In the pages of Cosmonaut a discussion has been sparked by the recent book Flowers for Marx related to the scientific status of Marxism. In my original review, I critiqued the authors for erroneous misreadings of Althusser, Marx, and in general for a deference to the epistemic authority of the bourgeois social sciences (in the case of the Jacobin crowd) and the epistemic authority of the party line (in the case of the more Marxist-Leninist-leaning authors). Conrad Hamilton, of the latter group, wrote an essay responding to many of my points. If I was to provide a summary of Hamilton’s essay, it’d probably be as follows: the epistemic authority of the bourgeois social sciences (the neoclassical economists, the empiricist historians) is deserved, and it is impetuous amateurism to suggest otherwise, whereas it is the epistemic authority of the bourgeois hard sciences (engineering, even physics) which needs to be overthrown or surpassed. Needless to say, I find this to be absolutely backwards.

Regarding Althusser, Empiricism, and Abstraction

Hamilton alleges that Althusser’s early work suffered from a lack of appreciation for the empirical as a sort of “theoreticist” deviation and that Althusser later corrected this deviation. What I wished to point out was that Althusser perfectly acknowledged the role of empirical data and facts having a relationship as the foundation of abstract structures, and that these abstract structures causally affect each other, rather than an abstract structure causally affecting empirical phenomena (e.g. economic structures directly affecting political structures rather than economic structures directly affecting political empirical phenomena). Hamilton’s response to this seems to be quite confused:

Althusser’s point here is that mechanical theories of historical materialism—as with the one found in Stalin’s Dialectical and Historical Materialism, though it goes unmentioned here for obvious reasons—simply ‘remix’ the approach of standard historians. Standard historians take for granted the primacy of ‘homogeneous-continuous/self-contemporaneous time,’ making this ‘empirical temporality’ preside over the ‘concrete’ contents of history (and thereby failing to recognize ‘the different temporalities produced by the different levels of the structure’—the way time functions differently within different modes of production, for instance). Similarly, within mechanical historical materialism the ‘economic’ is made the index of the ‘non-economic’ (politics, culture, etc.), in a way that fails to acknowledge any process of mediation (and thus the ‘(relative) regional autonomy’ of the non-economic).

This issue of the economic being erroneously made into the index of the non-economic is precisely what I’m talking about! That is the conflation of structural causality for empirical causality. What Hamilton does not understand becomes clear when he says: “Where [mechanical economic determination] errs is in turning what it does not know—what [Althusser] calls ‘non-knowledge’—into ‘empirical’ objects that it can thereby only partially explain.”

Let’s compare this to how Althusser actually uses “non-knowledge” and the “empirical” in the quote Hamilton draws from:

...instead of comparing a knowledge with a non-knowledge, the non-knowledge is put into parenthesis and the empirical existence of the unknown object (called the ‘circumstances’ or contingent givens, etc.) is substituted for it – which allows the terms to be crossed, achieving a fallacious short-circuit in which the knowledge of a determinate object (economic necessity) is compared with the empirical existence of a different object (the ‘circumstances’, political or otherwise, amid which this ‘necessity’ is said to ‘assert itself’).

Hamilton asserts that the problem of mechanical economic determination is that it cannot fully explain the empirical, but that is not what Althusser is saying. The empirical is incompatible with structural analysis not because it cannot be fully explained, but because it is already being structured by the abstractions of the empirical, already offering its own explanation. When we gesture to “empirical phenomena,” what we need to do to incorporate it into a structural analysis is to treat it as “non-knowledge” that is as purely unstructured data which must be duly processed.

Hamilton almost seems to acknowledge this is what Altthusser is saying as he discusses the problem of the role of the individual in history,^[1] but reverses course, saying that here the error lies within Althusser himself:

From a philosophical standpoint, the obvious problem with this gesture is that it cannot explain the origin of abstractions. For if the empirical is always-already theoretical, then where do theories come from? While Villarreal does not acknowledge the seriousness of this issue—partly perhaps because doing so would undermine his narrative of me thoughtlessly aping the invective of authoritarian party apparatchiks—it is a real problem; one many commentators of Althusser’s corpus have tried to resolve.

So, where do abstractions come from? This was already answered when I mentioned the “basic structuralist insight that before we can establish the relationship between certain phenomena, we must break up the continuum of experience of the phenomena into signs.” Hamilton notes that Althusser rejected certain aspects of structuralism, including preferring the language of “tendencies” to that of the “sign,” but it's crucial to see just as well what Althusser didn’t reject from structuralism, and which animated the entirety of the disparate tradition even into the dregs of post-structuralism.

What Althusser rejected was the formalist, idealist structuralism of those like Levi Strauss, or Jordan Peterson today, both of whom similarly believe in the structuring role of a limited set of elements such as myths in society, but it is only the limited, enclosed nature of these elements that Althusser rejects.^[2] As Althusser notes, relational, combinatorial logic is also fundamental in Marx. What Marxist materialists must reject, however, is that these elements can be laid out totally in advance. This insight, i.e. that meaning is relational (determined by one to all connections), not referential (determined by direct connections to what is represented), was the founding structuralist insight of Saussure, and which is maintained by Althusser in all his works, and was shared by all his students and by associates such as Lacan.

Once this is clear, it is easy to see what an abstraction is, it's the combination of two or more things represented by a single thing (a sign or symbol of some sort), where the things are points within a wider field of positions and oppositions which were, in the last instance, created by the transformation of non-knowledge into knowledge. If we must go further and ask where individual elements come from, the answer is the same as before, they come from breaking up some larger continuum into pieces which relate to each other. This of course points to things which exist outside of knowledge and systems of signs, but that is no trouble for Althusser or any materialist, it is simply saying that real objects exist out there, outside of knowledge, but that can be transformed into objects of knowledge via our labor.^[3] The formalists are idealists precisely because they reject this operation, which involves the creation of new elements.

Hamilton suggests I haven't taken this issue of the origins of abstraction seriously, however, this is the central topic of my research on AI and semiotics.^[4] I did not think it was necessary to dig into how abstraction occurs on the most basic levels of computation for my review of Flowers for Marx, nor do I think it's necessary to do such elaboration in order to assert that it is a scientific fact discovered by structuralism that this breaking up of continuums is a necessary feature of creating abstract structures. While I don't think elaborating on the origins of abstraction would have “undermined” my “narrative” of Hamilton's relationship to the epistemic authority of the Party, his response has illuminated precisely the way he defers to the bourgeois social sciences, which is similar to the Jacobin-affiliated authors, though it was obscured due to the nature of his disagreements with them.

Referential theories of semantics, until very recently being experimentally refuted by LLM development,^[5] were the dominant paradigm of the bourgeois social sciences, which all empiricism (as Althusser defines it) is based upon.^[6] When E.P. Thompson raged against Althusser's critiques of empiricism, he did so on the grounds that class should be defined merely in terms of how it is contemporaneously apprehended by people in a given time (how it is experienced empirically). When Noam Chomsky bemoaned French theory, he did so on the grounds that its terms were not defined so referentially as is favored among analytic philosophers, and he continues to deny that LLMs can ever produce language due to his preconceived ideas of language being syntactic operations of symbols with particular referential meanings. But for some reason, which is not very intelligible at all, Hamilton believes that the problem of the origin of abstraction is one for Althusser, but not the empiricists!

Hamilton's case rests on the idea that Althusser recognized the err of his ways and found a solution to the problem of abstraction within empiricism in his later work. This simply is not the case. If we return to Philosophy for Non-Philosophers, written in 1980, we are helpfully given the answer to this conundrum in the chapter titled “Abstraction.” Here, Althusser lays out precisely what the origins of abstraction are: it begins with language, and indeed Saussure's discoveries on language. If at all possible, I would urge readers to review it themselves. It is perhaps the most beautiful and rigorous philosophical digression on the topic. If Hamilton is correct about the evolution of empiricism in Althusser's work, we'd find here something altogether different, a treatise on the key role of empirical data as the foundation of abstraction rather than the linguistic sign.

Contra Althusser, Hamilton claims Sohn-Rethel provides the true origin of abstraction, that being the exchange of commodities. But Sohn-Rethel cannot help us here, as it's crucial to his thought that commodity abstraction does not occur in thought; this is his primary criticism of Althusser! For him, the commodity abstraction is an abstraction which exists in a “spatio-temporal sense.” That the historic discovery of commodities had an important impact on thought is neither here nor there, one must account fully for the connection between thought and material culture, something that Sohn-Rethel retreats from by refusing to articulate what precisely an abstraction in thought is. Where, then, are we supposed to find the origins of abstraction in thought other than language and the sign more generally?

Other Issues

Hamilton's objections are somewhat murkier and more self contradictory in the rest of the essay. Regarding the issue of bourgeois hard sciences, Hamilton’s "proletarian sciences” stubbornly continue to refuse to acknowledge the role quantification (that old bourgeois way of knowing) will play in solving problems of climate and social reproduction, which certainly Sohn-Rethel would have had no problem admitting. His quibbles about Popper's refutability standard and the transformation problem strain credulity of interpretation.

On supply and demand in Marx versus neoclassical economics, it's clear that Hamilton doesn't understand the significance of whether long term equilibrium prices correspond to constant returns to scale or not. If constant returns to scale was accepted doctrine of long term equilibrium, then something approximating Ricardo and Marx would still be taught in schools rather than the current psychologizing, demand driven framework. The assertion that Marx anticipated neoclassical economics should be taken as an insult and is doubly absurd if one actually asserts a break between classical and neoclassical economics on the understanding of supply and demand, as nothing Marx said on the issue of supply and demand was outside the mainstream of the classical tradition.^[7]

When I set out to read and review Flowers for Marx, as with Hamilton's piece at the request of Cosmonaut Editor Donald Parkinson, I was certainly not interested in embarrassing Jacobin authors. I was quite sympathetic to them regarding the controversy that had ended the book's first publishing attempt. What I was looking for, first and foremost, was someone, anyone, to rigorously defend Marxism as a science, and I was left disappointed. And after reading Hamilton's response, I remain disappointed.

-Nicolas D Villarreal

1. “The problem of the ‘individual’ here is the example he gives: for the ‘individual’ is simply a received notion, that is then taken for granted. Yet what is important to note for our purposes is that Althusser is so adamant here to not validate the language of the empirical that he uses another term: ‘non-knowledge.’ The ‘empirical’ is by its very nature a distortion—once ‘non-knowledge’ becomes ‘empirical,’ it has already been misrepresented.” See: Conrad Hamilton, “The Foothills and the Summits: A Response to Nicolas D Villarreal on 'Flowers For Marx',” Cosmonaut Magazine, December 5, 2025, https://cosmonautmag.com/2025/12/the-foothills-and-the-summits/.

   ↩

2. “Now no-one can claim that we ever gave way to the crazy formalist idealism of the idea of producing the real by a combinatory of elements. Marx does speak of the ‘combination’ of elements in the structure of a mode of production. But this combination (Verbindung) is not a formal ‘combinatory’: we expressly pointed that out. Purposely. In fact this is where the most important demarcation line is drawn. For example, there is no question of deducing (therefore of predicting) the different ‘possible’ modes of production by the formal play of the different possible combinations of elements; and in particular, it is not possible to construct in this way, a priori… the communist mode of production! Marx constantly uses the concepts of position and function, and the concept of Träger (‘supports’), meaning a support of relations: but this is not in order to make concrete realities disappear, to reduce real men to pure functions of supports— it is in order to make mechanisms intelligible by grasping them through their concept, and beginning with these (since this is the only possible way) to make intelligible the concrete realities which can only be grasped by making a detour through abstraction. But just because Marx uses the concepts of structure, elements, point, function, Träger, relations, determination by relations, forms and transformed forms, displacement, etc., that does not make him a structuralist, since he is not a formalist. Here the second demarcation line is drawn.” See: Louis Althusser, Essays in Self-Criticism, Humanities Press, 1976, 129-130.

   ↩

3. V.I. Lenin, “Lenin: 1908/Mec: 1. The Criticism of Kantianism from the Left and from the Right,” Marxist Internet Archive, 2025, https://www.marxists.org/archive/lenin/works/1908/mec/four1.htm.

   ↩

4. Nicolas Villarreal, A Soul of a New Type: Writings on Artificial Intelligence and Materialist Semiotics, 2025, 176-194.

   ↩

5. Leif Weatherby, Language Machines (University of Minnesota Press, 2025).

   ↩

6. To be sure, there are relational theories of semantics that do not rely on French structuralism or Saussure.

   ↩

7. “Villarreal’s other arguments don’t rise above this initial mischaracterization. He tells us that classicals were ‘capable of imagining hypothetical increases in supply versus demand and their effects on price’—a point that’s obvious, but when pitched against Beggs’ analysis ignores the difference between complex graphing procedures and the sort of verbal description of this problem provided by John Stuart Mill in his Principles of Political Economy.” See: Hamilton, “The Foothills and the Summits.”

   Complex graphing procedures my ass! Marx himself was doing the same sort of verbal descriptions as Mill and his predecessors, so I don't know on what grounds this special anticipation of the neoclassical could be found. By any objective measure, the marshallian cross has only set back economic science rather than advanced it, as Marx articulates perfectly in his day that what matters is what determines the equilibrium! Supply and demand on the level of a structural determination of capitalist society is a non-answer.

   ↩