Adapted from a talk given at the launch of Abolishing Capitalist Totality: What Is to Be Done under Real Subsumption? (Minor Compositions, 2026) at Historical Materialism, London, November 2024.

I completed the essay for this much-delayed book some years ago, and haven’t really returned to the topic since, so it seemed necessary to make this a sort of retrospective presentation of a past project, rather than of something active and ongoing—although, of course, there are going to be some continuities in the thinking of any single individual.

The original project dates all the way back to 2007 or 2008, when the Endnotes group was a lively, fractious discussion group developing its own approach to Marxism, particularly through a thinking-through of a critique that Theorie Communiste had made of the journal Aufheben, which most of the original Endnotes group had been involved in. In June 2008, before even the first volume of Endnotes had been published, me and Nick Gray, who was also involved in the group, presented the first draft of the project in an event organized by Tom Bunyard at Goldsmith’s; it was published in a book called The Devil’s Party the following year. That initial text sketched out some criticisms of TC’s conceptual framework, while remaining friendly to the outlines of their thinking.

A central feature of Theorie Communiste’s approach was an attempt to historicise and periodise class struggle in a way that called into question the abstract normativity and moralism of typical ultra-left and workerist narratives about the lost struggles of the past and the contemporary situation. Since many of the original Endnotes members felt that TC had a point, we set about working through the implications.

Key to that periodization was the concept of subsumption. This had been rediscovered in the drafts of Capital from the 1970s, and taken up by many people, including Toni Negri, Jacques Camatte and TC, as a periodizing term, with the distinction between formal and real subsumption registering some fundamental shift in capitalist society. Often, this was linked to some idea of a kind of total subsumption, in which the entire social world was viewed as saturated by capital. TC had added a touch of regulationism and Althusserianism to the mix, and attempted to further break down their period of ‘real subsumption’ into distinct phases.

What, I think, was compelling to the thinkers who picked up this term from Marx’s drafts, was that it seemed to provide a conceptual bridge between the abstract theoretical schema worked out in Capital, with its emphasis on capital as a sort of immanently self-producing totality, and the actual unfolding of capitalism over time. It was certainly compelling for us too, in a certain Hegelian-Marxist fashion, as a way of thinking through capital’s self-actualization in concrete history.

However, when I set about analysing this concept, I quickly ran into doubts about the viability of its use for these periodizing projects. It took me some time to fully articulate what was wrong with it, so there were several textual iterations over about a decade, gradually firming up a critique, before starting to indicate a different path onwards.

In the early versions of the project—in The Devil’s Party and Endnotes 2 (2010)—the main thing we noted was that there was something wrong with the logic by which the concept of subsumption was taken as a bridge from the immediate production process to capitalist society at large, and its history.

In trying to get to the roots of this conceptual problem, I searched through the oeuvres of the major German Idealists in the hope of identifying Marx’s source, and found the term used by Kant, Schelling, and occasionally Hegel. It was used in a fairly technical logical sense, to refer to the ranging of a mass of particulars under some universal. Kant considered the relation of the manifold to the categories of the understanding to be one of subsumption.

Hegel developed a critique of the apparent externality in this subsumption of the manifold: there was something absurd here if the point of these operations was to know something about particulars, for the particularity of particulars seemed to be obliterated by the subsuming concept. This was also violent: a domination of particulars by universals.

Nonetheless, Hegel used the term occasionally, notably in characterizing the relation of the sovereign’s decision to the particulars of civil society. It may be tempting to infer from the use of the term in this context a subtle critique from the left: viewing the sovereign’s decision as a violent and absurd imposition. But it is far from clear that this was what Hegel meant—this is the late Hegel, viewed by many as a thinker of the Prussian Restoration; or, at most, defended by liberals as a thinker of freedom in the context of the modern constitutional state.

And indeed, the use of the concept of “subsumption” was something that the young Marx criticized in Hegel, as the false imputation of a philosophical category to objective social processes. There was an irony in this, because Marx’s own later use would impute to capitalist society itself what he saw as absurd here about Hegel. Against his younger self, the mature Marx concluded—with Hegel—that subsumption really could be a social process, not just a matter of logic or epistemology. It might still be absurd, but it was also real.

So why did Marx turn to this odd philosophical category when he was drafting Capital? The answer seems to be that he saw a striking resemblance between these logical and ontological operations whereby a universal presents itself as the truth of some mass of particulars, and the operations of value and capital. In commodity exchange, the value form presents itself as the truth of particular use values; in capitalist production, capital presents itself as the truth of the production process. There is a violence in these relations, and if we follow Marx’s playful Hegelianism, there is also an absurdity.

Taking seriously the specific philosophical content of this term meant, for me, focusing on that moment of identification between universal and particular. Subsumption, in Marx’s use, was all about the relation of capital to the labour process. It was not about the relation of capital to capitalist society at large. Of course, the kinds of transformations in labour processes that were brought about through real subsumption under capital did have massive effects beyond the sphere of production, but that doesn’t justify any identification of society at large with capital.

What’s more, because the production process that is subsumed by capital is always a particular one, one cannot simply use these terms to periodize the history of capitalist society as a whole. While it is true that the distinction between formal and real subsumption implies a movement from the first to the second, formal subsumption in one production process may follow real subsumption in another. The temporality is an abstract one that does not map simply onto phases of society as a whole.

At most we can say that, at any given moment, capitalist production processes are “more” really subsumed than previously. This is somewhat akin to how we think about demographic ageing: many societies now are demographically older than they once were, but it would be fallacious to try to derive from this some socio-historical periodization into periods of “youth”, “maturity”, “senescence” etc. Thus, subsumption did not make sense either in the characterization of capitalist society at large, or in the periodization of its history.

This was underlined by the distorted interpretation of history that use of these concepts seemed to involve. If, as it was claimed by various thinkers, some phase of “formal subsumption” had lasted up until around the First World War, this implied that absolute surplus value reigned up to this point; and for that to be true, the consumption goods purchased by workers’ wages could not have been affected by productivity increases or changes in production methods.

While it makes logical sense that some kind of extensive moment would have to precede any intensive one, the idea that we could isolate some period of purely extensive accumulation, as the regulation school also did, does not stand up to historical scrutiny. For, of course, the production of consumption goods was capitalized early, not late: capitalism proper began in agriculture—in the 15^th^ and 16^th^ centuries—and with it “real subsumption”. In this, we drew on Brenner and Glick’s critique of regulationism.

We formed these rather technical objections early on, and these core arguments persisted until the final essay which is appearing in this volume. But over time, I became curious about the role of grand totalizations like these in Marxist theory. If subsumption didn’t get us there, was there nonetheless some important truth being intuited in these notions of a completely capital-saturated whole actualizing itself through history? This pattern of thought is a common one in Marxism, not only among those thinkers who became hooked on subsumption, but under different terms in others such as Tronti and Jameson. Arguably it goes back at least to Lukacs.

But, when I went looking in Marx I could find no equivalent: no great emphasis on this sort of notion of “totality”, which played such a role in Western Marxism—only passing uses of the word, to mean quite different things. Capital itself clearly formed a sort of self-enclosing whole for Marx, but I found no strong evidence that he identified this enclosure with society at large, the state etc. This came later.

One of my frustrations with the problematic that was established by this kind of totalization was the way in which it could slide into a rather metaphysical mode. The endless abstract riddling, in the circles close to TC, around the question of how the proletariat could abolish capitalism by abolishing itself without somehow affirming itself in the very same action and thereby reinstating itself. The puzzling over how we, being “inside” capitalism, might get “outside” it, when it encompassed the entire social world. All of this could seem miles from actual social struggles, although TC made valiant efforts to rediscover their theory immanently in the analysis of particular movements.

I responded to this with something called “Null Outsides”, and then “False Totalities Don’t have Exits”, to be presented at the little gathering in Bilbao in 2014 which formed the initial basis for this book. The initial move was to draw a distinction between totalities of the sort that can be elaborated in detail—sets of self-related mediations constituting a sort of whole, of which I took the paradigm case to be Marx’s analysis of capital—and totalities that involve a sort of vague holism.

This latter sort I termed “false” totalities, because they lacked any real theoretical definition. Little wonder—went the thought—we struggled with thinking our way out of them, when we evoked totalities so lacking in definition that they could seem to encompass the universe at large. The answer, then, was to insist on a more determinate kind of totality. This might get us out of our theoretical binds.

In 2015, Arts Against Cuts put out a fun little book called Bad Feelings, to which I contributed a very early version of a text called ‘Error’, containing a version of these arguments; this eventually came out in a fuller form in Endnotes 5.

But on further reflection, and in the process of working through the drafts of ‘Error’, I started to question my own notion of falsity here. Were such totalizations false or were they just indeterminate? Were indeterminate totalities necessarily false ones? Wasn’t the correct distinction between determinate and indeterminate totalities? Perhaps totalities could be indeterminate but still carry some truth? What might that truth be?

What distinguishes the text that appears in this book is primarily that it captures these lines of questioning. I assembled much of the text during long, sleepless nights spent watching over babies—the title is ‘Feeding the Infant’—so I was outside of the old patterns of debate around Endnotes and all that, and I was in a playful and provocative mood. Some of the text may be a bit frivolous as a result, but I’ll let it stand as it was, as an expression of a moment. What seems to me its probable main weakness is that the logic of my thinking took me onto terrain that I suspect has a whole scholarly sociological literature. I am no academic, and definitely not a sociologist.

The text does contain a fuller, more thoroughly fleshed-out version of the critique of the concept of subsumption that I have already surveyed, but it also posed questions about these kinds of totalization—indeterminate and determinate.

But I shifted the frame somewhat, from the Marxological focus of the beginning of this project, to a more general social-theoretical one, with the driving question: Why do we totalize indeterminately?

I was struck by this line from Durkheim’s Elementary forms of Religious Life: ‘the concept of totality is but the concept of society in abstract form. It is the whole that contains all things, the supreme class that contains all classes’.

But it seemed to me that Durkheim had things upside down. Isn’t it society itself that is the abstraction here? The concept of society encompasses so much: institutions, classes, laws, people, cultures, physical infrastructure and so on. It is surely impossible to articulate the concept of society as anything more than a hazy encompassing of these sorts of indeterminate sets. That is to say, it would seem impossible to articulate a theory of ‘society’ as such, as anything other than an indeterminate totality.

I was also struck by some arguments in Robin Dunbar’s famous work on grooming and gossip in the social bonds of primates, and this is where I was most in the playful and provocative mood. Dunbar’s work suggests that abstractions become completely unavoidable once we are dealing with groups above a certain, rather small, size. Certain primates may be able to keep track of all the social relations that constitute their groups, through constant grooming rituals, which is to say: the social whole may be determinate in small primate groups.

But with larger groups, this sense of the immediate self-presence of the social whole becomes impossible. Once group size surpasses this level, new phenomena come into play, such as increasingly hierarchical structures, and a displacement of direct personal loyalties by ‘more nebulous and less inspiring concepts such as “the regiment” or “the Queen”’. This emergence of nebulous concepts of the social would then be the emergence of indeterminate notions of totality.

But a key thing here is that these nebulous concepts themselves play a social role. In a sense, the indeterminate totality is constitutive of the social. Vague notions like ‘society’ and ‘the nation’ have a certain performative role in enabling the functioning of such large-scale groupings. The indeterminate totalisation of the social whole would then be not simply a failure of thought, but an abstractive function that enables that whole to exist in the first place. Social groups unify themselves under the sign of one or another concept that can stand in for the whole, yet without specifying it concretely.

Ideas of a completely integral, self-present social whole, such as we find in Hegel’s and the young Lukacs’s notions of ancient Greece would then be closer to a pre-human kind of social grouping, and it would be the human revolution itself that introduced a certain kind of social abstraction.

All of this has a significance for attempts at constructing a revolutionary theory. If our notion of revolution is pinned to one of these indeterminate notions of the social whole, we will be faced with a problem of scope that will nullify any attempt to do things like thinking strategically. If everything is to be overcome, then where must we start?

By contrast with these vague notions of the social whole, the kind of totality we find in Marx is specific and concrete: it is the specific set of mediations that constitute capitalist relations of production. It thus seems a more promising basis for revolutionary thinking. We can insist on a more limited scope specified by Marx’s analysis of the mode of production.

But there is a strong tendency to slide from one mode into the other, and to start imagining that Marxism gives us a theory not just of the mode of production in a strictly defined sense, but of the entirety of ‘society’—whatever that is.

The concept of subsumption has often played this role of enabling the sort of telescoping movement whereby the indeterminate whole of society is collapsed into the specific totality of the mode of production.

But, as enormously important as it is for what happens in any given society, the mode of production does not encompass everything. People clock off work at the end of the day, and what they do outside of the workplace is not entirely reducible to the reproduction of labour.

Most notably, Marx never managed to fully articulate any theory of the state. Other aspects of the social whole that have been of enormous importance in the actual history of capitalism, such as the nation, were almost entirely untheorized.

There may be a tendency to over-extend one’s notion of the mode of production to implicitly encompass these under-theorized terrains, which is particularly facilitated by concepts like that of subsumption if we fail to also address questions of the state, the nation and so on. If we are talking about some kind of social totality that fully encompasses us, and makes an ontological claim to represent our truth as particular individuals, the state and the nation seem better candidates than the mode of production.

This distended notion of subsumption would then be a sort of theoretical cipher for something that is repressed in theories which fail adequately to think through other notions of the social whole, such as nation and state. Perhaps, then, we end up back where Marx started: with Hegel’s sovereign subsuming civil society under the decision.