hung on the wall of the Sociology Staff Room? That is not very easy to say. DIT only got its name from its killers, just before the trigger was pulled.

A second, closer reading of DIT reveals a curious structure of the book. DIT is first defined by general reference to a number of Marxist theorists, then it is refuted by a series of arguments concerning what AHT hold to be false notions about the operation of ideology in feudal society and in early and late capitalism. This procedure assumes, with no systematic attempt at demonstration, that the criticized notions of feudal and capitalist ideology are those of the authors whose writings constitute the DIT. DIT contains a host of references, but the ones decisive for its authors' argument are conspicuously absent. A common and respectable procedure of scholarly debate is first to give a clear picture of what is to be scrutinized and criticized, and then to show the logical inconsistency of the object of analysis or to demonstrate its empirical inadequacy or falseness by bringing evidence to bear against it. For some reason, however, AHT have chosen a quite different path. The criticandum, DIT, is first defined in three different ways. Then the authors pool their knowledge to cast as much doubt as possible on one of the three objects of definition. The conclusion is that DIT is 'empirically false and theoretically unwarranted', presumably in all three meanings. To most people this will hardly be a convincing demonstration, however sympathetic they may feel towards much of the book's anti-idealist thrust. It remains to be seen whether AHT have arrived at a correct position, even though they have not succeeded in bringing their arguments together in a logically compelling way.

The three definitions of 'the' DIT which AHT offer are the following. First, what we might call the 'identifiable DIT' is defined by reference to known authors 'such as Habermas, Marcuse, Miliband and Poulant zas ' (p. 1), Gramsci, Habermas and Althusser' (pp. 11ff.). Secondly, we find something like a 'stress definition' of DIT: 'Our argument is that there has been an increased emphasis on the autonomy and causal efficacy of superstructural elements, and of ideology in particular, in modern Marxism. . . . This emphasis on ideology amounts to advocacy of what we have called the dominant ideology thesis' (p. 29 ). The third and final definition is of a 'constructed DIT', a product, most immediately, of AHT's talent for formulation:

-170-