Colin Barker Archive   |   ETOL Main Page

Colin Barker

Comments on Base and Superstructure

(Winter 1987)

From Three responses to Chris Harman’s account of base and superstructure, International Socialism 2:34, pp. 125–127.
Transcribed by Christian Høgsbjerg.
Marked up by Einde O’ Callaghan for the Encyclopaedia of Trotskyism On-Line (ETOL).

Chris Harman’s stated aim in his article Base and Superstructure, International Socialism 2:32, is a non-mechanistic and non-voluntaristic version of historical materialism, with adequate room for political action and organisation. The aim is fine, how about the argument?

I find much of his argument (up to page 24) unsatisfactory, but space permits only a focus on one issue: his treatment of the forces and relations of production.

What does Chris mean by productive forces, and by social relations of production? Alex Callinicos was wrong, he suggests, in treating production relations as ‘the independent variable’. Chris reserves that role for the forces of production.

I find both positions unhelpful. To treat either as ‘independent variable’ implies a sharp distinction between the two notions, such that the relation between them is one of external causality. Thus Chris treats the productive forces as a ‘first mover’, with changes in the social relations of production always as an ‘effect’ of changed productive forces.

This conception won’t do. [1] To sustain the position, productive forces must be thought of as being quite distinct from social relations. Yet, for Marx, it is clear that some social relations were themselves ‘forces of production’. One obvious example is human productive cooperation, which is simultaneously a productive force and a form of social relations. [2]

Chris is correct: Marx’s conception of human society and its development is centred on a key idea: production. But in Marx this idea is not limited to ‘the action of human beings on their environment to get a living for themselves’ (Harman, p. 17). Human production involves not only interaction with nature, but also the making of social relations and human character. Marx’s brief formulation in the Preface – ‘men enter into definite relations’ – is much better developed in Capital, where he emphasises how workers, in producing things, also reproduce the social relations of property and exploitation. The argument is an important one for Marxism: the very possibility of socialist revolution depends on the idea that the working class is the key force producing and reproducing the social relations of capitalist exploitation. What it makes in one way it can also remake and change.

Marx’s starting point in Capital is not the production of things, but immediately socially constructed things: commodities. Certainly a commodity has to be a use-value, but its exchange-value (a predominantly social aspect) is not caused by or reducible to its use-value. The relation between them is one of mutual entailment or presupposition, not ‘cause’. Value relations between commodities express the social relations between producers in their relations with the world of things they need: they entail ‘private property’.

Chris’s account of the ‘social relations of production’ is underdeveloped: they include not simply the immediate relations within which human beings actually work on nature’s resources and forces [3], but also all those social relations which are necessarily entailed by the whole mode of production. In the case of capitalism, for example, competition is part of the relations of production (hence Marx begins Capital with the commodity), along with a developed division of labour and property, and the means to guarantee those divisions. Some form of state is an inherent necessity for capitalist (reproduction to take place, and enters into the definition of the ‘mode of production’.

Is ‘base and superstructure’ such a necessary distinction? It is not a metaphor that Marx uses much in Capital, the mature version of his Critique of Political Economy. (Let’s remember that it is a metaphor, by the way.) Marx deployed a whole series of linguistic distinctions at different times, in an effort to point to and discuss a key problem in his theory. He writes of ‘essence and appearance’, of the ‘esoteric and exoteric’, ‘abstract’ and ‘concerete’, ‘general’ and ‘particular’, ‘innermost secret’ and ‘surface form’, etc. In the Introduction to the Grundrisse he writes:

In all forms of society there is one specific kind of production which predominates over the rest, whose relations thus assign rank and influence to the others. It is a general illumination which bathes all the other colours and modifies their particularity. It is a specific ether which determines the specific gravity of every being which has materialised within it. [4]

What Marx is getting at, in these different ways, is that there is a core set of social relations which characterise a particular epoch in human production, and to which reference must constantly be made if we want to understand any particular aspect of that form of society. The character of that ‘core’ is often concealed from view, appearing quite differently, and it is the work of Marxist science to reveal it.

Is this ‘core’ a direct causal result of the development of the productive forces, in the narrow technical sense in which Chris seems to want to use the term? This seems very dubious. On the ‘basis’ of essentially the same productive forces, different core sets of social relations of production – modes of production – have developed historically. A whole variety of pre-capitalist class societies are known to us, all resting on the achievements of the neolithic revolution: settled agriculture, towns, trade. [5] Similarly, the socialist project assumes that human beings can develop a quite different core set of social relations of production on the ‘basis’ of a technology developed essentially within capitalism.

Do the narrowly conceived productive forces, then, not ‘determine’ the relations of production? Of course they do – as long as we don’t try to pack too much into ‘determination’. It is surely better to see ‘determination’ as the setting of limits to possibility of variation. Class society and the state could not be consolidated and develop unless and until settled agriculture and urban development occurred. The collective democratic rule of the associated producers across the globe – socialism – cannot be achieved without the huge extension of productive forces that has occurred over the past few centuries. [6]

To make his argument work, Chris has to treat the development of the productive forces as an inevitable process, an ultimate cause. And he has to link this to the rise and ultimate victory or defeat of a ‘new class’. But is he really so sure that the decline of Mayan civilisation in central America was due to the Mayan ruling class’s suppression of a ‘new class’? It seems equally possible that the direct producers resisted exploitation (ran away? fought civil wars? who knows?) and made the civilisation unviable. All pre-capitalist class societies – in different ways – seem to have gone through cycles of development and crisis; no very significant process of ‘repression of new technology’ seems central to the history of most of them. Relative stagnation of the productive forces, rather that some regular crushing of technological advance and of an associated ‘new class’, seems quite as plausible an hypothesis.

There have, after all, not been that many technical revolutions in human history. One was the shift to settled agriculture, which permitted the growth of class society. Its merely ‘permissive’ character is indicated, perhaps, by the fact that the archaeological record seems to suggest that a period of some 60 centuries intervened between the first emergence of settled villages in the Middle East (c. 9250 BC) and the first clear evidence of state formation (3300 BC). On the very broad ‘basis’ of settled agriculture, there emerged a whole variety of pre-capitalist ‘modes of production’, which are best characterised by their specific forms of social relations rather than simply their different technologies. (This is not to deny that technological innovations did occur, but very slowly, and hardly in a way that adequately explains very much.)

Human history might have continued to be the record of the rise and fall of different agrarian civilisations, from ancient Sumeria to Rome, from the Mayan to the Aztec states of the Americas, without any fundamental alterations in human technology. Except that, in Europe, yet another form of agrarian civilisation emerged out of the decline of the Roman empire and through ‘fusion’ with the ‘barbarian’ peoples immigrating from the East. European feudalism was not so much distinguished by its specific technology as by the fact that its social relations of production gave it a peculiar capacity for development, which went beyond both previous and other contemporaneous agrarian civilisations. Feudal Europe was a single civilisation without an effective central state, in which new kinds of towns and new kinds of trade could develop, along with certain kinds of ‘freedom’. When feudalism ran into crisis in the late medieval period, it proved capable of providing the launching pad for an entirely new kind of class society, world capitalism.

The agents of that transition were not simply Chris’s new class of ‘burgers’. [7] It is curious, indeed, how little part Chris’s historical account allows for the class struggle between the ‘old’ exploited and exploiting classes. The emphasis – apart from the last and best part, with which I have no serious quarrels – falls on conflicts between ‘old’ and ‘new’ classes. Yet the peasantry and their struggles against serfdom were at least as significant in the break-up of feudal relations. Particularly in England, the outcome of the feudal crisis was a new pattern of rural social relations based on legally free labour which proved capable of developing in an entirely new direction.

Capitalism developed first in Europe because of the social rather than the technological peculiarities of feudalism. And it was the rise of capitalist social relations, as Marx spelled out at length, which promoted the enormous expansion of the productive forces which has marked the past few centuries. That development has now made possible a new social system resting on the abolition of exploitation. In turn, that new possibility depends on the development of a new productive force: the cooperative activity of the working class. Isn’t that what historical materialism is all about?


1. Let’s remember Engels: ‘What these gentlemen all lack is dialectics. They always see only here cause, there effect. That this is a hollow abstraction ... that here everything is relative and nothing absolute – this they never begin to see. Letter to C. Schmidt, 27 October 1890, Selected Correspondence, p. 425.

2. And there’s Engels again: ‘Force (that is, state power) is also an economic power!’ op. cit., p. 424.

3. This narrow reading characterises a writer like Laclau.

4. Marx, Grundrisse, pp. 106–7.

5. Marx’s 1859 Preface mentions only two, ‘ancient’ and ‘feudal’, which is far too restrictive.

6. There is a stronger sense, of course, in which we can suggest that socialism will involve the predominance of different productive forces. But when we suggest this we are no longer excluding social relations from our conception of productive forces. Thus, for example, we can suggest that socialist production will permit cooperative rather than competitive production to flourish, that human social powers will develop, etc.

7. Does the Harmanic terminology here reflect an obsession with more rapid subsistence production (‘fast food’)?

Colin Barker Archive   |   ETOL Main Page

Last updated: 3.3.2012