The ideological classes and the lumpenproletariat
Marx left us with a puzzle concerning what he thought about class. On the one hand he opens the Communist Manifesto with the line “The history of all hitherto existing societies is the history of class struggles” and goes on to say “Our epoch […] has simplified class antagonisms. Society as a whole is more and more splitting up into two great hostile camps, into two great classes directly facing each other — Bourgeoisie and Proletariat.” He goes on further to predict the extinction of all other classes, saying “The other classes decay and finally disappear in the face of Modern Industry; the proletariat is its special and essential product.”. But no sooner has he said that, then he adds a puzzling rider
The “dangerous class”, (lumpenproletariat) the social scum, that passively rotting mass thrown off by the lowest layers of the old society, may, here and there, be swept into the movement by a proletarian revolution; its conditions of life, however, prepare it far more for the part of a bribed tool of reactionary intrigue.
Either society is splitting into to two great camps, with all other classes disappearing, or it isn’t. Which is it?
If the manifesto was an early work, before Marx had formulated the critique he lays out in Capital, with its pivotal concepts of abstract labour and surplus value, the third volume of that work ends with an abortive chapter 52 on class that starts by asking “The first question to he answered is this: What constitutes a class? — and the reply to this follows naturally from the reply to another question, namely: What makes wage-labourers, capitalists and landlords constitute the three great social classes?”. He then follows on with “At first glance — the identity of revenues and sources of revenue.” But immediately shoots that down as a possibility and then we get Engels note “Here the manuscript breaks off” and that’s it. Talk about a cliff-hanger! In fact that ending is so unsatisfactory that in some quarters the myth has arisen that this was the last thing Marx was writing when he died. Not so, of course. The last sustained writing work Marx did on Capital was on the manuscripts for volume 2. Volume 3 was assembled by Engels after Marx’s death, based on older manuscripts and notes. The chapter on class simply breaks off where it does because Marx never finished the thought of what comes after the “first glance”.
Finally, in volume 1 of Capital Marx writes about unproductive labour as “all who are too old or too young for work, all unproductive women, young persons and children, the ‘ideological’ classes, such as government officials, priests, lawyers, soldiers, etc.”. What exactly are we to make of the “ideological classes”?
The puzzle that Marx leaves us with, is how to make up our minds between whether class, in capitalism, is essentially a binary, with other class fractions as residuals from earlier pre-capitalist modes of production — or… whether class is a plural concept, allowing for non-residual, persistent class formations, such as the ideological classes and the lumpenproletariat, that not only have not disappeared in the last century and a half, but are going to be a permanent feature of capitalist society as long as it continues?
I would argue that the way to resolve this question lies not in marxological hermeneutics, poring over the texts and manuscripts of Marx himself, but through working out what makes most sense to us from first principles, balancing theoretical and pragmatic concerns. In order to do this, I will address the concept of the lumpenproletariat first, as this also illuminates the question of the ideological classes.
The lumpenproletariat — class or reactionary prejudice?
The question of the status of the lumpenproletariat has been taken up in a recent (2017) book by J. Sakai, “The Dangerous Class and Revolutionary Theory: Thoughts on the Making of the Lumpen/proletariat”. The book is composed of two main essays, the titular one and on the reverse of the folio, a complementary historical text “Mao Z’s Revolutionary Laboratory and the Role of the Lumpen Proletariat”. Personally, I would advise reading the second text before the first, as it grounds the whole discussion.
In the historical Mao text Sakai notes an inner tension between the classical Marxist class dogmas embraced by the Chinese Communist Party (CCP) of the early days of the movement and the actual social composition of the Red Army. As much as official reports to the central committee tried to disguise the fact, Mao’s internal survey-based social science methods revealed that in composition the vast bulk of his forces were neither workers or peasants, according to the classical definition, but what he called the “yóumín”.
Unfortunately Sakai does not give us the original Chinese characters for this term, but tells us that it translates as “floating people”. The second part of the term is straightforward, 民 (min) is a common term for people and found in multiple compounds designating all types of people or social groups. The first part yóu has two possibilities — 游 (U-6E38) — yóu — (to swim — read in Japanese kanji as to float) and it’s homonym and virtual synonym 遊 (U-904A) — yóu — (travel, wander, roam). In both cases yóumín means roughly vagabond or vagrant. Presumably the underlying metaphor is that these “roving people” as we might call them in English (particularly Hiberno-English) flow over the land like water or rivers.
In the category of yóumin, Mao includes landless peasants, the unemployed, members of Chinese secret societies, soldiers and even lower bureaucrats. This is a familiar assemblage of different types of people to anyone familiar with the pre-Marxian traditions of Chinese Confucian ideas about class or occupational categories. In Confucian ideology, the “four categories of the people” represented the mutually-interdependent idealised occupational categories upon which a functioning society rested. In order of importance they were the shí (士 — scholar/gentry), the nóng (农/農 — peasants), the gōng (工 — artisans/craftsmen) and finally the shāng (商 — merchants).
This ideological categorising social roles into four, has its parallels in other ancient traditions, like Hinduism’s four varnas. It also looks familiar from the perspective of European feudal ideas of a social order based on those who pray, fight, grow and make. And in Japan, for example, this Fēngjiàn system, imported from Confucian scholarship, hardened into the feudal-like caste system of samurai, peasants, artisans and merchants. But in China, the dominant shí strata, made the transition away from being a purely martial warrior class, to a literate scholar-bureaucrat administrator class quite early in Chinese history. Which had certain effects, particularly on the social standing of soldiers.
The four occupations system, as an ideological ideal, clearly leaves outside all those that don’t fit neatly into its categories. Traditionally this outsider status was called jiànmín (賤民). Jiàn is commonly translated as “mean”, not in the sense of cruel, but in the sense of base, low-ranking, lowly, poor, outcaste, despicable, hateful. Pretty much all the derogatory social status associations you can think of. Which would explain why Marxists like Mao found it necessary to use a different term, less laden with class-hatred. Included in this outcast “fifth element” category were of course the landless and destitute, but also, significantly, soldiers and the military, at least below the rank of aristocratic generals. This partly due to the relative loss of social status of martial roles under the dominance of the scholar-bureaucrat ruling class, but also due to the habit of poorly paid soldiers to regularly “live off the land”, i.e. predate on the peasantry in a way little distinguishable from banditry.
The heritage of this long-standing Confucian ideological thinking about class and economic occupation, formed the cultural environment in which Chinese Marxism had to articulate its own version. The rebranding of the jiànmín as Mao’s yóumín may have been a valid attempt at de-stigmatisation, but the continuity of the thinking behind the category as those who have “fallen out” of the “proper” categories remains. Witness the creative accounting that Sakai demonstrates in Mao’s reports to the central committee, attempting to downplay how many of his forces were not worker-peasants in the conventional sense. It seems not unfair to characterise this specific combination of class thinking as a kind of Marxist-Confucianism.
Lennon Ying-Dah Wong CC BY-NC-ND 2.0
One of the effects of this combination has been the promotion of the concept of “the masses” relative to class, in Maoist discourse. A tendency that Ellen Meiksins Wood blamed for some of the dynamics of what she alleged was “The Retreat From Class” amongst the Western Marxist intelligentsia, due to their brief Maoist period in many European countries in the 1960s and 70s. Even if Wood’s thesis may or may not have merit in a European context, it seems unfair or just wrong in the American context, where a strong Maoist current in the left has sustained from the 1960s to the present day, without any substantial disappearance of class from the ideological discourse of those tendencies. Aside from the “masses” thing though, the most obvious effect of the submerged “jiànmín-thought” is a use of “declassé” distinctly at odds with the eurocentric tradition of that term.
To recap, the European meaning of the “declassé” is that of a loss of elite or privileged class status, from which one naturally falls into the category of those who have nothing to offer the state but their children — the proletariat. In 1993 the Welsh rock band Manic Street Preachers had a hit with “The Masses against the Classes”. The title is from a famous quote from William Ewert Gladstone’s speech to a mass electoral meeting in Liverpool on June 28 1886 in which he declared “All the world over, I will back the masses against the classes”. Far from being a proto-Maoist, what the classical liberal Gladstone meant was that he was taking the side of the common people (the masses) against the prejudices of the social elites (the “classes”). Gladstone was using the old sense of classes as restricted to the elite strata of “the quality”. That old usage still survives in our idioms today, when we use “classy” or “pure class” as meaning superior quality. At that stage in the 19th century, the socialist movement — of which Gladstone was very much an opponent — had begun promoting the dignity of labour by insisting that the workingman (sic!) was a not a member of an excluded degraded mass, but part of a working class, from whose labour all worthwhile things came. Nonetheless the older, elite usage of masses and classes as counterposed social orders was still current enough to make Gladstone’s speech unremarkable in this.
Switching track from Gladstone to Marx, its clear that in his use of the term he also understands “declassé” as one who has lost a class status above the proletariat and now been reduced into its ranks, or even that of the lumpenproletariat. But if the proletariat’s “radical chains” are defined by its lack of any property or resources, then it is a class that people fall into, but not out of. Because less than nothing, there’s nothing.
This fundamental difference between European and Chinese understandings of the relationship between the proletariat and the declassé has its roots, I propose, in the heritage of Confucian ideological thinking around class, occupational category, and the jiànmín. Marx is clearly on one side here and Mao on the other. Which is not to say that rectitude necessarily lies with the former rather than the latter, based on some appeal to socialist hero status, length of beard, or what have you. But it does mean that militants from the broadly eurocentric traditions (Trotskyists, Stalinists, anarchists and autonomists) and from Maoist heritage, do risk talking past each other on the question of how to understand the lumpenproletariat.
As Sakai unfolds in the titular essay, the main source in Marx’s writings for delving into the category of the lumpenproletariat, aside from its brief, puzzling flash of appearance in the Manifesto, is in “The 18th Brumaire of Louis Bonaparte”. This is a text well know to anti-fascist theory heads because it has been used as a source material in the past by various writers trying to outline a “Bonapartist” model of the nature and causes of fascism. In our context Sakai is of course more interested in what Marx has to say about the nature of the lumpenproletariat. We get a discussion of the role of the “Garde mobile” recruited from the ranks of the lumpen, according to Marx, who then played a key role in the violent repression of the proletarian uprising of June 1848. But Marx goes further and extends the category of lumpen from a (sub?) type of proletarian to talk of the future dictator Louis Bonaparte as “chief of the lumpenproletariat” and “a bohemian, a princely lumpen proletarian”. Sakai goes on to suggest that Marx went as far as implying that Louis Bonaparte’s dictatorship constitutes a kind of lumpen regime or state, and discusses what that might mean.
It’s worth quoting Marx’s somewhat infamous laundry list of the constituents of Bonaparte’s December 10th society, in full. Given its association with the lumpenproletariat in the text
On the pretext of founding a benevolent society, the lumpen proletariat of Paris had been organized into secret sections, each section led by Bonapartist agents, with a Bonapartist general at the head of the whole. Alongside decayed roués with dubious means of subsistence and of dubious origin, alongside ruined and adventurous offshoots of the bourgeoisie, were vagabonds, discharged soldiers, discharged jailbirds, escaped galley slaves, swindlers, mountebanks, lazzaroni, pickpockets, tricksters, gamblers, maquereaux (pimps), brothel keepers, porters, literati, organ grinders, ragpickers, knife grinders, tinkers, beggars — in short, the whole indefinite, disintegrated mass, thrown hither and thither, which the French call la bohème; from this kindred element Bonaparte formed the core of the Society of December 10. A “benevolent society” — insofar as, like Bonaparte, all its members felt the need of benefiting themselves at the expense of the laboring nation. This Bonaparte, who constitutes himself chief of the lumpenproletariat, who here alone rediscovers in mass form the interests which he personally pursues, who recognizes in this scum, offal, refuse of all classes the only class upon which he can base himself unconditionally, is the real Bonaparte, the Bonaparte sans phrase.
In that long and scathing, even bilious, list, it can seem a small thing to note that soldiers appear only in the form of discharged, i.e. unemployed ones. As we saw at the beginning, in Capital 1 Marx counts serving soldiers amongst the “ideological classes”. Unproductive labour, for sure (a category he takes mostly from Adam Smith, NB), but not connected to the lumpenproletariat by dint of being soldiers tout court. This is where Sakai follows Mao rather than Marx in relegating serving soldiers, the police, and other members of the repressive armed services of the state, to the lumpenproletariat, on the model of the yóumín/jiànmín.
And here is the crux of the matter. For Sakai’s redefinition of the lumpenproletariat to work, we have to valorise Marx’s category of lumpenproletarian, while simultaneously dismissing or supressing that of the ideological classes. It’s not clear why the second class category falls for Sakai, because it is never discussed as a possibility in its own terms. Sakai makes much of the Mobile Guard, recruited from the lumpenproletariat but still referred to as such. However the Mobile Guard is only mentioned a single time in the 18th Brumaire, as a member of a assemblage:
To this declaration of the Constituent National Assembly the Paris proletariat replied with the June insurrection, the most colossal event in the history of European civil wars. The bourgeois republic triumphed. On its side stood the aristocracy of finance, the industrial bourgeoisie, the middle class, the petty bourgeois, the army, the lumpen proletariat organized as the Mobile Guard, the intellectual lights, the clergy, and the rural population. On the side of the Paris proletariat stood none but itself.
A passing mention seems little to stand up a theory that the armed services of the state, police, army, feds, etc, are not part of the “ideological classes”, as Marx would have it in numerous places, but themselves as much lumpenproletariat as the criminals these internally repressive wings target.
A bigger question is raised by the tone of the passage quoted above. Marx is clearly having a bad tempered-rant in these passages about what a bunch of scum the lumpenproletariat are. Are we reading too much into these few brief passages, in the context of the mass, weight and careful working out of his main theoretical works? Are these anything more than the unfortunate reactionary outbursts of the embarrassing uncle at Christmas, spitting bitterness over the turkey while others change the subject and pass the gravy?
Again, recourse to practical experience of organising is the way out here. Sakai has not written this book out of crankiness, but from decades of practical organising experience. From an anti-fascist perspective, I can’t think of a single instance of urban organising in the last decades, where the problem of how to keep the lumpenproletarian criminal underworld (or at least the whiter sections of it) apart from the fascists, has not been a practical priority for antifa militants. It seems stupid to argue metaphysical niceities when we accept their existence as a real concrete problem in our practice. At least for the basic existence of the lumpenproletariat then, its a no-brainer from a pragmatic side. What to make of Sakai’s proposition that the lumpenproletariat is a “non-class”, however, is something we will need to look at presently. But first…
WTF are the “ideological classes”?
We know that Marx mentioned them a few times (not many, in fairness) but what on earth are the ideological classes? There are two problems here — the plurality of “classes” as opposed to class, singular — and the confusing nature of the name “ideological”.
Taking the last first, as stated in previous sections of this work, we take the results of Charles Mills’ work on the meaning of Marx’s use of the term “ideology” as a starting point. Particularly the difficult to swallow revelation that in fact what Marx meant by the term is unconnected to the conventional use of that term in common and Marxist discourse today. So virtually the entirety of existing Marxist discourse on ideology is based on a historic misunderstanding, due to the unavailability of certain crucial texts (principally The German Ideology) at the time when this classical Marxist discourse was first formed. Kautsky, Lenin, Luxemburg, Lukács, Gramsci, Althusser, E. P. Thompson, and the list goes on… all wrong. Even if the last two had access to the necessary texts, they were so caught up in the conventions of the previous discourse that they weren’t able to uncover its mistaken starting points. Mills’ work exposing the errors in those starting positions have unfortunately gone largely ignored and that’s were we are now.
What Marx meant by the term “ideological superstructure” was the extended state apparatus that sets up and maintains, the public legal, physical force, political and philosophical conditions that allow private property, private production and market forces to exist.
The apparently fashionable position amongst the Marxist intelligentsia of all stripes, that the base/superstructure distinction is old hat or discredited, is effectively the same absurdist utopia of the right-wing libertarians or ancaps that capitalism can survive without the state. The “ideological” superstructure is not some ideational realm of legitimating ideas, so much as a very large body of men and women, armed with a hella lot of guns and enough bureaucratic computing power to make the Large Hadron Collider look like Sunday League football. In sheer headcount alone, it far outnumbers the capitalist class. It is a multi-branched, large corporate body funded primarily through state revenues (tax) and for the most part does not generate capital, but consumes it (although this is somewhat of an oversimplification in macroeconomic terms).
For the most part, it is composed of waged labourers who fall into the category of what Adam Smith, and after him, Karl Marx, called “unproductive labour” — a controversial category, the problematic implications of which is a discussion for another time. Nonetheless, the superstructure has a huge wage bill, which if it trys to pay by simply printing money, experience has shown, it tends to collapse the economy through hyperinflation. Consequently, as it has to get its taxes from somewhere, it lives in symbiosis with the capitalist “base” private sector, a share of whose profits it relies on for revenue.
Having put some flesh on the bones of the superstructure itself, we need to see if we can find a more mundane term than “ideological” to characterise this body of workers, functionaries, managers, etc. Civil servants seems too specific to the civilian bureaucracy alone, “servicemen and women” is associated narrowly with military services, and the public sector is much too broad. So maybe something like “state-service class”, as uninspiring as that is, will have to do for now.
It’s worth examining why the public sector is too broad a category. In the first case, the public sector includes public enterprises like nationalised companies. In the era when the UK state owned the coal mining industry, a car manufacturer (British Leyland), British Steel and so on, that did not mean these enterprises — which manufactured and sold goods and services to the public like any private company — became part of the superstructure simply by dint of being publicly owned. The base/superstructure distinction has to do with relation to social production not legal ownership. Similarly health and education services can be part of the public sector or the private sector, or a mixture of both, depending on the individual state. That doesn’t make teachers and doctors and nurses part of the ideological classes if they work in the public sector side of their industries. These industries — health and education — are as much part of social production as agriculture, manufacturing or other service industries. As Marx said, in what little he said in that infamously truncated chapter on class, it would be a mistake to identify class with revenue (and by coincidence he referenced doctors specifically in his counter-example).
But then, the next question naturally becomes “isn’t everything part of social production, then?”. What’s the difference between doctors and nurses on the one hand, and cops and civil servants on the other? Its all very well defining the base/superstructure distinction on the basis of the alleged relative separation of the latter from the sphere of production, but that’s just deferring the question. How do we determine what entails “relative separation” from the sphere of production or what distinguishes the sphere of production itself, as contrasted to a putative “sphere of reproduction”, for example? Are cops just “workers in uniform”, as some Marxists would have it? Or are they separated from the working class they prey upon by more than mere contingency of occupational function?
Abstract labour as foundation for class
The way to approach this, imo, is through the prism of abstract labour. Michael Heinrich maintains that, after breaking from his early Feuerbachian phase with its central concept of alienation, Marx was for many years relying on the “division of labour” as the source of capitalism’s social divisions and contradictions. Specifically from the point at which mental labour is separated from, and put over, manual labour. This I would identify as Marx’s “Thompsonian” phase because this problematic is lifted directly from chapter 4 of the latter’s 1824 Inquiry. Heinrich asserts that the manuscripts of the Grundrisse were work of this phase, and that it was only some time after them that Marx discovers the key-concept of abstract labour that allows him to transcend division of labour — and the “trinity” formula of classical political economy (land, labour, capital) it implies — and make possible the critique begun in Capital.
I am going to propose that the way to move beyond the “first glance” of chapter 52 (note that it starts with the “old” trinitarian formula of landlords, capitalists and wage-labourers as the “three great classes”, before trailing off) is through reconceptualising class via abstract labour, rather than for e.g. the law of value. Even if at first sight these seem equally abstract or abstruse formulations.
What do we mean by abstract labour? There are two main schools of thought on this. The first is that the abstraction is a transhistorical “substantialist” one. A process of generalising from individual concrete labours by abstracting from their particularities to find the general “essence” shared by all concrete labours, namely its duration. In this “substantialist” conception, abstract labour is a scientific materialist concept that is not particular to any time or place in human history, but can equally be talked about in relation to the construction of the pyramids of ancient Egypt, as well as production in future socialist or communist societies.
The second school of thought, is that abstract labour is a historically-specific social form, related to the value-form, which presupposes certain very specific relations of production, namely commodified labour, purchasable by capital on a labour market via wage labour, to become productive variable capital in the form of labour-power. Which implies that abstract labour, as a concept, includes the alienation from the person of the labourer of his or her labour-power as an externally usable commodity. The form of wage labour thus separates the labour-power of the worker from the conditions and process of the reproduction of the worker, body and soul, and the renewal of the labour-power (of which the worker is merely the bearer, under these social relations). The wage is a way for the capitalist to externalise the reproduction of the labour-power upon which their enterprise rests, as “somebody else’s problem”. It is this separation and externalisation that divides social activity, under capitalism, into distinct spheres of production and reproduction — in a way that no previous mode of production has done, at least to the socially-transformative degree capital has. This separation permits for capital, as an emergent real abstraction, to axiomatise its own cycle of reproduction outside of, and indifferent to, the cycle of human (and environmental!) self-reproduction.
Obviously I am of the second school of thought here. For me abstract labour is labour-power that is abstracted from the conditions and processes of its own reproduction, aka the historically-specific form of capitalist wage labour.
The three forms of proletarian labour
So lets talk about the unholy trinity of proletarian labour forms: prison labour, lumpenproletarian hustle and grift, and wage labour. Although value-form analysis tends to begin and end with the last of these three, for the particular questions we are addressing here, it’s more useful to begin with prison labour. Generally there are some miniscule wages for prison labour, but it could hardly be called wage labour because of that. Mainly because here the surrounding conditions of abstract labour — i.e. that the employer hands over wages in exchange for not having to concern themselves with the feeding, housing and other reproductive needs of the worker — do not apply here. The prison houses, feeds and even occasionally provides some medical treatment, of sorts, to all its inmates. In short, prison labour — a form of labour “available” to the proletariat under capitalism — does not take the form of abstract labour.
Moving on from prison labour to the hustles, grifts and other forms of illegal or “greyzone” income of the lumpenproletariat. Under the category of the lumpenproletariat, Sakai includes legally-employed wage-earners like serving soldiers and cops. I’m going to restrict the category of “lumpenized” labour to that specifically earning a living outside of the legal labour market. When lumpenproletarians talk about the possibility of “going straight”, by that they generally mean joining (or submitting to, according to lumpen cultural values) the legal labour market and giving up “la vida loca” or “the game”. So, in this view, lumpen labour is predicated on trying to thread a path between prison labour (and ever-present threat for most lumpen) on the one hand, or legal wage-slavery on the other (assuming the latter is even available to them — which is not always the case).
If we accept that legal wage slavery falls under the terms of abstract labour, and prison labour does not, where does the liminal world of lumpen labour lie? This question is complicated by the existence of large scale organised crime enterprises, staffed with bosses, accountants and employees, in some cases. Nonetheless, I’m going to say that for the most part, the lumpenproletarian condition is defined by a rebellion against submitting to the abstraction of labour effected by selling your labour-power on the legal (or even “grey”) labour market. This is for many reasons, but the main one being the absence of any support from the superstructure for mediating economic relations between agents in the lumpen economic sphere in terms of rights of contract, employment or even freedom from violence or murder, in extreme cases. These unprotected relations of contingent risk are of an entirely different order than the working life of the wage slave.
The conditions of wage work in the legal or grey labour markets are the predominant mode of proletarian labour under capitalist conditions, and determine the norm that other marginal modes, such as prison or lumpen labour, define the boundaries to and exceptions from. Orpheus has returned from the underworld and once again takes his place in the capitalist sun of “straight” society.
Always assuming that work paying enough to stave off homelessness or hunger is readily available. A guarantee that capitalism does not make, of course. That said, in a number of core or metropolitan/imperialist countries in the 1930s, a welfare state was created to pay unemployed workers a minimal income to keep them outside of the illegal economy of the lumpen. To the extent the recipients of this social wage do so, then we can subsume them under the predominant mode of the legal labour market economy, without creating an unnecessary new fourth category. Similarly most of the cash-in-hand or “nixer” labour of the informal economy can be seen as the blurry boundary of the same — to the extent that it is primarily a matter of concern to inland revenue and the welfare office only, rather than the police and judiciary.
The wage-workers of the superstructure and abstract labour
We now move to the final question in the series. The employees of the various apparatuses of the superstructure are legal wage-workers. But are they still part of the general pool of abstract labour employed in social production by private capital — “workers in uniform”? Or are they set apart in some way, albeit a radically different manner than the apartness of the prisoner or lumpenproletarian?
My answer is yes. Those employed in the specific social function of being the abstractors of labour, cannot themselves be the abstracted. Otherwise abstract labour could breed abstract labour without the support of any external agency and we wouldn’t have to wait for communism for the “withering away” of the state, because it would have already happened under capitalism, so we would be already living in the libertarian ancap utopia (or dystopia, depending on your point of view).
The distinction between the sphere of production of the base and the superstructure is that the latter is engaged in a form of meta-production — the production and reproduction of labour-power, not as living labour, like educators, doctors and nurses do, but as abstract labour, as jailers, cops and bureaucrats do. Prisons exist for the reason of meta-producing the abstraction of labour, and the servants of the state that operate them and put people in them, are carrying out work of a socially-specific kind. A labour which sets them apart from the mass of workers who are constrained, not only by prisons, but the relations of production they underpin, based on being abstracted from their own human needs and desires. Or to put it another way — in the sphere of social production capitalists consume commodified labour power (abstract labour), whereas in the sphere of meta-production the ideological class produce commodified labour power. The state-service class construct their human objects one-sidedly as citizen-subjects. In turn the capitalists employ their labour-power, again one-sidedly, as workers. At no stage, between these two points of power, is the human object treated as a unitary whole, only as the specific facet relevant to that power.
In the specificity of their social function, their “superstructural” function, as it relates to the creation and reproduction of abstract labour, we could call this class the “separator class”. When we talk of the extended superstructure, we include also those marginal elements who provide a privatised service to the state. Bailiffs, debt-collectors, bail enforcement agents aka bounty hunters, etc. These agents may not derive their income directly from the state taxes payroll, but their function is licenced by special legislation granting them non-civilian powers and consists of separating ordinary workers from the possibility of living outside the yoke of wage slavery, state subjugation and prison.
Solidarity with workers in uniform?
Given that practice is the final test of theory, perhaps a supplementary, more pragmatic justification might be useful for people uncomfortable with putting too much faith in abstract formulations. Another way of looking at this is through the lens of solidarity. Everybody knows that different sections of the working class can be set against each other through the general competitive relations of our society. That class solidarity is an end to be achieved, not a given starting point. But the possibility of realising class solidarity has to be part of the definition of what constitutes a class, otherwise class politics, as such, is moot.
Could the police unions ever be part of a united working class solidarity against the state? Absolutely not. It’s common sense, at least to anyone not from the privileged or middle classes, that the police are scabs in uniform, not workers in uniform. As for soldiers, since the historic division of labour that took place with the formation of the police for internal violence, reserving the armed services for external violence, there have been cases of soldiers, especially conscripts, refusing to fire on “their own people”, when deployed against internal dissent. But there have never been any significant cases of solidarity between soldiers and the colonial peoples or foreign nationals of countries they were invading at the time.
On a less violent front, certain sections of the left that have strong membership in local government and central state civil services, uphold the idea that civil service unions can be part of a united working class solidarity front. But again the historical reality is that in confrontations between the populace and the government over class issues, the civil service unions have never defied their paymasters to the extent of refusing to process the new taxes, directives or what have you. Striking miners may have toppled governments, but never striking civil servants. (The limit case of a general strike of unpaid public sector workers against insolvent governments unable to fund the wage bill is more a case of state collapse than workers solidarity, so doesn’t count in this context).
During the late 80s anti-Poll Tax campaign in the UK, the official slogan of the non-payment wing was “The three Ds — Don’t register. Don’t collect. Don’t pay”. In reality the “Don’t collect” part was a sop to the Trotskyists, specifically the Militant, who still persisted with the delusion that somehow the local government unions could play a part in the class struggle against the tax, by refusing to do their jobs in collecting it. In reality the anarchists driving the non-payment wing knew damn well that this would never happen — and sure enough, had it been left to those unions we’d still be paying the Poll Tax today. Because as any trade unionist will tell you, there isn’t any labour law defence against being dismissed for gross misconduct — and for a civil servant to refuse to implement the government’s work is gross misconduct as surely as a soldier disobeying a direct order from his or her commanding officer. The employees of the state must implement the will of the state, class solidarity be damned, or look for another job, simple as.
It may look as if the argument for the distinction of the state servant class apart from the general working class that proceeds from abstract labour and that from the pragmatics of solidarity are unrelated to each other. I personally think they’re actually just two ways of looking at the same thing, but I’m not going to labour the point.
What about the plurality of the “ideological classes”?
The plurality of the state service class lies in two directions. One that it is functionally divided into different branches or apparatuses — the judiciary, the military, the bureaucracy, the police, the political representariat, and so on. The other that both sides of the wage relation, order-givers and order-takers, are necessarily included. This means that despite the discipline the state effects on all its employees — in limiting their capacity to be in solidarity with the general working class, at least in their workplace — the normal antagonisms between bosses and employees within each apparatus prevent them being a single united class.
In a different direction, the differences of the terms and conditions of that command relation between “civilian” apparatuses, like the civil service, and more militarised services, where disobedience can be, at the limit, be punished by summary execution, or imprisonment in agency-specific, non-civilian prisons (military prisons, intelligence agency “black sites”, etc), produces strong cultural and ideological differences between the different groups. At the elite, commanding level, each apparatus evolves its own self-specific forms of “protagonist” ideology. The Supreme Court thinks the evolution of the law determines the history and future of the state. The military believe that foreign policy, as sighted down the barrel of a gun, determines the rise and fall of states. And so on.
However, despite the boss/worker and inter-agency conflicts, the differentia specifica of the superstructure compared to the base it rules over, is that it is articulated under a unified hierarchy of command. Unlike the private enterprises of the capitalist sphere who are united only by attention to the bottom line, investment flows and losses or gains of market share. Regardless of their specific location, the employees of the superstructure know that they serve the will of a single machine and that the interests of that machine are to some degree their interests. Its reasoning, their reasoning. The view that the state is natural, eternal and the true agent of history, is the shared protagonistic ideological delusion of this body that is at the same time, both classes plural and class-in-itself singular.
Are the lumpenproletariat and ideological classes “non-classes”?
For Sakai the lumpen are a non-class because they are literally dis-integrated — as in not integrated into capital’s cycle of production, through the wage relation (or, by extension, in an associated social production role like the peasantry, in articulated modes of production). Which of course presents a problem when it comes to the police (or the Garde mobile, to go back to the 18th Brumaire) because they are integrated, albeit into the repressive arm of the superstructure. Sakai’s argument that the police and the military are separate from the working class by their separation from social production — following the Maoist-Confucian logic discussed above — has some elements of the arguments for the base/superstructure distinction, but he does not incorporate this distinction in his class theory, so by extension, potentially all the workers of the superstructure, not just the police and the military, fall into an undifferentiated lumpenproletarian mass — of huge proportion.
Sakai registers the incoherence of motives and actions in a category that puts cops alongside sex workers, hustlers and pimps alongside soldiers, beggars and homeless alongside firefighters, by making that the reason for the non-class status of the lumpen. While it might be appealing, from the perspective of a radical rejection of bourgeois morality, to say that being on different sides of the law, does not necessarily divide cops from street gang members. Indeed, it’s often wryly observed by the streetwise that the cops are the biggest gang in town… and there’s truth in that. But even so, cops are legally-employed wage-workers, dependent on bosses and paymasters, with pensions and mortgages. Gang members live outside wage-slavery, violently autonomous, dependent only on each other, and are in the majority dead or in prison before they’re 30. These two things are not the same. Being down on the cops is one thing. But potentially dissolving the whole superstructure into the lumpenproletariat is a move that would create a “non-class” category so large, so incoherent, so contigently unpredictable in its motives and actions, that it threatens the very viability of class analysis as a materialist frame of analysis at all.
We need to separate the questions — is the lumpenproletariat (more narrowly defined, as I have sketched above) a “non-class”? Is the state-servant class a “non-class” as well? Here we have a bit of an Anna Karenina problem. That novel opens with the line “Every happy family is alike, but every unhappy family is unhappy in its own way”. For the category of “non-class” to make sense in a class analysis, we have to decide whether they are like either happy or unhappy families in Tolstoy’s witticism. If all “non-classes” are alike, then whatever groups of people fall into that category present as one undifferentiated mass, from an analytical point of view — one I have just rejected. If, alternatively, each different group of people categorised as “non-class” are “unhappy in their own way”, then that means analytically we can say things about their different characteristics, dynamics, and behaviours, determined by their material situation. Which is to say they have all the characteristics of a class, which is a contradiction with our starting premise. My conclusion is that the idea of a non-class that is circularly defined by its analytical incoherence leads inescapably to contradiction and breakdown by the principle of explosion. Which may be a great title for an insurrectionist journal, but from a class theory point of view, is not viable. Having an analytical system of categories where one of the categories is basically “Here be dragons” is not a runner.
As I indicated in the section on the plurality of the ideological classes above, there are circumstances where all the different groups of the superstructure will come together and act as a united class. Generally when they are under threat. When the shit is hitting the fan and the population is uprisen, the cops will be shielding city hall, and the judiciary will have the cops backs. When the whiff of tear gas and burning barricades is in the air, the various arms of the superstructure will pull together in a united front in the face of existential threat and social disorder. Despite their pluralism, internal and inter-agency tensions and rivalries, it would be a strategic error to treat the state-service class a non-class.
What about the lumpenproletariat? The material condition of trying to survive outside the legal labour market means that the competitive pressures of society in general, are multiplied and intensified. On the one hand, the lumpenproletariat doesn’t even have a job to lose, so having nothing to lose, there’s nothing to prevent them taking the side of the proletariat in struggles against employers and/or state denying their reproductive needs. But how long they can sustain that before having to go back to whatever it was they were doing to survive before, is an open question. Similarly, there have been episodes in the past, in the wake of the Rodney King beating verdict, for example, where rival street gang federations like the Bloods and the Cripps, have declared truces in support of the demands of their working class communities against the racist violence of the state. But the brutal, almost social-Darwinist competitive pressures that relentlessly drive inter-gang conflicts can’t be held off forever, and such truces can never last long.
So yes, as a class formation, the lumpenproletariat is almost uniquely fractious. And while the maxim “there’s no honour amongst thieves” may be the sneer of the socially-privileged, there’s no doubt that chronic division and internecine conflict present severe challenges to the possibilities for any sustained class unity, either on its own account, or with the wider proletariat. The problem of the “by whatever means necessary” survival ethos of the lumpenproletariat means it will always provide some willing recruits to sources of ready money looking to hire muscle for “dirty deeds, done dirt cheap”. This makes them a potential problem whether you’re social movements in Central and South America having to deal with the army on one front and the Narcos on the other, or tenants unions in the metropole having to deal with developers and slumlords hiring gangsters for evictions. But by the same token, the reality that most non-proletarian classes use lumpenproletarian services of all kinds (not just muscle) means they have access to information and intelligence regarding hidden agendas and strategies of the class enemy that are a potential goldmine for organising movements. It’s not just the cops that rely on lumpen informants for their intel.
But before we leave the “non-class” question behind, we have to deal with a related question of whether these are “ambiguous” classes — i.e. does the lumpenproletarian or the cop not remain a proletarian, for all that, thereby violating the idea of a class membership as being unique and exclusive? Does it not break the idea or very point of class to imagine that individuals might be more than one class at once?
The answer to this lies through consideration of related questions raised by the feminist movement about class, long since. For e.g. — If a middle class woman marries a working class man and becomes a housewife, does she become working class? Similarly — Are married women’s class status dictated by their husband’s status?
The answer from sociologists would be that we need to distinguish between an individual’s class “background” and their current class “position”. In the natural language of ordinary conversation we often use these originary qualifiers like “background”, “from an X family”, “grew up in a Y neighbourhood” and so on, when talking about individuals and class.
Lets call these two factors “native” and “occupational”. Where the two factors coincide, no question arises. Where they diverge is where it gets interesting, because the question is which predominates over the other? The “nativist” view, popular amongst “self-made men” is that their humble origins story defines them much more than their current billionaire boss status. Similarly the notion of police as workers in uniform leans heavily on the working class origins of most cops.
The alternative view, that occupational class swiftly overcomes native class origins, is the one, I submit, more coherent with historical materialist views of class in general. The idea of the “declassé” for example, is one where the bourgeois or petty aristo origins of a “ruined” person, does not prevent them being as resourceless as the rest of the proletariat, once deprived of their money and social networks. Acquaintances whose schoolmates became cops recount how quickly their former friends lose social contact with the outside world and restrict themselves to socialising and associating exclusively with other cops. Women who marry soldiers rapidly become army wives, due in part to having to relocate periodically to follow military postings, children attending army-run schools, etc. Regardless of their own native class origins, women who gave up their own employment when marrying, back in the days of the “housewife” as a common social institution (albeit a historically and geographically limited one), by default did assume the occupational class position of their husband.
In summary on the question of the “ambiguity” of lumpenproletarian and state-service classes, this is mostly resolved by asserting the primacy of occupational over native class. In other words, that on joining either, people are confronted with the choice to either “shape up, or ship out” — i.e. adopt and internalise the institutionalised ethos and culture of their new association, or be spat out again. In the case of the lumpenproletariat this is enforced more by the intensely and brutally competitive relations of the underworld economy, rather than any organised or centralised disciplinary apparatus. In the realm of the state-service classes, there are obviously substantial differences in degree to which this social separation from the rest of the working class is effected, depending on whether the individual branch joined is part of the repressive apparatuses of the superstructure (cops, army, feds, screws, bailiffs, debt-collectors) or its more civilian arms. Its clear that minor-ranking civil service bureaucrats are safe drinking and socialising in working class pubs and clubs in a way that cops are not. The ability of known lumpenproletarians to make use of the same pubs and clubs depends to a certain extent of what branch of “the industry” they’re in, and the character of the establishment, as determined by the owner or management.
Finally — is class a binary or not?
We started this piece with a brief review of Marx’s scanty writings on class that left us with a puzzle — is class a binary of capitalist/proletarian, or is it a plurality? It’s time to answer that question.
I’m now going to commit that most irritating and heinous of dialectical dodges by playing the “it’s not either-or, but both-and” card when it comes to the binary vs plurality question. Putting abstract labour at the centre of the class question allows us to both have our cake and eat it. Yes, the binary opposition between capitalists and proletarians, mediated by the indissoluble trinity of abstract labour, the law of value and the class struggle, remains the dominant paradigm of all capitalist social relations. But there remains the necessary load-bearing structural support of a small plurality of additional classes — superstructural, lumpen, landlord, etc — necessary to reproduce the social relations required for this centrally defining antagonism to play out. The flying buttresses of the capitalist class cathedral, if you like.
How, or in what way, can class still be a binary if there’s more than two classes? What happens to “the” class struggle? Does it become class struggles, plural? The answer is, it’s partly a matter of perspective.
From our perspective, the proletarian perspective, our struggle for survival, freedom from want, domination, exploitation, environmental destruction, is all one struggle. Our opponents in that struggle may wear many hats, that of the boss, the landlord, the cop, the welfare clerk, the banker, and so on. But our struggle is against our separation from the conditions of our own reproduction, our material and social needs — one single struggle, even if fought on multiple fronts.
From the perspective of other classes, however, things get more complicated. Historically, in the UK the capitalist class had to make an alliance with the working class against the then dominant landlord class in the struggle against the Corn Laws. Then the landlord class later attempted to make its own alliance with the working class to oust the capitalists. In the end the capitalists won out (ish) but reports of the death of the landlord class, even in the metropolitan heartlands of the old imperialist powers, have been greatly exaggerated. In the UK the monarchy and the aristocracy are still the largest non-state landowners, leaving aside certain big tax-exempt charities (National Trust, RSPB) also controlled by the hereditary landlord and aristo class. Both capitalist and landlord classes have to fend off not only each other and the working class, but also have to struggle for control of the extensive resources of the state itself. At the other end of the class pyramid, the lumpenproletariat may not have to worry so much about the capitalist class in the shape of a legal employer, but if their “natural enemy” the police can often be placated with informing and bribes, they have to keep a weather eye over their shoulder for the occasional backlash from working class community organisations (like Ireland’s Concerned Parents Against Drugs in the 80s and 90s). In countries where the peasantry remains a significant proportion of the working masses, the permutations for the non-proletarian classes can increase again, without mentioning the role of indigenous groups also, in certain parts. About the only constant in all of these varying scenarios and situations, is that for all these classes and fractions, the role of the proletariat remains the potential decider in any contest, the sleeping giant that can upset the applecart if aroused. Of course the strategic question for all sides then becomes how the proletariat can be kept predominantly passive or divided against itself. But that is a much longer conversation, that is out of scope for our exploration of the binary versus plurality question of “how many classes?”
Conclusion
In all the community organising and anti-fascist work in the living memory of militants still with the movement today, the lumpenproletariat has been as much a practical problem as the police and other repressive wings of the state. I agree with Sakai that our class analysis exists to aid and support our work as movements for social change. And so the category of the lumpenproletariat should be taken as seriously in our theory as it is in our practice. Yes, the antagonism between the needs of human self-reproduction and that of capital’s relentless accumulation provides the background of a two-way fight for all social struggles. And that abstract dichotomy is made flesh in the division of peoples into two “great camps”, proletarians and capitalists. But that does not mean these are the only social classes in society. Nor that those other classes, from landlords to peasants, from the state-service class to the lumpenproletariat, to segments like the intelligentsia or the middle classes “inside and against” the proletariat, are mere historical residues, disappearing strata destined to recede from view. To the extent that they play currently irreplaceable roles in reproducing the system as a whole, then they are all also part and parcel of it, and need to be analysed in their specificity.
While I argue against Sakai’s inclusion of cops and soldiers into the lumpenproletariat, for the reasons outlined above, I want to propose the return of the “ideological classes” by way of a “friendly amendment” rather than a motion to strike out and replace in full, as it were. It’s no surprise to anyone that anarchists the world over have always rejected the idea of our natural enemy, the police, as “workers in uniform”. Some Marxists have taken this as a sign of anarchist’s lumpen composition and “a-theoreticism”. In turn anarchists have responded with accusations of bourgeois intellectualism, and so forth. Each side trading ritualised slurs and insults that reflect our cultural and sociological differences without really advancing the project of reformulating a class theory — a “second glance” — adequate to our 21st century struggles. And I feel that there is still far more that still needs to be said on these topics than has been touched on above. But this piece is already far too long and its pointless aspiring to start a debate with an opener too lengthy for any reasonable person to read. So I’ll end it here, for now. For those of you who made it this far, I hope the effort was worthwhile or at least left you with some things to think about.