Atheoretical model of class formation must undergird any serious theory of capitalist politics, just as the concrete analysis of class formation must be the prerequisite for the realistic examination of the historical development of any particular capitalist country.footnote1 So far, in contemporary Marxist research, two approaches to the problem of class formation have predominated. On one hand, there is the current of social and labour history whose unrivalled exemplar remains Edward Thompson’s The Making of the English Working Class. Primarily concerned with the nineteenth century, the main contribution of this historiographic work has been to tell us when and how, and sometimes to what extent, a distinctive, self-conscious working class first came into existence, set apart from the rest of the population. On the other hand, there is another body of research and debate which commences from the reformist versus revolutionary dichotomy, and has the aim of explaining why, when and to what extent this or that working class became reformist—as the supposedly unexpected norm—deviating from the vision of revolutionary socialism. The literature on this topic is almost inexhaustable, but its fulcrum may be said to be the debate—for and against—the labour aristocracy thesis.footnote2 Here, however, I am mainly interested in class formation as a twentieth-century explanans, not as a nineteenth-century explanadum. Likewise, I am not so much concerned with the characteristics of a given class per se, as with its capacities and achievements in conflict and in other types of relations with other classes. From this relational perspective, the question of whether and why a certain class at a certain time should be called reformist or revolutionary loses its centrality, and is replaced by a question of the capacities of a given class to act in relation to others and the forms of organization and practice thereby developed.

Behind this reformulation of the classical question is obviously the hindsight that all working classes under advanced capitalism have become predominantly reformist, thus making variations amongst ‘reformist’ classes more interesting, and politically salient, than the abstract reformist–revolutionary distinction per se. But other considerations are also involved. First, revolutions—so I will argue—do not spring so much from revolutionary class consciousness, cultivated in situ, as from revolutionary situations of institutional breakdown in which masses become revolutionized. Therefore the degree of revolutionary ideology in a non-revolutionary situation has little definitive explanatory power. Secondly, from the standpoint of a materialist concept of history, what is being done and what is being achieved are more important than what ideas are being held. Forms of practice are, typically, more interesting than states of consciousness. Thirdly, reformist and revolutionary practices and postures are important to an analysis of societal development, not in themselves, but in their effects upon social relations of power. Which type of practice is more capable of advancing the positions of a given class in a certain society in a certain period cannot be determined a priori. For all these reasons, the reformism–revolutionism dichotomy can be more fruitfully subsumed under the problematic of ‘class capacity’.

By ‘class’ we mean, with the mainstream Marxist tradition, a concept designating an aggregate of people having a common location in the relations of production. The rationale for paying attention to class and class formation is our assumption that such a common economic location ensures an inherent tendency to common collective action. In general, class is simultaneously an objective and subjective phenomenon, both something independent of members’ consciousness and something expressed in conscious thought and practice. From this it follows that ‘class formation’ must be conceived as a double process. In its objective aspects, class formation is a socio-economic process accompanying the development of a mode of production: the process of agents moving into, being shaped by, and being distributed between the different kinds of economic practices which constitute the given mode of production. In the case of the working class, this process first of all entails the formation of a mass labour-force for industry and other capitalist enterprises. In its subjective aspect, on the other hand, class formation is an ideological and political process of the tendential unification of class members into forms of common identity and of concerted action as conscious class members in relation to members of other classes. This second constitutive process is manifested in the development of class-specific collective actions and institutions. Here again we part ways with the Thompsonian current by not treating the making of a class as exclusively a process of conscious self-identification. The reason is our suspicion that the capability of a given class depends not only upon its degree of self-identity, but also upon its concrete economic location and the organizational and power resources available to it.

In a problematic specifically concerned with explaining twentieth-century socio-political developments, the questions of how and when classes are formed also raise other problems not encompassed within nineteenthcentury historiographic discussion. In particular the question of class formation no longer is pivoted around one point in time, as in enquiries of the type: When can we for the first time talk of a self-conscious working class in England? Was the German proletariat at any point ever predominantly revolutionary? When did the Swedish working class become reformist?—And so on. Instead, in a perspective of the class analysis of contemporary social change, the problem arises of how to tackle class formation as an open-ended process with no fixed destination. Classes must be seen, not as veritable geological formations once they have acquired their original shape, but as phenomena in a constant process of formation, reproduction, re-formation and de-formation.

We will, therefore, have to distinguish crucial moments or periods of the formative process. We may, for instance, distinguish a ‘founding’ moment, a possible ‘de-forming’ moment, as well as possible ‘re-forming’ moments or periods. We should also look out for ‘mechanisms of reproduction’ which prevent de-formations and re-formations. To take the case of the modern working class, the founding moment in the subjective sense would be the emergence of class-specific concerted action: the first mass labour movements. An early example of working-class de-formation in a subjective sense, on the other hand, would be what happened in England after the smashing of Chartism. Prior to contemporary deindustrialization, and the mass unemployment of the Depression partially aside, a working-class de-formation in an objective sense was always at the same time a re-formation, such as that brought about by the rise of the giant us industrial corporations at the turn of the century. In the history of capitalist dictatorships, moreover, the current military regime in Chile stands out as unique in its deliberate attempt at a de-formation of the industrial working class, not only in the latter’s conscious, subjective existence but also as a structured labour-force.