partly 'descriptive', that means first and foremost that this descriptive 'theory' is without the shadow of a doubt precisely the beginning of the Marxist theory of the State, and that this beginning gives us the essential point, i.e. the decisive principle of every later development of the theory.

Indeed, I shall call the descriptive theory of the State correct, since it is perfectly possible to make the vast majority of the facts in the domain with which it is concerned correspond to the definition it gives of its object. Thus, the definition of the State as a class State, existing in the repressive State apparatus, casts a brilliant light on all the facts observable in the various orders of repression whatever their domains: from the massacres of June 1848 and of the Paris Commune, of Bloody Sunday, May 1905 in Petrograd, of the Resistance, of Charonne, etc., to the mere (and relatively anodyne) interventions of a 'censorship' which has banned Diderot's La Religieuse or a play by Gatti on Franco; it casts light on all the direct or indirect forms of exploitation and extermination of the masses of the people (imperialist wars); it casts light on that subtle everyday domination beneath which can be glimpsed, in the forms of political democracy, for example, what Lenin, following Marx, called the dictatorship of the bourgeoisie.

And yet the descriptive theory of the State represents a phase in the constitution of the theory which itself demands the 'supersession' of this phase. For it is clear that if the definition in question really does give us the means to identify and recognize the facts of oppression by relating them to the State, conceived as the repressive State apparatus, this 'interrelationship' gives rise to a very special kind of obviousness, about which I shall have something to say in a moment: 'Yes, that's how it is, that's really true!' And the accumulation of facts within the definition of the State may multiply examples, but it does not really advance the definition of the State, i.e. the scientific theory of the State. Every descriptive theory thus runs the risk of 'blocking' the development of the theory, and yet that development is essential.

That is why I think that, in order to develop this descriptive theory into theory as such, i.e. in order to understand further the mechanisms of the State in its functioning, I think that it is indispensable to add something to the classical definition of the State as a State apparatus.

Let me first clarify one important point: the State (and its existence in its apparatus) has no meaning except as a function of State power. The whole of the political class struggle revolves around the State. By which I mean around the possession, i.e. the seizure and conservation of State

-108-