It is now two decades since Lucien Goldmann published his monumental Le Dieu Cache, believing that this was but the first of a series of empirical studies which would put the sociology of literature on a sound basis. Since then, it is true, a few courses (even the odd lectureship) have been established in the sociology of literature, the rare book written which relies heavily on Goldmann's theories, such as Laurenson and Swingewood's introductory text,1 the occasional article published which tentatively applies his methods, but there has beerl virtually no extension nor even testing of his theories via major case studies. In fact the most recent contributions avowedly ignore the remarkably unworked potential of Goldmann's theories. In an article depressingly entitled 'Can there be a sociology of literature ?', Ivan RuS2 reduces this potential to no more than the portmanteau use of the concept of alienation and, thus reduced, sweeps Goldmann and the Marxists aside for a very limited and ill-defined sociology of literature indeed, somehow related to linguistic deprivation in mass society. Forster and Kenneford are equally tentative in their alternative approach derived from Parsons, but expressly reject as significant sociology a concern for art as individual action. They are particularly severe with the 'world view' approach of Goldmann. Theoretically vague and over-general, as they see it, 'it is difEcult to see what contribution is made by work of of this sort. Whatever contribution to intellectual life is made, it is not sociological theory which benefits.'3 All this seems unnecessarily pessimistic, as well as restricting. Since not only literary critics of the stature of Raymond Williams,4 but also sociologists working in other fields of theory5 value the potential of Goldmann's work, it is surely premature for sociologists of literature to forget him. There is little doubt that the 'world view' approach has itself been unnecessarily reductionist, particularly in its neglect of artistic conventions (and often, with Goldmann, arbitrary in application as well6). This is an argument of course for greater sophistication, whereas a feature of recent articles has been a narrowing down of options, and a characteristic silence on theories from outside sociology proper (such as those from linguistics and semiotics), as well as on complementary approaches from within the discipline, such as Elizabeth