The historic debates of the 1930s between Ernst Bloch, Georg Lukács, Bertolt Brecht, Walter Benjamin and Theodor Adorno have now been assembled into a single volume, with an Afterword by Fredric Jameson.footnote1 Readers of nlr have already had a foretaste of its contents: Brecht’s sardonic deflation of Lukács, for example, published in nlr 84, has been absorbed with significant rapidity to swell the slim corpus of a Western Marxist aesthetics in dire need of nourishment. But it is only with the coherent ordering of these complex interchanges that we can pose to them some fundamental questions. Why does this pivotal debate take the recurrent form of a quarrel over ‘realism’? What is the political secret of these varied contentions over painting, theatre, fiction? And how are we to receive and appropriate these polemics today?

Consider this curious paradox. A Marxism which had for too long relegated signifying practices to the ghostly realms of the superstructure is suddenly confronted by a semiotic theory which stubbornly insists upon the materiality of the signifier. A notion of the signifier as the mere peg or occasion for a signified, a transparent container brimfull with the plenitude of a determinate meaning, is dramatically overturned. On the contrary, the signifier must be grasped as the product of a material labour inscribed in a specific apparatus—a moment in that ceaseless work and play of signification whose sheer heterogeneous productivity is always liable to be repressed by the bland self-possession of sign systems. A centuries-old metaphysic of the signified is rudely subverted: the signified is no more than that always half-effaced, infinitely deferred effect of signifying practice which glides impudently out of our reach even as we try to close our fist upon it, scurrying back as it endlessly does into the privilege of becoming a signifier itself.

In trying thus to close our fist upon the signified, we are in fact attempting nothing less than the risible task of nailing down our very reality as human subjects. But what we will nail down, of course, will not be the subject, but the paranoic knowledge of the ego and its various identifications. In this ceaseless cat-and-mouse game, the subject, which is no more than the effect flashed cryptically from one signifier to another, the ‘truth’ which can be represented only in a discourse from which it is necessarily absented, will hunt frantically for its self-recognition through a whole fun-hall of mirrors, and will end up fondling some fetishized version of that primary self-miscognition which is, in Lacanian mythology, the mirror phase. Terrified of the very linguistic productivity of which it is the endlessly transmittable effect, the subject will attempt to arrest the signifying chain in order to pluck from it some securing signified—a signified within which subject and object will blend infinitely into each other in an eternal carnival of mutual confirmation. The literary names for this are realism and representation—those recurrent moments in which the comedy of writing—the incongruous flailings by which, in heroically attempting to ‘refer’, it will finally do nothing but designate itself—is gravely repressed for the ritual enthronement of some unblemished meaning which will fix the reading subject in its death defying position.

Yet there is a problem. For Marxism was supposed to have something to do with ‘history’; and history, in this schema, seems effectively to have evaporated. It is not only a question of the signified, which is now no longer Saussure’s obedient if arbitrary echo of the signifier, but an effect which slides along in its own sweet way, sporadically buttoned by a chain of hegemonic signifiers which moves athwart it. It is also a question of the referent, which we all long ago bracketed out of being. In re-materializing the sign, we are in imminent danger of de-materializing its referent; a linguistic materialism gradually inverts itself into a linguistic idealism. In evolving a practice upon literary texts which every English University greets with a certain nervous contempt, we have played straight into the hands of the Yale English school. For nothing could suit that particular group of modish academics better than the notion that even ‘history’ is a text. The proletariat is to be neither supported nor repulsed; it is to be textualized. And since we have rejected the theological concept of hierarchies, then it is the crassest folly to insist that the proletariat is in principle a more significant text than Adolphe or Crotchet Castle.

Now that ‘history’ is a text is surely true. But it follows that the text of history, like any other, is constituted by certain determinate absences and contradictions which call for that symptomatic reading which is historical materialism. It also follows, as Fredric Jameson has remarked, that if history is a text it is a text-to-be-(re-)constructed.footnote2 And that (re)writing of the historical text which is socialist revolution has little in common with that sportive, privatized re-fashioning of the unreadable which characterizes the more decadent works of Barthes. The re-materialization of the signifier has indeed provided the most fruitful impetus in recent Marxist aesthetics; but there is a sense in which, like Wittgenstein’s Tractatus, and for rather similar philosophical reasons, it leaves everything exactly as it was. It is instructive to be told, as we are by the later Macherey, that the problem of literary ‘realism’ inheres in the real conditions and possibilities of a text rather than in some imaginary notion of truth-to-life. But one does not wholly escape a Lukácsian problematic merely by shifting the ground of debate from epistemology to ontology.