Around 1958, the American artist Robert Rauschenberg undertook—over a two-and-a-half year period—a canto-by-canto ‘illustration’ of Dante’s Inferno. Rauschenberg’s method was the comparatively restrictive technique of solvent transfer on paper, with watercolour, gouache and pencil, in small (14½ x 11 inch) format—as distinct from the ‘flatbed form’ of his earlier large-scale combine paintings and in anticipation of his later use of industrial silkscreen processes. The Dante ‘drawings’ were limited to direct one-to-one quotations of found images, transferred from magazine pages or other printed matter, in what approximates a grid-like arrangement. Each transferred image was, due to the constraints of the transfer method, ‘framed,’ producing an effect that in certain respects mirrors the structural organization of Dante’s text—of loggia within loggia, of categories within categories, of a prolific regularization.

In an essay on Rauschenberg, art critic Rosalind Krauss makes the observation that the Dante drawings work in tension with the more ‘open’ forms of his combine paintings, affecting a ‘mirrorlike photographic surface’ in which the organizational logic, or techne, of Dante’s work is duplicated. The ‘veil like character of the image,’ however, produced by the transfer technique, introduces a fundamental ambiguity into the notion that these ‘drawings’ in any straightforward sense illustrate Dante’s text.[ref]Rosalind Krauss, ‘Perpetual Inventory,’ Robert Rauschenberg, ed. Brendan W. Joseph (Cambridge, MA: The MIT Press, 2002) 101.[/ref]

Krauss’ ‘reading’ of Rauschenberg commences from a distinction attempted by Roland Barthes in a by-now notorious essay on photography, in which he identifies a primarily indexical function of the photographic image—which is to say, a type of one-to-one correspondence between the image and its ‘object,’ stripped of connotation (the truth, as Derrida says, in pointing). The image itself is the product of a mechanical procedure (like Rauschenberg’s solvent transfers)—a procedure which, independent of any other consideration, is strictly an optical, photochemical process. Its ‘content’ points to its object. Whatever autonomous interpretive framework we seek to surround the image with, or impute to it, will inevitably be confronted with ‘the allegorical requirement of a master text.'[ref]Krauss, ‘Perpetual Inventory,’ 113.[/ref]

This at least was the conventional view inherited from certain mimetic traditions. What photography, by way of montage (in film) and collage (from Dada onwards), does to this view, is to explode the notion that the image depends for its ‘meaning’ (its decipherment, so to say, as though it were a rebus simply waiting to be undone) upon external referents. Consequently—in contrast to the mimetic view—not only is the image shown to ‘invoke connotational fields,’ as Krauss says, but to constitute its own reality—an idea perhaps most forcefully advanced in the 1950s by Barthes’ contemporary André Bazin, whose film theory argues for an ‘ontology’ of the cinematic image: that the reality of the cinematic ‘image’ is no less real than objects ‘in the world.'[ref]André Bazin, ‘Ontologie de L’Image Photographique,’ Qu’est-ce que le cinéma? vol 1: Ontologie et langage (Paris: Editions du Cerf, 1958).[/ref]

In discussing Rauschenberg, Krauss sees the Dante drawings as ‘a work whose very fabric is woven from the rich multiple strands of associations'[ref]Krauss, ‘Perpetual Inventory,’ 97.[/ref] — constituted both by the internal organization of the image, the techne of its composition, and its referential recodings of Dante’s text — both in terms of its topical ‘content’ and of its structure (the one appearing, in any case, to mirror the other). Rauschenberg’s image thus, in a sense, appropriates and reconfigures a Dantesque image logic. Where Dante’s text suggests certain pictorial content—as it were (in Canto IV, Dante’s ‘roar and trembling of Hell’ becomes a racing car; ‘putrid slush,’ a [presumably stinking] fish)—Rauschenberg’s ‘transfers’ evoke a metaphoricity that extends to the composition of each ‘drawing’ as a whole (within and between ‘images’). Insofar as the form of Rauschenberg’s work evokes a type of ‘engine’ of rigid designation—Barthes’ indexicality—it does so by appropriating the very rigidity of Dante’s prodigious stratifications (the circles of Hell, the mount of Purgatory, and so on.)

Rauschenberg himself commented upon the way Dante’s moral allegory coincides with a structural system—a coincidence which Rauschenberg felt increasingly compelled to account for in his own work—responding in particular to ‘the self-servicing of the text disguised as righteousness.’ On a practical level, this entailed for Rauschenberg a number of questions, of which one in particular was to address the relation of abstraction to figuration. The discipline of Rauschenberg’s ‘subjection’ to the text of Dante demanded an examination of the very nature of that subjection and of its articulation at the level of the ‘image’ and of the ways in which its reality is constituted. Krauss asks: ‘does the avowed desire to break with abstraction … demand figuration and textual support?'[ref]Krauss, ‘Perpetual Inventory,’ 113.[/ref] Or, might we say, is (dis)figuration the outcome of an attempt to come to terms with the apparently abstract status of the image itself and of its structure? The result, in either case, is a serial, disciplined regularity which derives from a ‘matrix of slippages’ and ‘veils’ a system. Not the closed system of Dante’s organized vision of Christian metaworlds, but the generative recombinatory system of textual structures that underwrites it. For this reason we can say that Rauschenberg’s Dante drawings are critical, and neither merely denotative or allusively connotative.