MODES OF POWfcK 59 to rest on a non-economic basis, such as the distinction of elders and juniors or male and female. It seems to me that a satisfactory resolution of this debate crucially hinges on our ability to adequately conceptualise the political and the cultural-ideological instances of a social formation in their relations to the economic instance. In the particular case of the formation of classes in lineage societies, this will necessarily mean a. concrete conceptualisation of the community and its relation to a specifically political institution such as the state. Marxist anthropologists have made a number .of observations in this regard which are specific, ad hoc and limited. I do not think a satisfactory conceptualisation of the community or the state in the evolution of lineage societies has yet emerged. My fundamental argument then is that an explicit conceptualisation of the political instance of social formations (as also the ideological-cultural instance) is essential if we are to address ourselves to the problem of the transition from one mode of production to another. I reject the view which asserts that the political forms are merely aspects of the 'superstructure' and thus have only a residual role to play in historical analysis in terms of the mode of production. I also believe that such a view can be shown to be fundamentally inconsistent with Marx's own discussions of the transition problem in several places in his later work. Take, for instance, Marx's identification of the fundamental difference between the Asiatic and the feudal mode of production. It is a different matter altogether to judge whether the concept of the Asiatic mode of production is useful in understanding the histories of countries such as India or whether Marx's evidence on these countries was accurate or adequate. But there is no doubt that Marx repeatedly located the fundamental difference in the absence of a notion of'private propert/ in the Asiatic mode. To say that this was merely a difference in the 'legal-juridical form' of the rights over the means of production would lead us to the sort of mechanistic classificatory scheme drawn up by, say, Barry Hindess and Paul Hirst in theii Pre-capitalist Modes of Production (London, 1977) in which they concluded that the Asiatic mode could not be defined as a distinct mode of production. The fact is that the difference in the so-called 'legal' form is crucial in determining the possibilities of a transition to the capitalist mode. Marx repeatedly argued that it was the specific process of development of the legal form of'property' in Europe, from the ancient formation in the Roman period, its breakdown and the evolution of feudal property, which was a crucial element in facilitating the subsequent separation of the labourer from the means of labour in the period of the transition to capitalism. A quite different configuration in the relation between producers, the 'village community' and the state, and thus a different notion of'property', was to Marx the main condition which impeded the possibility of a separation of the labourer from the means of labour in the Asiatic mode. Once again, the question of how far Marx was accurate in his historical description of Asiatic countries is'another matter, but the thrust of his analysis clearly highlights the point that the political form of the class struggle in a period of transition is a struggle over the redefinition