CRITICISM ANTONYMY, LANGUAGE, AND VALUE IN CONRAD's HEART OF DARKNESS I Arnold Krupat . . .Conrad's writing. . . has a critical place in the history of the duplicity of language which since Nietszche, Marx and Freud has made the study of the orders of language so focal for the contemporary understanding. Conrad's fate was to have written fiction great forits presentation, and not only for what it was representing. 1 Joseph Conrad's Heart of Darkness has been interpreted from nearly every conceivable perspective, noris there, even now—the recentarticles by Ian Watt are significant instances—an abatement of commentary on the story.2 It is only from the mid-'sixties, however—from thebeginnings of widespread American attention to French "new criticism" with its exaltation of the linguistic model as the paradigm for the study of literature—that American attention to "the orders oflanguage" inHeartof Darkness may be dated. The application of certain structuralist and post-structuralist insights to the story may well appear to some as superfluous as it is inevitable: but this may also offer some further understanding of the story's well-known problematics.3 What I understand Conrad to be trying to do inHeart ofDarkness is to "represent" the discovery of moral value. This effort is founded upon a view oflanguage which assumesa directcorrelationbetween the word (or sign) and its referent (either a concrete quantity, a "thing," or an abstract quality, a "moral value") enabling an "imitation" of a fairly stable, ontologically prior "reality." Such an "imitation" need hardly offer itself as a straightforward mirroring of the "world"; rather, as Cesare Segre writes, because the writer "privileges only certain aspects of reality, emphasizing themin relation tohis personalvision ofthe world," literary mimesis can offer a "critical representation," one polemical, or obviously deformed.4 We can determine the kind and quality of the "representation " by examining its mode of "presentation," the distinctive features of diction and syntax, or its "style." "Representation" and "presentation" are conceived of as complementary andcontinuous: the "vision" requires a certain style to convey it and only the peculiarities of that style can convey the "vision," or constitute the work's "meaning." Literary language is thus cognitive and constative according to this view which we may call the mimetic or semantic view of language. To adhere to this view of language is what Conrad's experience as a writer everywhere revealed to be impossible. The contemporary not only of Nietszche, Marx, and Freud, Conrad is the contemporary of Mallarmé The Missouri Review ¦ 63 and Saussure as well; and he, too, feels the slippage between word and referent characteristic of his period.5 Conrad suspects what the Swiss professor of linguistics will assert,6 the necessity to break the unitary sign (or word) into its constituent parts, the signifier and the signified, and to shift the emphasis to what can be specified with some assurance, the order of the signifier, and away from what cannot, the order—the ontological and epistemologjcal status—of the signified. This movement to what may be called the semiological view of language has important consequences for the representational project. For, as Fredric Jameson has written, "Ultimately, if the process of thought bears not so much on adequation to a real object or referent, but rather on the adjustment of the signified to the signifier...then the traditional notion of 'truth' itself becomes outmoded."7 Itis, from here, only a short step to the assertion of the unqualified "primacy of the signifier," and—I cite now a more recent formulation by Jameson—the "conclusion that therefore the 'referent '... does not exist."8 This, of course, is exactly the "conclusion" of those who adopt the radical form of the semiological model we may call, from Derrida, the grammatological. Thus Hillis Miller unequivocally states that "all referentialityinlanguageisa fiction."9At this point, to speak of"representation " in literature is simply to make a categorical error; and "presentation " can indicate no more than the production of the "text" as locus for the play of an infinite galaxy of self-reflexive signifiers. To Nietzsche's question, Who speaks? there is only Mallarmé's answer, The Word.10 But—to return to Nietzsche—there can be no "correct...