Reading Guide to: Poulantzas, N (1975) Classes in Contemporary Capitalism, London: New Left Books.


Social classes in Marxist theory are seen as groups of agents defined by their place in the production process, but not determined by it. Politics and ideology are also very important. Classes are defined by struggle, which coincides with class practices, but they occupy objective places in the division of labour independent of their will as agents. The social division of labour includes the political and ideological levels: class as a social matter denotes the effects of this overall structure. However, class positions are different in each conjuncture -- but the issue of class determination is not reducible to locations in conjunctures. It is possible to have a 'distance' between determinations and eventual position.

Therefore, a class, fraction or stratum can have a class position which does not correspond to its interests. For example the aristocracy of labour takes up a bourgeois position, even though it is not a bourgeois fraction, but a stratum of the working class. Some fractions of the petit bourgeoisie can do the opposite. Technicians, for example, can take either position, even though they are not part of either class.

Ideological and political relations are themselves part of the structural determination of class -- it is not that the economic factors determine, and the ideological/political factors merely position. So if the working class exists in the economic level, there must be a space for it in the ideological/political level even if this is not actually developed. There is a 'class instinct' in these levels, a potential to burst bourgeois discourse. The ideological level is not just a series of ideas, but material practices: there is no need to develop a specific consciousness at the level of ideas.

Class consciousness refers to the terrain of class positions and conjunctures, and represents the conditions for intervention by classes as social forces [this is the sort of economic analysis that Parkin needs?]. The point of analysis is to locate places, rather than explore the class origins, feelings and so on of participants. Consciousness is important for the reproduction of classes and agents: it is irrelevant to emphasise belonging to a class, or the statistical or empirical boundaries between social groups -- what matters is how classes act in class struggle. Class is not a matter of social inequalities either, since these are only effects on the agents of objective places.

A principal role is played in determination by the economy, meaning the process of production. This places agents and organises their distribution according to the relations of production. Production here means the whole reproductive cycle. So, for example, levels of income vary according to productive relations, and this joins in consumption as part of the reproductive cycle. Production is a form of labour process, expressing human relations to nature, but it takes place in historically determined social formations. The relations of production include relations between agents and the objects and means of labour, and the relations between men and their fellow men [sic] -- in other words, class relations.

In capitalism, the relations of production include economic ownership, which provides real economic control over the means of production, and the capacity, deriving from possession, to put the means of production into operation. In every class society, ownership confers the right to exploit labour. Economic ownership is a basic determinant of class relations. Usually, but not always, economic ownership becomes legal ownership, that is, law develops as an important ideology. The relations between labour and the means of production also vary. In capitalism, there is nothing but abstract labour, labour as a commodity, which leads to a generalisation of the commodity form (the determinations work this way, and not as in Lukacs, where the commodity form has an independent effect).

So the relations of production reveal an articulation of these various relationships between human beings, and nature, between economic and legal forms. In any mode of production, one relation of exploitation becomes dominant, and this determines the nature of the exploited class. It is not a matter of ownership as such, since not all non-owners perform productive labour -- forms of ownership are seen as an index of underlying relations of production rather than a defining characteristic as such. Similarly, the production process is not defined by technology but by the articulation between the productive process and the relations of production: productive labour is not just a technical or neutral term, but something 'which gives rise to specific and dominant forms of exploitation'. Production always means class division in class societies.

It is not the fact that they earn wages that defines the working class. The wage system is also an index for relations of production, corresponding to a distribution system and the organisation of labour power. All members of the working class are wage-earners, but the reverse does not apply. The hierarchy of incomes is an effect of class. These divisions of income are important for concentrating inequalities in particular groups, such as the young, the old, and women [this is an example of a specific articulation between class and sexual divisions in the overall division of labour ]. [Poulantzas also offers the basis of an explanation here for the so-called new social divisions which have allegedly replaced social class]. The relations of production are responsible for this process of production and reproduction of social divisions.

The domination of relations of production over forces of production and the labour process itself produces an important (constitutive) role for political and ideological relations. Economic relations provide the basis for class powers, but these are tied to ideological and political expressions and legitimations. Such legitimations are always present in the constitution of relations of production. Thus economic production and reproduction must involve the reproduction of political and ideological relations of domination and subordination. The articulation of all these aspects produce the social division of labour, which determines the technical division of labour. Apart from anything else, this helps us to grasp the significance of splits like those between mental and manual labour [so we can even explain the so-called emergent divisions among the working class on the basis of skill, as long as we go through this rather convoluted series of articulations between economic, political and ideological forms, and then between social and technical divisions of labour].

There is a distinction between a mode of production and the social formation. The mode of production already implies economic, political and ideological levels, but social formations are concrete and singular. There may be several modes of production in a social formation, maybe with one dominant mode of production dissolving or conserving the others. There are always two distinctive classes, the exploited and the exploiters, in a mode of production, but they may not appear in this distinctive form in a social formation, even though classes are still fundamental [so there can be class fractions etc left over from earlier modes of production -- the peasantry or aristocracy, say?. This is close to Marx's analysis in the 18th Brumaire...] Social formations occupy concrete sites of class struggle: thus the existence and role of actual classes cannot simply be deduced from an analysis of modes of production at some essentialist level, without an analysis of actual fractions and social groups. Fundamental classes clearly affect other classes, and there are international dimensions too to be taken into account: imperialism is an issue requiring the analysis of social formations, rather than something which can be derived from the capitalist mode of production as such.

Marxist analysis allows for the emergence of fractions and strata in classes, according to the differentiations of the economy, and the role of political and ideological levels. The categories which arise depend on the political and ideological levels too (the emergence of state bureaucracy, the rise of intellectuals and so on). These fractions and strata can become relatively autonomous, but they are still class fractions -- class is not merely just another dimension of social inequality, as in Weber.

Class determinations and class positions are articulated according to 'concepts of strategy', concepts such as a 'power bloc', or 'the people', which arise from social polarisation and the development of social alliances. However, it is still class fractions which form these alliances.

The State apparatuses have a role in the existence and reproduction of social classes. They aim to maintain the cohesion of social formations, and thus to reproduce social relations. The political and ideological levels are materialised as practices in State apparatuses [there is a clear link to the work of Althusser]. There is also an economic apparatus [producing what are sometimes known as paratechnical relations, the appropriate social relations between workers and bosses, say]. Social classes are constituted by these apparatuses. Of course, the apparatuses themselves are not independent, but are a materialisation of class relations -- quite unlike Weber's analysis again, where some separate dimension of power produces classes [see Turner for a retort] . There is no separate dimension of power, not even when describing State power: this is merely a 'condensation of class relations'. Nevertheless, State power is independent enough to provide a specific role for its specific apparatuses.

Reproduction is the key to the material existence of modes of production in social formations. Reproduction is not just located in the economic system, but in the political and ideological levels too. It is this that provides a role for the specific apparatuses, such as education.

Reproduction is a matter of reproducing places and agents. The relations that determines social classes also have to be reproduced. There is thus a material base in class struggle to explain the functioning of apparatuses. The educational apparatus, for example, reproduces, not from its specialist interest role in promoting [educational] ideas, but as an effect of production and its relations, and the struggle that surrounds them. The historical forms of struggle produce the specific forms of State apparatuses.

The splits between mental and manual labour are located not only in the economy, but in political and ideological relations and the social division of labour. Thus, for example, mental labour has an air of secrecy and control. Mental labour is used to define places [via the social role of qualifications?]. Schools emphasise this role, but this is not a necessary or abstract role: it takes place only in capitalism.

The apparatuses do not themselves create ideology, but rather materialise dominant ideology -- it is not the Church that creates religion, as in Weber's analysis, but vice-versa. The cycle of valorisation is responsible for commodity fetishism, and class struggle for the emergence of party, and not vice-versa. Finally, the organisation of production produces paratechnical relations, and not the other way about.

Finally, the role of agents takes place against the necessary background of places: bourgeois studies of social mobility show this perfectly. Training, of the kind that goes on in schools, is also an ideological training [clear links here to Bowles and Gintis on the reproduction functions of schooling]. The determining role here is played by the labour market, rather than families or schools, however [who could possibly deny this in modern Britain?].

Part 3, Chapter 3 The Class Determination of the New Petit Bourgeoisie

In terms of economic relations, this group is not a bourgeois group, since it relies on wage labour. However, wage-earners in service groups are not productive labour. Productive labour can indeed be reproductive, however. Members of this group are exploited, since their wages arise from the extraction of their surplus labour, but this arises from their unequal market position, where capital is the consumer. It is not even the case that all wage-earners are exploited in this group -- sometimes there can be a genuine exchange of equivalents. Sometimes, surplus labour is extracted through non-wage forms.

[Poulantzas wants to add to the strict definition of productive labour in marxism, developing a more general definition rather than that which is specific to capitalism: it is possible to produce surplus value while producing use values, he argues, instead of just exchange values, as in the classic analysis. An interesting aside ensues on whether the term 'productive' means the same as 'utilitarian', and whether commodities are produced solely in the interests of extracting surplus value. The point of all this is to argue that some labour may not be productive in the strict sense of the creation of surplus value, but it does make a material contribution nevertheless, playing a part in material production, even though it does not create exchange values. An example is the useful role of science in production, although even here we need to distinguish between pure and applied sciences. Pure science is not productive, not even today, even though its products take a commodity form, and even though it is invested in it and institutionalised. As a result, it cannot be associated directly with the class structure.] [So where the **** is it located, and is science the only practice that escapes class determination?].

Economic relations alone are not sufficient to understand the position of scientists, technicians, or supervisors. Economic relations separate these groups from the working class, but political and ideological relations affect their position in terms of other social groups.

Management and supervision arise from a technical division of labour, but this is related to the social division of labour, and thus to the political and ideological levels , as above. There is a tendency in Marx to emphasise the technical as economically determined, of though. It may be necessary technically to supervise and co-ordinate, but this is possible only because there is already a social divide between bourgeois and labourer. The direction [as in giving directions to ] of labour is a function of capital, since capitalist ownership penetrates all the means of production. The separation of labourers from the means of production does not arise from technical factors, but from social ones. It is always capitalist technology that produces a technical division of labour. Supervisors and managers take part in productive labour, but in a position of ideological domination. This domination is the most important aspect of their work: the function of capitalists has been delegated to them. While higher managers act as if they were capitalists, the lower ones are exploited as subalterns.

Productive labour is organised so as to permit mental specialisation. Thus mental labour can be seen as a part to productive labour, and people like engineers can be seen as working class. However, there is a real split here: the issue cannot be reduced to a simple technical one -- it shows the 'concentrated expression of political and ideological relations to relations of production'. The split between mental and manual does not correspond to the split between productive and unproductive labour, which shows that it must take a specific form at the political and ideological levels. The role of mental labour grows as production itself grows, seen best in the increasing role for science. However:

(a) Science, even pure science, is closely tied to ideology, especially where it leads to a technology. Indeed, so close is the relation, that technology can be seen as materialised ideology, and thus technologists are reproducing ideology. The monopolisation of knowledge is important here, and in so far as they support this, technologists are ideological functionaries [echoes again of the role of knowledge in class closure in Parkin?].

(b) Non-technical practitioners of mental labour can be treated similarly. They have more general legitimising functions, for example in politics and law, seen as forms of social engineering, or even as apparent opponents of the system, as in ecology [!] (239).

If this has explored the ideological dimensions of the new petit bourgeois, what of their political relations?

Firstly, it is important to see that most technical applications help the domination of labour in factories, as when work is regulated by machines. There is also direct supervision too, such as when technicians monitor efficiency. So does this help reveal a class determination for engineers and technologists? They are not members of the working class, even though they are increasingly used in productive labour [productive labour here is seen as a collective matter, rather than referring to the functions of specific labourers]. They are involved in the same contradictions as with all 'improvements' in capitalism [such as when modernisation leads to increasing competition, the destruction of economic growth, the falling rate of profits and so on]. The applications of engineering and technology do serve monopoly capitalism, but this is not enough: the main reason engineers are not members of the working class is because they maintain political and ideological relations of domination.

Even this is too simple [!], since they are important subdivisions again within this group. They also still a separate stratum, since it is not the case that the working classes are all becoming mental labourers [and there are some odd empirical bits to support this view, on page 245]. Shifts within the overall groups are not the result of technical change, but rather the result of shifts in the progress of dominance and exploitation.

Differences between skilled and unskilled labourers are not so significant. They are still recognisably members of a working class, and the shifts between them are easier to analyse compared with the shifts towards technology and technologists [more order empirical bits followed on the barriers which still exist between the working classes and technicians].

Of course, the petit bourgeois are not proper bourgeois either. They are still exploited by capital, and there is still a massive separation between them and capitalists rather than the official view of the smooth meritocratic hierarchy. In terms of political relations they are subalterns. There are victims of their own interest in the pursuit of secret knowledge as an ideology: there are liable to fragmentation just like the working class and there are petit bourgeois fractions. However, it is still possible for them to adopt either bourgeois or working class class positions [very close to Marx's own analysis here, in the Manifesto, where increasing polarisation would force the middle classes to choose the side of the working class].