Iam grateful for Francis Mulhern’s penetrative comment on my essay, which raises some methodological problems of first importance to Marxist criticism. He is right, I believe, that the necessarily elliptical character of my article excluded certain central issues (notably, ideological determinations other than those of ‘class situation’, and the writer’s submission to an ineluctable lineage of literary forms) to the point where some of my formulations could reasonably be accused of residual historicism. But the thrust of the essay was, in intention if not in effect, quite different. Rather than grasping ideological positions as totalizations of class-determined authorial situations, my point was to demonstrate the relations between the ideological sub-ensembles in which the authors were historically set, and the dominant ideology itself. It is this, not some ‘expression’ of class-ideology, which sets the ideological matrix of the fiction itself. The ‘real’ ideological situation of the writer is significant, not as the source of the text’s ideology, but as an index of the specific mode of insertion of an ideological sub-ensemble into the dominant ideology as a whole.

The ‘ideology of the text’, then, is neither merely the hegemonic ideology of its historical moment, not the mere effect of its formal operations. It is, rather, the effect of literary form’s transmutation of an ideological matrix partly determined by the writer’s own specific mode of ideological insertion into the dominant ideology. It is the writer’s partial mode of access to the ideological ensemble—as Lenin sees in the case of Tolstoy—which is one (though only one) determinant of the mode of refraction of that ideology into his text. It is true that that refraction is then, so to speak, doubly refracted by literary form; and Mulhern is certainly right to charge my essay with paying too little attention to this process. Yet to claim that the text’s ideological positions are merely the determinate effects of its form is surely to attribute to form an unacceptable degree of autonomy, as well as to by-pass the question of what precisely it is that form ‘works’. Ideological and aesthetic structures are certainly irreducible to one another; yet they are not mysteriously independent of each other either. O’Casey’s humanism is indeed, but surely not merely, the effect of his naturalistic forms; for those forms mutate materials which contain the historical possibility of such humanism—contain it, indeed, as the tired other face of revolutionary struggle. Similarly, the doubled narrative of Heart of Darkness is in my view a way of formally negotiating, rather than definitively transcending, the ambiguities of Conrad’s attitude to imperialism; it is used elsewhere in his work for precisely such purposes. A writer’s ‘textual’ and ‘extra-textual’ ideologies may certainly be distinct; but the relation between them, and its historical variability, is measurable and determinate rather than arbitrary and opaque. It is precisely this point which Lukics fails fully to grasp in his studies of Balzac and Tolstoy, attributing the productive contradiction between their art and ideology to their ‘inexorable veracity’ and artistic ‘genius’ as individuals, rather than to conflicts and capacities within their ideological positions which their fiction activates.

I agree with Francis Mulhern that the ideology of the text is internal to it—but it is external also. The correct opposition is not, as Mulhern appears to suggest, between ‘textual’ ideology and the ‘extra-textual’ ideology of the author. It is between ‘textual’ ideology and an objective ideological formation—a formation of which the text is a unique transmutation, and authorial ideology a determinate structure.

Terry Eagleton