that together define it as a recognizable historical entity. Production is therefore fundamental in a universal and unique sense, and a scientific approach to the understanding of history has to take this as its starting point. Furthermore, it is obvious that the possibilities of production at any given time and place establish narrow, though certainly not rigidly defined, limits and constraints on what a particular society can actually accomplish.*
But when it comes to the processes of change that are always going on in every society (or mode of production if you prefer), even if at times they are so slow as to be practically invisible, there is no a priori reason to assume that they must have their origin or derive their impetus from the realm of production, as prescribed by the schema of the Preface. As Marx and Engels wrote in The German Ideology:
The fact is … that definite individuals who are productively active in a definite way enter into … definite social and political relations. Empirical observation must in each separate instance bring out empirically, and without any mystification and speculation, the connection of the social and political structure with production. (3rd revised ed. [Moscow: Progress Publishers, 1976], p. 41, emphasis added)
My attention was called to this striking formulation by the recent book Socialist Construction and Marxist Theory, by Philip Corrigan, Harvie Ramsay, and Derek Sayer. The authors go on to draw the appropriate conclusion:
On this premise, Marxism could sustain no general theory of the “connection of the social and political structure with production” comparable to the classical base/superstructure conception. For any such theory must involve either a priori or inductive generalization, and Marx rules out the first when he requires that we ascertain this connection “empirically,” and the second when he enjoins us to do so for “each separate instance.” (New York: Monthly Review Press, 1978, p. 6)
There is, of course, no contradiction between this view and the schema of the Preface if we interpret the latter, as I have done here, not as a statement of a general law of history but as a shorthand summary of findings empirically derived from a study of the separate instance of capitalism. Like all shorthand summaries, it is naturally oversimplified: it provides a useful framework for further study of capitalism but by no means a rigid or infallible formula. That this is the way Marx and Engels themselves understood the matter is amply demonstrated by their own historical writings that show few if any traces of base/superstructure orthodoxy.
Engels’ famous letters to younger German followers in the years after Marx’s death (Conrad Schmidt, Bloch, Mehring, Starkenburg) should be interpreted in the same sense. These were mostly written in response to questions about the role of the economic factor in historical materialism, questions that reflected a strong tendency, which still exists today, to turn historical materialism into a form of economic determinism. Engels argued against this tendency, without denying, however, that there was some ground for it:
Marx and I are ourselves partly to blame for the fact that younger writers sometimes lay more stress on the economic side than is due it. We had to emphasize this main principle in opposition to our adversaries who denied it, and we had not always the time, the place, or the opportunity to allow the other elements involved in the interaction to come into their rights. But when it was a case of presenting a section of history, that is, of a practical application, the thing was different and there no error was possible. (Engels to Bloch, September 21, 1890)
It is true that some of Engels’ formulations in these letters suggest that the noneconomic elements in the “interaction” affect the form and timing rather than the substance of change and in this way give support to a doctrine of “determination in the last instance” that is pretty close to an economic determinism. But it is also possible to consider these formulations as an indication that Engels himself had not completely overcome the tendency to overstress the economic factor mentioned in the passage cited above from the letter to Bloch. Once the principle of interaction has been admitted, it is no longer possible to salvage a general law of the determination of noneconomic elements by economic (or superstructure by base). The problem becomes the one Marx and Engels had already identified in The German Ideology, i.e., the “empirical observation … in each separate instance … [of] the connection of the social and political structure with production.” This does not rule out generalizations for which an empirical basis can be established, but it does most certainly rule out universally valid laws of history.
* This seems to me to be the rational kernel of the concept of “determination in the last instance” to which some Marxists—notably the French philosopher Louis Althusser and his followers—have assigned a privileged position in their version of historical materialism. In my view, however, the determination which is said to prevail in the last instance is of a largely negative kind and hence of little value in explaining what happens as distinct from what does not happen. The danger is that in the hands of uncritical practitioners “determinism in the last instance” tends to become a mere formula, a pseudo-explanation that closes off rather than opening up the way to a serious analysis.
[ 2 ]
THE CONTRADICTIONS OF CAPITALISM
In this lecture I am going to focus entirely on capitalism, and I want to say right away that in this area it seems to me Marxist theory has made important forward strides in recent years. These advances have of course not been evenly assimilated, and it may even be true that an appreciation of their scope and importance has hardly penetrated to what may be called the textbook level, where what is taught seems generally to be a more or less adequate summary of selected parts of Marx’s Capital. Not that this is necessarily bad: a good case could be made that the best introduction to Marxism will always be Capital itself. But capitalism has greatly changed and expanded in the last hundred years, and its analysis requires that the theory expounded by Marx should be supplemented and to some degree modified to take account of these developments as well as of the vastly increased amounts of knowledge that a century of accumulated research has provided. My purpose in this lecture and the one to follow is to try to sketch, in desperate brevity as Schumpeter used to say, the main outlines of an overall Marxist theory of capitalism in the last quarter of the twentieth century.
A commodity is something—a good or a service—produced for sale, not for use. All societies since the most primitive have been characterized by some commodity production, but only under capitalism has it become the dominant type of production; and only under capitalism has labor power, the capacity of the worker to perform useful labor, become a commodity, not exceptionally but in general. Workers, however, would not sell their capacity to perform useful labor to others if they possessed the means and materials of production necessary to produce goods and services for their own account, i.e., either for direct consumption or for sale on the market. It follows that the very existence of capitalism implies that a tremendous and indeed traumatic upheaval has already taken place in the structure of social relations. Producers, especially peasants, have been uprooted and separated from their traditional means of producing and acquiring a livelihood, with the result that they are obliged to sell their labor power in order to keep alive and reproduce their kind. And, for this to be possible, there must be another class of people in existence who possess means of production and enough money or capital to buy labor power and materials that can be combined in a process of producing commodities for sale on the market. Capitalism, therefore, comes into the world in the wake of two great and more or less contemporaneous historical processes—the formation of a propertyless working class on the one hand and of a property-owning capitalist class on the other (the former’s lack of property as well as the latter’s ownership of it presuppose and necessitate a coercive state that is, therefore, as essential to the existence of capitalism as the workers and capitalists themselves).
These twin processes of class
