On Filmic Rewriting
Contamination of the Arts or Destruction of Art’s Identity?

Marie-Claire Ropars

I would like to attempt here an oblique take on the general project of this volume [Porous Boundaries. Texts and Images in Twentieth-Century French Culture]. This will involve interrogating the notions of contamination and transfer that are thought to govern the relationship between literature and the visual arts by confronting them with another proposition which takes into account the singularity of the repetition experienced in cinematic remakes of literary works. Rather than cinematographic transformation, which assumes the grafting of an art onto language, I will oppose to that classical model of adaptation the hypothesis of a rewriting which proceeds by duplicitous reflection and identity-destruction: one art – literature – comes to be reflected in another – cinema – which removes from it its own power of representation in order that within language itself, only the outline of a writing, without meaning, is permitted to remain.

This dispositif of destruction governs the cinematographic experimentation of Marguerite Duras, whose work of defection will be analysed by way of the example of the Aurélia Steiner films; but it illuminates numerous examples of film-texts just as well, whether by Frears, Visconti, Oliveira or Astruc, whose film remakes create fault lines in the source texts. The approach taken to its extreme by Duras the author – Détruire dit-elle – actually develops a more common aesthetic hypothesis which sees in the plurality of arts a challenge to the possibility of an identificatory approach to each art form. It is this originary heterogeneity of all thinking about art that will be brought into play in order to make the connection between the encounter of the filmic and the literary and one of the critical aims of literature, itself considered as an art.

The modern invention of aesthetics rests, in fact, on a paradox the full extent of which must be uncovered. On one hand, with the founding gesture of Lessing, a split is declared between the different arts, in particular between the plastic arts of which sculpture provides the prototype, and the verbal arts of which Homeric poetry constitutes the model: established in their plurality and their specific difference, the arts are separated one from the other, ceasing to obey rhetorical precepts of imitation and mutual subordination; each one becomes autonomous as art, and each art is found to be unified in itself. On the other hand, paradoxically, the declared partition does not put an end to the movement of attraction, particularly in literature: in German Romanticism, French poetry from Baudelaire to Valéry and a certain Anglo-Saxon line of fiction from Poe to James, the attraction of another art does not cease to manifest itself nor to form the basis of each art’s reflection on itself. The attraction of the other thus becomes constitutive of a non-unitary thinking about art that is most acutely formulated by Adorno, but which can be found at work in the very foundations of aesthetics.

1. Gotthold E. Lessing, Laokoon, oder über die Grenzen der Malerei und Poesie (1766); Laocoon (Paris: Hermann, 1990) for the French edition. See especially chapters XVI, XVII and XVIII (pp. 120–37 in Hermann). English translation: Laocoon (New York: Noonday Press, 1969).

2. Georg W. Hegel in B. Teyssèdre (ed.), Esthétique des arts plastiques [extract from Cours d’esthétique, 1828–1829] (Paris: Hermann, 1993), pp. 65–7 (Book I, chapter II).

3. Charles Baudelaire, ‘Exposition Universelle’ (1855), reprinted in Curiosités exthétiques (1868). See chapter III, ‘Eugène Delacroix’. All other quotations are from the collection of Baudelaire’s texts, Pour Delacroix (Paris: Editions Complexe, 1986), p. 132 and p. 129.

4. Edgar A. Poe, ‘The Oval Portrait’ (1842); Balzac, Le Chef-d’œuvre inconnu [The Unknown Masterpiece] (1831); Henry James, The Author of ‘Beltraffio’ (1884), published in French in the collection Les Deux visages (Paris: Les Lettres Nouvelles/Maurice Nadeau, 1977). For a confrontation of these three texts, see my study: ‘En rêvant à partir de fictions énigmatiques ... ’, in Marie-Claire Ropars (ed.), L’Art et l’Hybride (Saint-Denis: Presses Universitaires de Vincennes, 2001).

5. Theodor W. Adorno, ‘L’art et les arts’ (‘Die Kunst und die Künste’, conference held in Berlin in 1966, published originally in Anmerkungen zur Zeit, no. 12, Berlin, 1967). My quotations are taken from the collection of four texts by Adorno entitled L’art et les arts (Paris: Desclée de Brouwer, 2002), pp. 43–74, here pp. 54–5.

6. Jean-Luc Nancy, ‘Les arts se font les uns contre les autres’, in Marie-Claire Ropars (ed.), Art, regard, écoute – La perception à l’œuvre (Saint-Denis: Presses Universitaires de Vincennes, 2000), pp. 157–66. See also by the same author ‘Pourquoi y a-t-il plusieurs arts et pas un seul?’, in Les Muses (first part) (Paris: Galilée, 1994; English translation: Stanford University Press, 1997).

7. Adorno, op. cit., p. 45.

One will not be surprised, then, to find that the same modes of attraction continue to govern until the Hegelian classification of the arts. The pyramid constructed by Hegel rests on the very categories of Lessing, which permit him at once to measure out the progressive degrees of abstraction from painting to poetry and to bring forth those sensible echoes by which music can emerge from painting and be fully accomplished in the poem: for Hegel, one art is subordinate to the other, which nonetheless reflects and returns it to itself. (2) It is this reflexive capacity that Romantic poetry places at the heart of a new foundation: the self-constitution of literature, as conceived by German Romanticism, finds support in the mirror of a pictorial mediation; for Baudelaire, literature thinks itself by reflecting itself in painting, which ‘projette sa pensée à distance’ [‘projects its thought into the distance’] and thus brings the poet ideas; whereas the dramatisation of colour and movement in Delacroix translates the painting’s aptitude for making itself a memory of the dramas transmitted by history and literature. (3)

Attraction is reciprocal and attempts at resolution can operate either to the advantage of poetic language – the highest degree of abstraction according to Hegel – or in the wake of a pictoriality which alone is capable of lending plenitude to speech by concealing a poetic aporia. But this double movement, where discursive synthesis and sensory seduction are alternately at play, does not disrupt the partition of the arts except to reintroduce a new division with a dominant feature restored to both sides of the mirror. Many works of fiction bear witness to this: the attraction of the oval portrait in the eponymous short story by Poe is resolved by its disappearance from the text; the ‘unknown masterpiece’ of Balzac’s painter is all the more desirable since it must be destroyed; and the pictorial ambiance given off by the family of the writer of ‘Beltraffio’ is effaced with the death of the child-portrait and the return to the mere reading of written works. (4) Fiction thus traces the rule of a paradoxical rejection: why construct a work from pictorial attraction if this can only be accomplished in the destruction of the painting in question?

Attraction goes hand in hand with rejection; the partition of the arts is at once resisted and required. Without putting an end to this bind of secession-attraction, Adorno proposes an explanation of its origins in a short essay entitled ‘Art and the Arts’, where he comments on the artistic practices of the twentieth century: if the arts do not cease to encroach on one another, if their mutual ‘unravelling’ is the rule of a modernity where literature is made serial in the image of a music itself nourished by a new plasticity of figure, it is because they respond to the very insistence of the idea of art, heterogeneous in its principle: ‘to become art, art requires something that is heterogeneous to it’. (5) The consequence of this theoretical displacement will be fundamental to our thinking: if heterogeneity institutes the possibility of art, then the interaction of the arts can only resolve itself in some form of hybridisation; however, it is this paradoxical movement of attraction without synthesis which establishes at once the principle of an aesthetic idea perpetuated solely by alteration and the principle of its disruption. Art persists only in becoming other in itself, and this attraction of the other is at work in the artistic encroachment according to which one art – or technique? – ceaselessly eats away at another without in any way being assimilated to it. For this reason, uninterrupted attraction becomes the condition for realisation of each of the arts: ‘les arts se font les uns contre les autres’ [’the arts are made one against the other’], says Jean-Luc Nancy some time later in an ambiguous formulation inspired at once by Hegel and Adorno. (6) But, whereas Nancy intends to preserve the unity of each art through the act of making, so that an art is understood to take shape in itself and find meaning in the process of incarnation, one must emphasise to what point Adorno insists instead on the ruin of a totalisation not only of the arts, but of each art in itself.

There can be no synthesis of the arts – hence the pursuit of attraction; but no more is there internal unification of an art, the need for which seems once more to determine alteration – might this explain rejection? Adorno hesitates on this point, markedly so in his reflection on music, which is alternately seen to participate in the general mutual unravelling of a divided aesthetics, and to bring into play an essential alteration in the terms of its internal logic. (7) Adorno neither resolves nor even addresses this ambiguity; but it becomes decisively important for an ultimate question that he does not formulate: does heterogeneity pertain to a negative principle of art as such, or does this negativity concern each of the arts, which could only become themselves by becoming alien to themselves? It is not possible to choose between the two terms of this alternative; but the second – that would pursue a negative identity of each of the arts – could singularly illuminate the coexistence of attraction and rejection between the arts: if one art finds its reflection in another, thus literature by painting and painting in music, then that represents a deflection that is at once necessary – in order to manifest the attraction of the other – and disavowed – the operation can only be intrinsic to one of the two. Hence the disappearance of the portrait, the putting to death of the child-painting, the burning of the ‘muraille de peinture’ [‘wall of painting’], in Balzac’s terms, which charge the text alone with the responsibility to become other. At stake in the attraction of the outside will be the transferral of exteriority to the very principle of identity.

8. Paul Valéry, Introduction à la méthode de Léonard de Vinci (1919) (Paris: Gallimard, 1957), p. 129 and p. 134.

9. Marguerite Duras, India Song film (1975), Son nom de Venise dans Calcutta désert film (1976). A storyboard [découpage synoptique] in three columns (two different visual tracks to either side of a shared soundtrack) is to be found in L’Avant-Scène Cinéma, no. 225 (April 1979). For Aurélia Steiner, see note 17.

10. In relation to this, see Emile Benveniste’s well-known ‘Sémiologie de la langue’, which first appeared in Semiotica I (The Hague: Mouton, 1969), 1, pp. 1–12 and 2, pp. 127–35. Benveniste, analysing the double meaning of language (semiotic and semantic) considers that this privilege gives it the capacity to interpret all systems of expression. English translation: ‘The Semiology of Language’, in Robert E. Innis (ed.), Semiotics: An Introductory Anthology (Bloomington: Indiana University Press, 1985).

Duras’ experiments follow this line, but she forces an illuminating rupture in it: the double practice, successive then simultaneous, of literature and cinema, underwrites a speculation on the status and function of the film remake, where filmmaking is charged from film to film (India Song and Son nom de Venise dans Calcutta désert, divided by a common text), and from text to film (Aurélia Steiner, book and film in turn), with unmaking the text in the process of repeating it. (9) This system of double input, which subverts the partition of the arts, relies on the practice of an artistic deflection which would alone be capable of disrupting the written object, by amplifying it: nor is this an external mirror, as is da Vinci for Valéry, but instead an internal remodelling arising from a dual practice of the languages of literature and film, one by the other, one against the other. In this sense, experimentation in filmic rewriting would only be an exacerbated modality, and exemplary in this respect, of an aesthetic and not poetic constitution of literature. The risk must be accepted: not the invention of a new art, proceeding by hybridisation, but instead the collapse of a unitary conception of the work of language, whereby the privileging of language has made of verbal procedures the sole interpreter of themselves as well as of all other languages. (10) What is belied by Duras’ filmic essays, which is even more paradoxical, and for this reason effective, is that each film, made as reflection and interpretation of a reading text, is submitted in turn to the principle of destruction of which it is the carrier: partition is restored, but the work altered forever. Yet an investment in the singularity of cinema is required in order to support such experimentation.

Cinema and Literature: An Aesthetic Capture

In what way might cinema be fit to exercise a tighter critical grip on literature than other arts? Two theoretical perspectives are implicated here, dealing with the heterogeneous character of cinema and with the singular notion of cinematographic rewriting.

11. André Bazin, Qu’est-ce que le cinéma?, vol. II: Le Cinéma et les autres arts, I: ‘Pour un cinéma impur: défense de l’adaptation’ (Paris: Cerf, 1959), pp.7–32. English translation: ‘In Defense of Mixed Cinema’, in Hugh Gray (ed. & trans.), What is Cinema? Volume 1 (Berkeley: University of California Press, 1967), pp. 53-75.

First of all, recall the evidence: the plurality of visual, graphic and audio materials and the horizontal and vertical multidimensionality on which the making of a film depends lend cinema a strong material and formal potential for heterogeneity. To attempt to capture its specifity by reducing it to the moving image is to ignore the real basis of this seventh art, the singularity of which is that through musical accompaniment, superimposed text and the physical arrangement of scenes, it joins the multiple traces of all artistic materials at the very outset: and if it were necessary to privilege the figurativity of the filmic, it would be to the benefit of the movement that animates the image, and not the static figure as such. André Bazin forcefully emphasises that cinema is by definition impure. (11) But from this technical and sensory impurity emerge contrary approaches to the artistic practice of film: to proceed by syncretism, borrowing from all sides – from literature, from fairground spectacle, from melodrama or the serial, from orchestral score to décor to architecture – free to unify the ensemble in a dominant narrative mode; or alternatively to play incessant alterations, the emergence of one from the other, the renewed explosion of forms by an excess of quotation that comes to double each visual or verbal event with a memory, concealed or advertised, removing it from its specific frame. On one side Hitchcock or Carné, on the other Godard or Greenaway. Cinema becomes art only by responding, contradictorily, to the attraction of other arts, the heterogeneous factor of which persists in the diversity of solutions retained.

12. See in particular the different studies collected by Sergeï Eisenstein (on Zola, El Greco, Piranesi, landscape music or the montage pathétique of Potemkin ... ) in his unfinished work, La non-indifférente nature, vol. I and II (Paris: UGE, 1976 and 1978 for the French edition).

13. Alain Resnais, Hiroshima mon amour (1959); screenplay and dialogue by Marguerite Duras (Paris: Gallimard, 1962; English translation: Grove Press, 1961). Resnais, Providence (1976); screenplay and dialogue by David Mercer who, in 1977, published Providence, un film pour Alain Resnais (Paris: Gallimard).

14. Adorno, op. cit., pp. 72–3.

15. Alexandre Astruc, Le Rideau cramoisi (1952); remake of a story by Jules Barbey d’Aurevilly, dating from 1867 and published in the collection Les Diaboliques in 1874 (English translation: Dedalus, 1986).

16. Luchino Visconti, Morte a Venezia (Death in Venice) (1971), after Thomas Mann, Der Tod in Venedig (1913), and Doktor Faustus (Doctor Faustus) (1949); the rewriting chain extends further to Louis-René des Forêts’ Le Malheur au Lido (Paris: Fata morgana, 1985–1987), which rewrites Thomas Mann’s original story and relies on the use of Mahler and other music in the rewriting project represented by Visconti’s film. I have developed this analysis and that of Duras’ Aurélia Steiner in Ecraniques (chapter VII: ‘Sur la réécriture’) (Lille: Presses Universitaires de Lille, 1990), pp. 161–223.

I would emphasise here to what point literature can become the active element in this short-circuit of practices: not only because the construction of the literary work itself pertains to montage, as Eisenstein maintains, but also because the linguistic dimension intervenes, with talking cinema, at the heart of a soundtrack in which the verbal takes its place between the sonic attraction and relation to the visual. Supplementary material for a plural montage, language makes sense only through a sensory multiplicity by means of which it is consoled, but also transported: aesthetic distortion, the disrupting power of which Resnais experimented with when he appealed to the literary writing of Duras or David Mercer in order to shatter the unity of the verbal even while drawing on its incantatory force. (13) The projection of the voiceover, the disparity between the spoken and the seen, the concerted desynchronisation of sonic and visual signs do not just shatter the language of the text, extorted from the writer in order to be twisted by a newly filmic writing. It is the propriety of both genres that is undermined in the impossible partition woven between them by montage. Is it literary cinema, a bastard rejected by the purists of a cinephilia of the image? Or is it rather literature, subject to a cinematographic erosion that characterises cinema from its raw beginnings? As Adorno emphasises in the essay already mentioned, cinema, however pure it might be, participates in the general unravelling of the arts, where the idea of art lays down its negative principle: the absorption of the aura in technical processes does not preclude the flourishing of signifying procedures, which return cinema to an artistic status it had wished to elude but has helped instead to expand. Such a contradiction is, according to Adorno, ‘the vital element in all properly modern art’; but it is in cinema that he finds it revealed, in the ‘vehicles of meaning’ it carries with it, despite itself and despite its ‘immanent legislation’. (14)

Linguistic proximity and duplicitous deterritorialisation go together in the cinematic enterprise. Hence, no doubt, the abundance but also the double game of these various ‘adaptations’ of literature that base filmic invention on a literary remake. The term rewriting, substituted for that of translation, marks here the stakes of a theoretical shift in the understanding of this transition. On the one hand, a repetition is involved, all the more paradoxically so since the second work must become other – a film – while remaining the same – a text with a shared title: Astruc played on this paradox in Le Rideau cramoisi (The Crimson Curtain, 1952), when he allows the entire story of Barbey d’Aurevilly to be spoken in voiceover, adding a faithful illustration, but by mute figures. (15) On the other hand, putting this differential pair into play modifies the identity of the original, of which it proposes an analogic version capable of illuminating, albeit by altering, the principle of writing: thus, in Death in Venice (1971), Visconti takes up Thomas Mann’s story only to intersperse it with other works and art forms, including those of Mann himself, but also of Mahler and Schoenberg, whose musical appeal suffused the Doctor Faustus that followed. (16) Literal or dislocated, the remake avoids echo – the film differs, sometimes radically – by flaunting its borrowing – the film resembles to the point of being identical. It will not therefore be possible to compare two distinct works, corresponding to two different art forms, to measure the exact degree of equivalence, of contamination or of irreparable incompatibility. Taking separation as the principle of the remade work, it will be a question of maintaining both constituents simultaneously, entering this difference to the account of a work in two parts, the development of which is determined by its aptitude to alteration: attraction of the fantasmal in Barbey’s figures, the serial inversion attributable to Platonic reference that guarantees Mann’s work. In this sense, cinematic rewriting would have as its role to precipitate but also to reflect that which, in the work of language, already belongs to the unworking of language and of the work.

17. Duras, India Song (Paris: Gallimard, 1973; English translation: Grove Press, 1994).

Filmic Detour/Textual Return: The Becoming-Heterogeneous of Language in the Filmic Rewriting of the Various Aurélia Steiner

Duras’ experimentation obeys the principle we have been discussing and follows its consequences to the extreme limit. The entire work clearly rests on the principle of original translation, according to which the written text cannot take place without renouncing any identity of its own; and the famed ‘texte théâtre film’ which forms the subtitle of the libretto to India Song (17) designates less the polyvalence of the written than the necessity it obeys, of not being a written piece: passing, therefore, from art to art, undoing the propriety of each, starting with that of writing. In this system of destruction, cinema occupies a special place, precisely because of its capacity for differential repetition. From the moment where the voices, offscreen and autonomous, invade the soundtrack, cinematic rewriting takes up the task of undoing the works – those of the Indian cycle – by repeating them to the point of reworking them, both literally and otherwise. But a further turn is taken with the three Aurélia Steiner, written for film, realised as such in two cases, and published without any direct reference to those productions: no scene descriptions, no mention of visual or audio directions, unlike the previous shooting scripts. These are double versions, then, cinematographic and literary, the strangeness of which derives from the fact that in both cases, film and book, the text is the same and the book is written in the absent memory of the film for which it was written and of which it bears the trace by default.

18. Duras, Aurélia Steiner (three texts), pp. 115–200 in the collection of film-texts published by Mercure de France (Paris: 1979), Le navire Night,… . Aurelia-Melbourne can be found on pp. 115–35. The corresponding film (35 min., colour) was made in 1979 as was Aurelia-Vancouver (48 min., black and white), but they are not included in the ‘film section’ which concludes the collection. Aurelia-Paris was not produced as a film.

19. Duras, Aurélia Steiner, op. cit., pp. 135, 166 and 200.

Indeed, doubling the written, the Melbourne film recites in its entirety the text of the book: entrusted to a monotonous voice off-screen that reels it off dully, the letter (‘I am writing to you [...]’) occupies the soundtrack, which envelopes it in noise and silence. Become speech, trapped in an invasive space of sound, nonetheless it is not placed in an enunciative role: the voice of the ‘I’, where we recognise Duras’ own voice, has become a reading voice, all the more determining for being that of the author herself, becoming reader of her own text. Yet this insistent instruction – to read and not say, to make understood and not enable understanding – finds a striking amplification in the visual track: nothing of what is pronounced by the voice appears onscreen. Although the letter that is read out speaks of love and suffering, addresses a far-off lover, evokes a garden, roses, a skinny cat, the rising fog, the death of the cat and of many others, all that is seen and heard in the film is the descent of a barge on the Seine in Paris, its advance along the canal banks, the slow passage under bridges and the resumption of the journey in fading light. No story is illustrated, no place is converged upon, neither addresser nor addressee appears; the film reading of the text works toward the erasure of those representations that could be evoked in a reading of the text itself.

Positioned thus between voice and image, the spectator believes herself to be reading something like a tale of passion that is disavowed by its visual anchoring. This replaces narrative illustration with sensory propositions dominated by light, colour, shadows and reflections, variations of tone in a liquid element that renders shapes opaque or fleeting despite the lines of bridges and quays. There is a fundamental discord between the materiality of the visual aspect of evanescence and the murmuring, rustling sounds that occupy the filmic space on the one hand and, on the other, the continued intelligibility of the text that finds itself trapped there. Should one trust sensory perceptions, navigate among non-signifying materials, or align oneself with the meaning uncovered by the voiceover? That alternative is refused us, inasmuch as it is the filmic supplement that simultaneously repeats the text, which remains readable, and rewrites it by disfiguring it, that is to say preventing it from bodying forth figures.

20. Duras, Aurélia Steiner, in Le Navire Night, op. cit., p. 132.

21. Ibid., p. 124.

22. Ibid., p. 121.

In reading the text, the film surrenders it to reading. In doubling it, through the eruption of the visual and the audible into the verbal, it reflects a multiple splitting to which language is bound, caught between constraint and rejection of meaning. To rewrite is here to read the writing that signs the text (‘I write’) and that instils in the language as in speech the submerged process of semantic defeat, in which writing unsays the spoken. Filmic distortion thus becomes a property of language: if the text becomes other, it is solely according to the logic of a language divided from itself by the writing in which it takes shape: not two languages, as Valéry and Mallarmé suggest, but a single language turned back on itself. The film will therefore only have been a deflection. And the aesthetic requirement it brings into play – to listen and see in order to recognise sensory materiality and perceptive mobility – designates only so many operations that can be assigned to the materiality of a text that does not feature, that does not make itself heard or felt, and to which the film identifies itself as the only authorised sensory approach: tracing, in a palimpsest, the critical relation of writing to language and to the representation for which it is the vehicle. Such is the paradox of a programmed destruction which has recourse to sight and hearing in order to reject from the undone work all possibility of visible manifestation or harmony in sound.

Not a graft then, nor a hybridisation of text and film. Only characters to read, as is borne out by a publication without any identifiable cinematic trace. These, then, are the ‘immanent motives’ of literature, in Adorno’s terms, that are reflected at the heart of a text restored to its proper strangeness by its passage through an alien medium. If the partition of the arts is resolved in their encounter, it will only be that the dividing point has been displaced: each work of art, in itself, depends on an internal partition, not between two arts nor even between two forms of language, but within the work itself, an internal declaration of its realisation as art. But a difficulty appears here, unleashing a new paradox: if no trace of the film remains in the text, how might a consciously obstructed filmic reading be nevertheless maintained? How do we read a text while remembering a film that has become invisible? All that is left to attest to its presence are the blank spaces that separate the paragraphs, fragmenting the textual mass and suspending its movement. These passing gaps, these momentary spaces might then be the only perceptible signs of a process of writing that the filmic detour only unveiled in its linguistic duplicity in order to transfer to language alone the responsibility of its own disavowal. A mere operator, the film itself can be destroyed. The spatial organisation of the text becomes the index of an ultimate destruction, which removes the film from the text having relied on the film for a destruction for which the text alone is answerable.

23. Maurice Blanchot, ‘Détruire’, L’Amitié (Paris: Gallimard, 1971), pp. 132–6. English translation: Friendship (Stanford University Press, 1997).

24. Blanchot, L’Entretien infini (Paris: Gallimard, 1969), p. 459 (‘Oublieuse mémoire’) and pp. 630–1 (on ‘L’Absence de livre’). English translation: The Infinite Conversation (University of Minnesota Press, 1992). With the exception of L’Amitié/Friendship, these two passages from L’Entretien infini do not relate to Duras in particular.

To interrupt, to separate, to lead elsewhere, to divert, to recommence – so many acts made possible by movement into blank space. But they persist only as possibilities, their particular interest being however that they dispel any misunderstanding as to what there is of art in writing: there is nothing there to be seen or heard that might have escaped us at first reading: it is the least that should be seen, the scarce reality of an imaginary vision, that film would certainly have blocked, but at the risk of letting its other side truly emerge. The cuts, the fractures, the suspensions will be the only aesthetic operations that the erasure of the film permits the book. ‘Est-ce un "livre", un "film"? "l’intervalle des deux?"’ [‘is it a book, a film, the interval between the two?’], asks Blanchot, thinking of Détruire dit-elle. (23) But with the Aurélia Steiner, the ‘intervalle’ opens within the book itself and the forgotten memory of the film pleads for an ‘oublieuse mémoire’ [forgetful memory] attributable to literature when it is left to its own devices, which is also to say, in Blanchot’s words once more, left to the double writing that initiates it: the blank, invisible and violent, and the black, made of articulated signs and sense. (24) Once more it is necessary to make perceptible that white writing, a colourless flame elusive by definition; to which end Duras applies herself when she speculates on the dissimulation of the film at the heart of the text: that is, a distortion reduced to a pure discontinuity.

Such is the purpose of filmic deflection in the rewriting of literary works: to double the work, by superfluity or plundering, thus weakening the one split by means of which it is made. In this sense, it is the letter of the text, following Oliveira’s title [The Letter, 1999], that is the film’s aim, even though it diverts it from its end, the singularity of which is that it can only be attained within the text and at the cost of its own defeat. The unravelling of the arts, to which the relation between literature and cinema constructed here is vowed, has then a negative function: to declare the attraction of one to the other that leads to the rejection of one, and in this capacity to give over to the ‘concept of art’ the negativity of its ‘content’, which it does not cease to chart, but outside of which it could not take place.

The aesthetic attraction of literature rests on this singular paradox, which remains valid no doubt for the encounter between other arts: an expansive interaction of forms, but which leads to an intensive deprivation of effects. What is not formulated explicitly in this contrary dialectic, which is found within Adorno’s proposition, is indicated by the deflection into filmic rewriting. The immanent motivations of an art, on the subject of which Adorno hesitates, cannot appear except through an inverse test: the attraction of another art, that menaces its integrity – hence the risk of hybridisation – but which acts too as a reflexive mirror – hence, internal rejection and retreat. The work of art is inscribed within this paradoxical and inevitably insoluble process: for art has no essence other than this becoming, which makes it heterogeneous to itself. Duras’s filmic rewriting, as well as many other possible experiments, furnishes a particularly illuminating model, because it is always in expansion, resorting to image in order to abolish representation, resorting to the order of sound in order to disavow in turn image and music, theatre and text. But each time it is a matter of writing: of making a work, therefore – literary, in this case – by dint of an unworking that belongs to the double constraint of the principle of art, the negativity of which, however, depends on the work, on the making of the work, in order that it might appear. Blanchot will not suffice in this regard without Adorno, and vice versa.