govern the structure directly on a day-to-day basis, but they still reserve the right to choose who governs it and when—preserving all the power while remaining out of the spotlight. It is perhaps a good metaphor for the fact that the link between “structural” and “cultural” may be complex and in some way indirect, but it doesn’t exactly carry much analytical weight.

What all of this means for the idea of relative autonomy is that it becomes a profoundly unstable notion. One can tip it in the direction of economic determinism and admit that it’s really an illusion produced by the sheer complexity of the causal chains, but in such case the problem faced by Chibber (that is, the obvious irreducibility of certain cultural forms to structural processes) remains unsolved. Or one can go in the opposite direction and accept that even relative autonomy from domination implies in fact—at least functionally and in practice—genuine autonomy from economic determination and decouple the superstructure from the base. This can lead to a position similar to Desan and McCarthy’s or, in Althusser’s case, a redefinition of class itself, the prioritising of ideological hegemony over other forms of political power, and a late-life return to voluntarism.

6. Base and superstructure II

And it is precisely here that the intentionalist theory of action comes into play and helps us find a dialectical way out of the problem by producing a position that can fully embrace both the fact of autonomy (absolute, rather than relative), as well as the causal determination of the superstructure by the base.

In more concrete terms, it means accepting that meanings are causally linked to Chibber’s “structure” while also embracing the fact that it doesn’t make them available to us through a causal account—and that this is indeed, maybe somewhat counterintuitively, at least as much an ontological issue as it is an epistemological one. The meaning of every single work of art and every single ideological statement may indeed be fully determined by the underlying economic (or “structural”) processes and relations—and this doesn’t change the fact that this meaning by the nature of intention itself will only ever be available through interpretation (and as such remains irreducible to any causal account). In fact, its very existence already implies the need for a normative account and as such appears, from the standpoint of the “structure,” as something qualitatively different from it. When faced with such a realisation, the only thing causal analysis can do is acknowledge it, and when it does, it can—vide Voloshinov and unlike in Chibber, who only accounts for half of the problem—posit something like the existence of a superstructure. It could perhaps be said that the notion of superstructure is the way in which a causal account—an account of the base or Chibber’s “structure”—points to its own outside, an operation that remains, to emphasise once more the fundamental asymmetry of the problem at hand, unnecessary from the standpoint of meaning and interpretation, from which what is relevant about the outside is never outside at all.

When Chibber encounters the issue of the irreducibility of intentions to causes, he is right to abandon some of his previous intellectual commitments; the problem is that he abandons exactly the wrong ones. Insofar as his model doesn’t account for the fact that the meaning (or “culture”) cannot be explained fully by a causal account of its links to the “structure,” it has to be modified. But Chibber focuses on preserving the absolute authority and sufficiency of the causal account and in doing so surrenders the primacy of class analysis, which now has to be supplemented by an undefined number of other causal accounts. In a way, he chooses cause over class. However, rather than follow in his steps, it is possible to do the exact opposite: admit the ultimate insufficiency of causal accounts (in the sense that they can never replace interpretations) and preserve the primacy of class analysis within the domain of the causal.

The insufficiency of causal accounts should not be taken as a proof of their inherent incompleteness (which may still be the case but for different reasons). Rather, the point is that even a complete causal account of an otherwise intentional “thing” will never be sufficient to explain it; in other words, from a certain point of view it will always be irrelevant to interpretation, or interpretation will always be indifferent to it.

The implications of all this for Marxist theory seem by now quite obvious—what might be less obvious is what the stakes are from the standpoint of interpretation. And insofar as nothing in the above changes anything in the nature of interpretation, there are indeed none. The complexities of the relationship between the base and the superstructure may change the way one does causal analysis, but it doesn’t impact the nature of interpretation—be it interpretation of literature, art, or ideology. (Another way to put this would be that from the standpoint of the interpretation, the non-identity of cause and intent never even appears as an issue to be solved.) In this sense, the problem is not symmetrical. However, it still has some implications for the practice of criticism more broadly, mostly in that it alleviates a certain fear, a distrust of a causal-materialist analysis among those who perhaps see the latter as reductive and indifferent to the normative claims of the field. Insofar as a kind of mutual indifference is the point, however, it is clear that there is nothing in such an analysis—in striving for a causal account that is as coherent, as exhaustive, and even as totalising in its aspirations as possible—that poses a challenge or any danger to interpretation or autonomy. The latter, once achieved, is always full and in some sense absolute rather than relative. (Another way to put this would be that a full causal account of a work’s creation is perfectly compatible with it being fully autonomous.)

Once dialectically “fixed,” Chibber’s model reveals at least one more interesting implication: a suggestion that the base-superstructure metaphor itself could be rewritten in terms of causes and meanings. It would now appear not as an inevitably arbitrary division between various human practices, types of activities, or material processes but as a more ontological relationship based on both continuity and discontinuity between certain processes (such as relations of production broadly understood) and the meanings that may ultimately emerge from them—a relation, in other words, between the wage labour and any meaning it may or may not have for someone. This is a crucial point worth emphasising: if at the end of the day the notion of base and superstructure shouldn’t be seen as a way of dividing social activity into distinct spheres of varying importance, then it’s not just because it describes a certain relation instead but because the relation in question emerges as more fundamental and universal than any particular mode of social organisation.

Seen this way, the base-superstructure model doesn’t even have to necessarily privilege class and class antagonism as the primary components of the “base.” It could hypothetically be reconciled with any set of beliefs about what constitutes the primary driving force in human history. Of course, Marxists usually argue for the causal primacy of the relations of productions and the productive forces, but this is a claim about the nature of the base, not its existence as such.26

7. Homologies and transitions

Any attempt at aligning the base-superstructure model with the good picture of intention—that is, the account of intention as immanent to what is intended—will have significant implications for more specific areas of Marxist theory and criticism. Although there’s obviously no point in trying to review them all here, two questions seem to require immediate attention: the status of homologies and the so-called “real abstractions.”

Homologies are often seen as a practical necessity, if not the philosophical precondition, for any historical-materialist criticism worthy of the name (or indeed any socially- and historically-oriented criticism in general); after all, if one cannot draw structural parallels between the political and the aesthetic, what is at stake in such criticism? Meanwhile, the Anscombean disconnect between causes and intents may seem to preclude the use of homologies altogether, leaving room only for strictly causal analysis on the one hand and interpretations of particular works on the other. Is it possible, within such framework, to find a place for homology as a distinct mode of analysis?

It’s important to note that the problem at hand is very different from that identified by Fredric Jameson as the main issue with homologies; in fact, if anything, the good picture of intention provides an immediate solution to the latter. According to Jameson, homologies inevitably rely on something external to the objects of interpretation in that, unable to explain the source of identity, they need to invoke a universal device that would give legitimacy to the act of comparison, something of which the compared objects would just be a particular instance: “in order to secure that indifference or nonhierarchy of the various subsystems, an external category is required, that of ‘structure’ itself.”27 Thus, insofar as they present the objects of comparison as mere versions or expressions of something existing outside and independent of them, homologies repeat the error of Althusser’s “expressive causality”—that is, they imply an account of history as a progression of superficial forms shaped by an external force of a higher order. “[A] sequence of historical events or texts and artifacts is rewritten in terms of some deeper, underlying, and more ‘fundamental’ narrative, of a hidden master narrative which is the allegorical key or figural content of the first sequence of empirical materials.”28 The original sin of structuralism, the problem of homologies is perhaps even more present in the so-called new formalism where forms themselves become not just an analytic principle allowing one to link together various “subsystems” (a sonnet, a TV show, a political party, a mode of monastic life) but the proper matter of politics itself and—in all but name—something similar to a force of nature: they pre-exist actions and limit them through “affordances,”29 a property they share with physical materials and natural resources.

In this sense, homologies seem to reduce the aesthetic form back to a link in a causal chain, presenting its particular instances as simply determined by the original pattern (the “master narrative”).30 But this is once again where the intentionalist understanding of the relationship between cause and intent becomes instructive. If at its root the problem with homologies is that, unable to otherwise explain the source of identity of the compared objects, they rely on an account of an external “thing” that is logically prior to and independent of any particular interpretation (you need to know what a structure is and what it “does” causally in order to notice it at work at different levels of analysis), then this is precisely the problem that the good picture of intention solves in the sense of preventing it from arising altogether. What determines the form is the intent, which is in itself a part of, rather than external to, what is intended, and insofar as one needs any account of it at all (in order to explain the source of identity), such account is simply implied by every act of interpretation. In other words, by showing that intention is immanent to the act, the Anscombean account relocates Jameson’s “external category” or “master narrative” within the logic of the particular form itself.31 All of which is just another way of saying that in order to recognise two meanings as identical, one doesn’t need anything beyond interpretation: two things can mean the same thing, in full or in part, because anything can be used to carry any meaning (the relation between the signifier and the signified is in each instance arbitrary).32

However, the problematic instances of homologies only emerge as such whenever they seem to bridge the gap between different ontological levels—by linking something clearly meaningful (a work of art, an ideologem) to something we don’t necessarily think about in terms of meaning or intention (a mode of social organisation, an economic tendency). And here, ostensibly, they may seem as incompatible with the notion of base and superstructure: it’s not hard to accept that two people could mean the same thing on two different occasions (it could even happen on accident, that is, without any causal connection at all); it’s harder to accept that what either one means could be somehow “identical” to the logic of a monetary system. To limit homologies to their unproblematic instances would be to deprive them of analytical usefulness (no one denies the possibility of comparative interpretation), while to abolish the distinction between problematic and unproblematic homologies seems to risk eradicating “the base” altogether by turning everything—including all social life—into a series of aesthetic-like forms.

Here the problem of homologies turns into the problem of what is sometimes dubbed transitions in the context of intentionalist criticism; as Michaels put it in The Gold Standard, “How can we theorize the connection between trompe l’oeil painters and hard money Democrats, much less the connection of any of them to the novel?”33 One possible answer is that normative judgements about acts are not limited to the latter’s organising intentions—that is, interpretation—but can extend to the logical structure of the entailments, implications, and necessary presuppositions that intentions produce. As Nicholas Brown puts it:

Since every intentional act can be described in terms that are nowhere to be found in the moment of intention—think of the endless descriptions of jumping over a railing, turning a tap, or boiling water in the “Ithaca” episode of James Joyce’s Ulysses—nothing in the analysis of meaning to intention prevents us from chasing down what a meaning might entail as a logically necessary consequence (as opposed to an effect) or condition of possibility (as opposed to a cause), even if these are not intended.34

And of course, under the account of intention as immanent to what is intended, this claim extends far beyond art and literature and into actions as such. In fact, this is one of the major lessons of the good picture of intention—not just works of art but all actions may be said to have logical entailments. This makes possible claims of structural identity not between the meaning of different aesthetic and social forms but between their shared entailments. If it seems far-fetched to propose that a monetary system means the same thing as a certain type of novel, it is much less preposterous to suggest that such system, when remembered as a product of collective action, relies on the same presuppositions or abstractions as a particular novel, its structure both maintained and implied by the aggregate social action. In this sense transitions or homologies would constitute a momentary recovery of an intentional account of what otherwise presents itself as purely causal. Only momentary because it doesn’t so much produce a “social form” as a kind of thing in the world but as another type of description: an account of what we usually take to be socially causal as originally entailed by action. (“We are all literalists most or all of our lives.”35)

8. Real abstractions and the political unconscious

All of this leads naturally to the problem of “real abstraction,” a notion which seems to pin down something important about action in Marxist theory—that is, the apparent existence of a class of “things” that at first glance seem to belong neither to the base nor the superstructure even though they are directly rooted in social activity. Alfred Sohn-Rethel, who is usually credited with the popularisation of the term, used it to denote just a single real abstraction—that is, the “commodity or value abstraction,”36 which originates with the abstraction of exchange-value from the object of exchange. Such abstraction, of course, is logically necessary for the very existence of the capitalist market and for the money to function at all as a medium of exchange; however, throughout the introductory part of Intellectual and Manual Labour Sohn-Rethel struggles precisely with the question of what it means to say that exchange-value (and ultimately commodity value) exists in reality. It clearly cannot only be present in the minds of the buyer and the seller, much less an economist or a sociologist who observe the exchange; it cannot be reducible to ideology or institutions or law but has to be directly rooted in actions. Still, this realisation doesn’t save Sohn-Rethel’s account from a certain contradiction:

The essence of commodity abstraction, however, is that it is not thought-induced; it does not originate in men’s minds but in their actions. And yet this does not give ‘abstraction’ a merely metaphorical meaning. It is abstraction in its precise, literal sense. … [T]he economic concept of value is a real one. It exists nowhere other than in the human mind but it does not spring from it. Rather it is purely social in character, arising in the spatio-temporal sphere of human interrelations. It is not people who originate these abstractions but their actions. [As Marx says in the first volume of Capital,] ‘They do this without being aware of it.’ In order to do justice to Marx’s Critique of Political Economy the commodity or value abstraction revealed in his analysis must be viewed as a real abstraction resulting from spatio-temporal activity. (IML, 16–17)

Even though “the economic concept of value” doesn’t “spring” from the human mind, it still “exists nowhere other than in” it. There is clearly a step missing here, a point at which the real abstraction exists in the world before it enters the mind—otherwise, how could it be said to “originate” somewhere else in any sense other than that in which all concepts refer to something? Sohn-Rethel remains torn between these two ideas—of value as a mental and non-mental entity—but ends up leaning towards the latter—that is, the concept of real abstraction as something rooted directly in activity and fact (“resulting from spatio-temporal activity”):

[T]he exchange abstraction is, first, a real historical occurrence in time and space …. (IML, 18)

Wherever commodity exchange takes place, it does so in effective ‘abstraction’ from use. This is an abstraction not in mind, but in fact. (IML, 21)

Thus, in speaking of the abstractness of exchange we must be careful not to apply the term to the consciousness of the exchanging agents. (IML, 21)

In fact, one of the preconditions of the real abstraction is that it is not something done consciously or indeed intentionally by the actors of exchange in the sense that, under their own description, they have to be driven by use-value (even if on the part of the seller what counts is the market demand, rather than the use-value his commodity could have to himself):

The individualised consciousness also is beset by abstractness, but this is not the abstractness of the act of exchange at its source. For the abstractness of that action cannot be noted when it happens, since it only happens because the consciousness of its agents is taken up with their business and with the empirical appearance of things which pertains to their use. One could say that the abstractness of their action is beyond realisation by the actors because their very consciousness stands in the way. Were the abstractness to catch their minds their action would cease to be exchange and the abstraction would not arise. (IML, 22)

Because of this, Sohn-Rethel goes as far as to say that “the action alone is abstract” (it is also “social,” while the “minds are private”) (IML, 24). His struggle with the apparent contradiction echoes that of Marx in Capital; to go back to the fragment referenced by Sohn-Rethel, “whenever, by an exchange, we equate as values our different products, by that very act, we also equate, as human labour, the different kinds of labour expended upon them. We are not aware of this, nevertheless we do it.”37 Of course we do many things without being quite aware of what we do; this is no more than to say that we are not always conscious of our own intentions. But clearly what both Marx and Sohn-Rethel mean is something else entirely: the idea that what our actions bring about goes beyond their result in the conventional, or causal, sense (the result of the exchange being the goods changing hands), as well as beyond the intention expressed in the action, and that this peculiar side-effect isn’t located in the mind itself. Of course, if we locate intention in the mind (which Marx himself seems at times prone to,38 but this is an issue for another time), at this point we will be at a loss: what the buyer and the seller think might have certain implications, entailments, preconditions, etc.—it may produce a concept of exchange-value—but the point has been precisely that they do not have to have an account of what the exchange is from the standpoint of commodity form (indeed, according to Sohn-Rethel it’s necessary that they do not think about what they are doing). However, as soon as we acknowledge that intention is not located in the mind but in the act, the problem all but disappears: whatever is implied is just implied by the action itself. Exchange-value is what has to exist logically in order for the capitalist exchange to make sense—that is, to happen as an action rather than a purely physical movement of objects.

The assertion of the non-mental status of real abstraction is a step that even some of Sohn-Rethel’s followers crucially miss; Alberto Toscano, for instance, for all of his detailed account of the history of real abstraction as an idea,39 goes on to extend it to religion,40 thus blurring the line between real abstraction, ideology, and social construction. Such operation is, again, an inevitable gateway to political culturalism: if any mass belief can be seen as equally causally determinative as the law of value itself, then “the base” can be reshaped as a matter of cultural hegemony. However, to recast the real abstraction as something in any way similar to a social construct is to substitute a bad picture of intention in place of the good one by trying to derive the real existence of a certain force purely from what social actors believe in (while not equating it exactly with their belief), as if the sheer mass popularity of some beliefs turned them into something ontologically different. Crucially, whereas all claims about social constructs can be reduced to claims about someone’s beliefs, this is not the case with real abstraction, which requires a judgement on what is entailed by actions instead. This is why, for instance, “races” do not exist even though “categories of ascriptive differentiation”41 in an important sense do.

Of course, to criticise the kind of unreflexive extension of the notion of real abstraction proposed by Toscano is not to say that it should be limited to Sohn-Rethel’s original example—that is, commodity exchange. Rather, what Sohn-Rethel did by designating it as the real abstraction is both highlight its peculiar status as a logical implication of social actions and assert its structural significance for capitalism as such. Hence it might be more precise to call commodity exchange the foremost of a much larger set of real abstractions organising market societies. In fact, in order to grasp the problem, one doesn’t need to invoke real abstraction at all. When Brown notes, similar to Sohn-Rethel, that the abstraction of exchange-value is decidedly non-mental in nature—“The logical contortions embodied in the act of exchange … are nowhere in the mind of the capitalist. Rather, they are the logical preconditions of the act of exchange itself”42—he links this observation to the Jamesonian notion of political unconscious instead:

In this Hegelian-Marxian sense, the unconscious is simply everything entailed or presupposed by an action that is not present to consciousness in that action. Such entailment is often, in the Phenomenology of Spirit, an action’s necessary interaction with the universal in which it subsists. Such interactions yield a properly Hegelian mode of irony: think, for example, of the fate of Diderot’s sensible man in Hegel’s retelling of Rameau’s Nephew (or think, in our time, of the “outsider” artist) confronting a culture of wit that necessarily turns every attempt at plain truth telling into its opposite. An intention necessarily calls such necessary presuppositions or entailments into play. (The identity of intention and meaning insists upon a political unconscious.)43

Of course, relying on the unconscious in a broadly Freudian sense would pose the danger of once again placing form and meaning at the level of causality; in fact, this is what psychoanalysis in criticism and literary theory often does whenever it assumes that certain formal structures are reproduced on a different level of human activities because they persist unconsciously in the mind of specific actors (or are perhaps universal to humans as such). And certainly on many occasions such continuity may in principle exist (the possibility of a transition/homology certainly doesn’t preclude objects from being otherwise causally linked). However, Sohn-Rethel’s point is that there is a way in which the emergence of “things” such as value is irreducible to a purely causal account, as well as anyone’s beliefs or even intents themselves. (In fact, it seems plausible that at least one purpose of Sohn-Rethel’s insistence that the participants of an exchange be actively preoccupied with something other than the exchange itself is to emphasise that it is not enough for them to just push the notion of exchange-value deep enough into their unconscious; rather, it has to be analytically unavailable.) Brown acknowledges this problem, and puts forward a negative—Hegelian rather than Freudian—account of the political unconscious:

It may be that in Jameson’s work, the Freudian positivity of the unconscious is relatively inconsequential and can be rewritten in terms of the negative, Hegelian-Marxian unconscious, but it would take some work to ascertain. When Jameson writes about class-consciousness in Wyndham Lewis, for example, the point is that petty bourgeois class consciousness logically presupposes working-class consciousness and is unnecessary and unthinkable without it, and that Lewis is not aware of that entailment and presumably would have disavowed it. It does not mean that some secret part of Lewis’s brain is aware of that entailment. Any Freudian “return of the repressed” would then have to be understood instead as the Hegelian “ruse of reason”—that is, as an example of the fact that logical entailments are real entailments. The claim here is not that Jameson never relies on a positive unconscious but that work that follows his lead would be better off working with a negative one.44

This Hegelian reworking of the political unconscious into a “real entailment” has the effect of bringing it closer to real abstraction by specifying the former as explicitly non-mental in nature; it also clarifies—or potentially fixes—an important step in Jameson’s original account. It suggests that the notion of the political unconscious could be hypothetically expanded beyond art and ideology and into actions as such; just like in the context of homologies/transitions, there is no reason for the “entailments” in question to be limited to something like literature or art or distinct and interpretable ideologems, as such entailments are “called into play” whenever and wherever intention is present. In this sense, exchange-value itself could be seen as a form of political unconscious—with the latter just one possible way of accounting for the existence of undoubtedly real but ultimately neither causal nor mental nor fully intentional products of social actions.

If the base-superstructure model is to be seen as an account of how meanings, from art to ideology, exist in a world that is otherwise causally determined by the relations of production and the development of productive forces—an account which violates neither the principle of this determination nor the irreducibility of interpretation to any causal analysis—then such model not only leaves room for something like real abstractions and the political unconscious or homologies/transitions as a mode of analysis; it makes them all necessary as a way of accounting for the fact that all acts, including collective social action, have logical entailments. It is, so to speak, the final step in rethinking the notion of base and superstructure as a fundamental social-ontological relation, rather than an arbitrary division between historically contingent areas of social life.