‘Too many words’: the diagnosis is repeated whenever the crisis of art, or its subservience to aesthetic discourse, is denounced. Too many words about painting; too many words that comment on its practice and devour it; that clothe and transfigure the ‘anything goes’ it has become or replace it in books, catalogues and official reports – to the point of spreading to the very surfaces where it is exhibited and where, in its stead, we find written the pure affirmation of its concept, the self-denunciation of its imposture, or the registration of its end.
I do not intend to respond to these claims on their own ground. Instead, I would like to ponder the configuration of this ground and the way in which the particulars of the problem are set out in it. From there I would like to turn the game upside down and move from polemical denunciation of the words that encumber painting to a theoretical understanding of the articulation between words and visual forms that defines a regime of art.
At first sight, things seem clear: on the one hand, there are practices; on the other hand, their interpretations; on the one hand, there is the pictorial phenomenon and on the other the torrent of discourse about it which philosophers, writers or artists themselves have poured out, since Hegel and Schelling made painting a form of manifestation of a concept of art that was itself identified with a form of unfolding of the absolute. But this simple opposition starts to blur when one poses the question: what precisely does this ‘pictorial phenomenon’ set against the supplement of discourse consist in?
The commonest reply takes the form of a seemingly irrefutable tautology. The peculiarity of the pictorial phenomenon is that it only uses means specific to painting: coloured pigments and flat, two-dimensional space. The simplicity of this answer accounts for its success, from Maurice Denis to Clement Greenberg. Yet it leaves many ambiguities unresolved. We can all agree that the peculiarity of an activity consists in using the means specific to it. But a means is specific to an activity in as much as it is apt to achieve the purpose that is specific to the activity. The particular purpose of a mason’s labour is not defined by the material he works on and the tools he uses. What, then, is the specific purpose that is realized by putting coloured pigments on a flat surface? The response to this question is in fact an intensification of the tautology: the specific purpose of painting is solely to put coloured pigments on a flat surface, rather than to people it with representative figures, referred to external entities situated in a three-dimensional space. ‘The Impressionists’, Clement Greenberg thus says, ‘abjured underpainting and glazing, to leave the eye under no doubt as to the fact that the colours used were made of real paint that came from pots or tubes.’1 Let us accept that such was indeed the intention of the Impressionists (which is doubtful). There are nevertheless many ways of demonstrating that you are using tubes of real paint: you can mention the fact on the canvas or alongside the canvas; you can stick these tubes to it, or replace the canvas by a small window containing them, or arrange big pots of acrylic paint in the middle of an empty room, or organize a happening where the painter takes a dip in the paint. All these methods, which are empirically attested to, indicate that the artist is using a ‘real’ material, but they do so at the expense of the flat surface where the demonstration of painting ‘itself’ was to be staged. They do it while uncoupling the two terms whose substantial unity it was supposed to demonstrate: the material – pigmental or other – and the two-dimensional surface. They therewith pose the question: why must painters ‘leave the eye under no doubt’ that they are using real paint from tubes? Why must the theoretician of ‘pure’ painting show us that the Impressionist use of pure colours has this as its purpose?
The reason is that this definition of the pictorial phenomenon is in fact an articulation of two contradictory operations. It wants to guarantee the identity of the pictorial material and the painting-form. The art of painting is the specific realization of nothing but the possibilities contained in the very materiality of coloured matter and its support. But this realization must take the form of a self-demonstration. The same surface must perform a dual task: it must only be itself and it must be the demonstration of the fact that it is only itself. The concept of medium guarantees this secret identity of opposites. ‘To use nothing but the medium specific to an art’ means two things. On the one hand, it means carrying out a purely technical operation: the gesture of squeezing some pictorial material on to an appropriate surface. It remains to be known what the ‘peculiarity’ of this appropriation is and what makes it permissible, as a result, to refer to this operation as pictorial art. For that the word ‘medium’ must refer to something quite different from a material and a support. It must designate the ideal space of their appropriation. The notion must therefore be discreetly split in two. On the one hand, the medium is the set of material means available for a technical activity. ‘Conquering’ the medium then signifies: confining oneself to the use of these material means. On the other hand, the stress is placed on the very relationship between end and means. Conquering the medium then signifies the converse: appropriating the means to make it an end in itself, denying the relationship of means to end that is the very essence of technique. The essence of painting – simply casting coloured matter on a flat surface – is to suspend the appropriation of means to an end that is the essence of technique.
The idea of the specificity of pictorial technique is consistent only at the price of its assimilation to something quite different: the idea of the autonomy of art, of the exception of art from technical rationality. If it is necessary to show that you are using tubes of colour – and not simply to use them – it is in order to demonstrate two things: firstly, that this use of tubes of colour is nothing but the use of tubes of colour, only technique; and secondly, that it is something quite different from the use of tubes of colour, that it is art – i.e. anti-technique.
In fact, contrary to the claim of the thesis, it always has to be shown that the material displayed on some surface is art. There is no art without eyes that see it as art. Contrary to the healthy doctrine which would have it that a concept is the generalization of the properties common to a set of practices or objects, it is strictly impossible to present a concept of art which defines the properties common to painting, music, dance, cinema, or sculpture. The concept of art is not the presentation of a property shared by a set of practices – not even that of one of those ‘family resemblances’ which Wittgenstein’s followers call upon in the last resort. It is the concept of a disjunction – and of a historically determinate unstable disjunction – between the arts, understood in the sense of practices, ways of making. Art as we call it has existed for barely two centuries. It was not born thanks to the discovery of the principle shared by the different arts – in the absence of which tours de force superior to those of Clement Greenberg would be required to make its emergence coincide with the conquest by each art of its specific ‘medium’. It was born in a long process of rupture with the system of beaux arts – that is, with a different regime of disjunction in the arts.
That different regime was encapsulated in the concept of mimesis. Those who regard mimesis as simply the imperative of resemblance can construct a straightforward idea of artistic ‘modernity’ as the emancipation of the peculiarity of art from the constraint of imitation: the reign of coloured beaches in the place of naked women and war horses. This is to miss the main thing: mimesis is not resemblance but a certain regime of resemblance. Mimesis is not an external constraint that weighed on the arts and imprisoned them in resemblance. It is the fold in the order of ways of making and social occupations that rendered them visible and thinkable, the disjunction that made them exist as such. This disjunction is twofold. On the one hand, it separated the ‘beaux arts’ from the other arts – simple ‘techniques’ – in accordance with their specific purpose: imitation. But it also shielded the imitations of the arts from the religious, ethical or social criteria that normally governed legitimate uses of resemblance. Mimesis is not resemblance understood as the relationship between a copy and a model. It is a way of making resemblances function within a set of relations between ways of making, modes of speech, forms of visibility, and protocols of intelligibility.
That is why Diderot can criticize Greuze on the paradoxical grounds that he has darkened the skin of his Septimus Severus and represented Caracalla as an absolute rogue.2 Septimus Severus was the first Roman emperor of African origin and his son Caracalla was indeed a mischievous Frank. The painting by Greuze under criticism represents the latter at the point when he is convicted of attempted parricide. But the resemblances of representation are not reproductions of reality. An emperor is an emperor before he is an African; and the son of an emperor is a prince before being a rascal. To darken the face of the one, and accuse the other of baseness, is to transform the noble genre of the history painting into the common genre of painting that is appropriately called genre painting. The correspondence between the order of the painting and the order of history is the affinity between two orders of grandeur. It inscribes the practice of art, and the images it offers for our inspection, in a general order of relations between making, seeing and saying.
There is such a thing as art in general by virtue of a regime of identification – of disjunction – that gives visibility and signification to practices of arranging words, displaying colours, modelling the volume or evolution of bodies; which decides, for example, what a painting is, what one does by painting, and what one sees on a painted wall or canvas. But such a decision always involves the establishment of a regime of equivalence between a practice and what it is not. To know whether music and dance were arts, Batteux asked whether they were imitations; whether, like poetry, they recounted stories, organized actions. The ut pictura poesis/ut poesis pictura did not simply define the subordination of one art – painting – to another – poetry. It defined a relationship between the orders of making, seeing, and saying whereby these arts – and possibly others – were arts. The issue of flatness in painting, imitation of the third dimension, and rejection of that imitation is not in any way a question of discriminating between the peculiarity of pictorial art and the peculiarity of sculptural art. Perspective was not adopted in order to demonstrate the painter’s ability to imitate the depth of space and the contours of bodies. Painting would not have become one of the ‘beaux arts’ merely by offering proof of this technical capacity. A painter’s virtuosity has never sufficed to open the doors of artistic visibility for him. If perspective was linear and theatrical before becoming aerial and sculptural, it is because painting first of all had to demonstrate its capacity for poetry – its ability to tell stories, to represent speaking, acting bodies. The bond between painting and the third dimension is a bond between painting and the poetic power of words and fables. What can undo this bond, assign painting a privileged relationship not only to the use of flat surface but also to the affirmation of flatness, is a different type of relationship between what painting does and what words make visible on its surface.
For painting to be destined for flatness, it must be made to be seen as flat. For it to be seen as flat, the links that connect its images in the hierarchies of representation have to be loosened. It is not necessary that painting should no longer ‘resemble’. It is sufficient for its resemblances to be uncoupled from the system of relations that subordinate the resemblance of images to the ordering of actions, the visibility of painting to the quasi-visibility of the words of poems, and the poem itself to a hierarchy of subjects and actions. The destruction of the mimetic order does not mean that since the nineteenth century the arts have done ‘anything they like’; or that they have freely embarked on the conquest of the possibilities of their particular medium. A medium is not a ‘proper’ means or material. It is a surface of conversion: a surface of equivalence between the different arts’ ways of making; a conceptual space of articulation between these ways of making and forms of visibility and intelligibility determining the way in which they can be viewed and conceived. The destruction of the representative regime does not define some finally discovered essence of art as such in itself. It defines an aesthetic regime in the arts that is a different articulation between practices, forms of visibility and modes of intelligibility.
What inducted painting into this regime was not the rejection of representation, not a revolution in the practice of painters. It was primarily a different way of seeing the painting of the past. The destruction of the representative regime in painting started at the beginning of the nineteenth century, with the revocation of the hierarchy of genres, with the rehabilitation of ‘genre painting’ – the representation of ordinary people engaged in ordinary activities, which used to be contrasted with the dignity of history painting as comedy to tragedy. It therefore began with the revocation of the subordination of pictorial forms to poetic hierarchies, of a certain bond between the art of words and that of forms. But this liberation was not the separation of painting from words, but a different way of conjoining them. The power of words is no longer the model that pictorial representation must take as its norm. It is the power that hollows out the representative surface to make the manifestation of pictorial expressiveness appear on it. This means that the latter is only present on the surface to the extent that a gaze penetrates it; that words amend the surface by causing another subject to appear under the representative subject.
This is what Hegel paradigmatically does in his endeavour to rehabilitate Dutch painting – pioneer of the labour of re-description which throughout the Romantic era, when faced with the works of Gerard Dou, Teniers or Adrian Brouwer, like those of Rubens and Rembrandt, developed the new visibility of a ‘flat’ painting, an ‘autonomous’ painting. The true subject of these despised paintings, explains Hegel, is not what we see at first. It is not tavern scenes, episodes of bourgeois existence, or domestic accessories. It is the autonomization of these elements, the severing of the ‘threads of representation’ that attached them to the reproduction of a repetitive style of existence. It is the replacement of these objects by the light of their appearance. What occurs on the canvas is now an epiphany of the visible, an autonomy of pictorial presence. But this autonomy is not the installation of painting in the solitude of its own peculiar technique. It is itself the expression of a different autonomy – the autonomy the Dutch people succeeded in wresting in their triple struggle against a hostile nature, the Spanish monarchy, and Papal authority.3
For painting to attain flatness, the surface of the painting has to be divided in two; a second subject has to be shown under the first. Greenberg counter-poses to the naivety of Kandinsky’s anti-representative programme the idea that the important thing is not the abandonment of figuration, but the conquest of surface. But this conquest is itself a work of de-figuration: a labour that renders the same painting visible in a different way, that converts figures of representation into tropes of expression. What Deleuze calls the logic of sensation is much more a theatre of de-figuration, where figures are wrenched from the space of representation and reconfigured in a different space. Proust calls this de-figuration denomination, when characterizing the art of pure sensation in Elstir: ‘if God the Father had created things by naming them, it was by taking away their names or giving them other names that Elstir created them anew.’4
The surface claimed as the specific medium of pure painting is in fact a different medium. It is the theatre of a de-figuration/denomination. Greenberg’s formalism, which would reduce medium to material, and Kandinsky’s spiritualism, which makes it a spiritual milieu, are two ways of interpreting this de-figuration. Painting is flat in as much as words change their function with respect to it. In the representative order, they served as its model or norm. As poems, as sacred or profane history, they outlined the arrangement that the composition of the painting had to translate. Thus, Jonathan Richardson recommended to painters that they first of all write the story of the painting in order to know whether it was worth painting. As critical discourse, words compared what was painted with what should have been: the same story conveyed in a more appropriate attitude and physiognomy or a story more worthy of being painted. It is often said that aesthetic criticism, the criticism which emerged in the Romantic era, no longer proceeds normatively; no longer compares the painting with what it should have been. But the opposition between the norm and its absence, or between the external norm and the internal norm, conceals the main thing: the contrast between two modes of identification. In the aesthetic age, the critical text no longer says what the painting should be or should have been. It says what it is or what the painter has done. But to say that is to arrange the relationship between the sayable and the visible, the relationship between the painting and what it is not, differently. It is to reformulate the like of the ut pictura poesis, the like whereby art is visible, whereby its practice is attuned to a way of seeing and falls within a way of thinking. It has not disappeared. It has changed places and functions. It works towards the de-figuration, the alteration of what is visible on its surface, and hence towards its visibility as art.
To see something as art, be it a Deposition from the Cross or a White Square on White Background, means seeing two things in it at once. Seeing two things at once is not a matter of trompe-l’oeil or special effects. It is a question of the relations between the surface of exhibition of forms and the surface of inscription of words. But this new bond between signs and forms that is called criticism, and which is born at the same time as the proclamation of the autonomy of art, does not work in the simple form of retrospective discourse adding meaning to the nakedness of forms. It works in the first instance towards the construction of a new visibility. A new form of painting is one that offers itself to eyes trained to see differently, trained to see the pictorial appear on the representative surface, under representation. The phenomenological tradition and Deleuzian philosophy readily assign art the task of creating presence under representation. But presence is not the nakedness of the pictorial thing as opposed to the significations of representation. Presence and representation are two regimes of the plaiting of words and forms. The regime of visibility of the ‘immediacies’ of presence is still configured through the mediation of words.
This labour is what I would like to show at work in two texts of nineteenth-century criticism – texts that reconfigure the visibility of what painting does. The first, by placing a representative painting of the past in the new regime of presence, constitutes the new mode of visibility of the pictorial, appropriate for the contemporary painting that it nevertheless disdains. The second, in celebrating a new form of painting, projects it into an ‘abstract’ future of painting that does not as yet exist.
I take my first example from the monograph on Chardin published by the Goncourt brothers in 1864:
Against one of those muffled, mixed backgrounds that he knows so well how to rub in, and amid which a coolness, as in a grotto, mingles vaguely with the cast shadows of a sideboard, on one of those coloured tables like moss and topped with dull marble, which so often bear his signature, Chardin pours out dessert plates: here is the shaggy velvet of the peach, the amber transparency of the white grape, the sugary rime of the plum, the moist crimson of strawberries, the hardy berry of the muscatel and its smoke-blue mist, the wrinkles and tubercules of an orange skin, the embroidered guipure of melons, the copperas of old apples, the knots in a crust of bread, the smooth rind of the chestnut and the wood of the hazel nut … In one corner there is apparently nothing more than a mud-coloured texture, the marks of a dry brush, then, suddenly a walnut appears curling up in its shell, showing its sinews, revealing itself with all the details of its form and colour.5
The whole of this text is governed by one aim: to transform figurative particulars into events of pictorial matter, which themselves convey metamorphic states of matter. The operation might be conveniently summarized starting from the last lines: the curling up of the walnut, the appearance of the figure in the mud-coloured texture and the marks of a dry brush. The ‘matterism’ of the Goncourts’ description prefigures a major form of visibility of pictorial ‘autonomy’: the working of the material, of the coloured paste asserting its sway over the space of the painting. It configures in Chardin’s painting a whole future of Impressionism and Abstract Expressionism or action-painting. It also prefigures in it a whole future of descriptions and theorizations: conceptualization of the formless à la Bataille, of original mimesis à la Merleau-Ponty, or of the Deleuzian diagram – the operation of a hand that cancels one visibility in order to produce another: a ‘tactile’ visibility, the visibility of the gesture of the painter substituted for that of its result. The domestic still life does not possess any privilege in this regard. The description of Rubens’s great religious paintings observes the same principle: ‘Never has a paintbrush so furiously rolled and unrolled lumps of flesh, tied and untied clusters of bodies …’.
This transformation of the visible into the tactile and of the figurative into the figural is only possible through the highly specific labour of the writer’s words. It first of all consists in the deictic mode of the utterance, a mode of presence indicated through the operation of a literalization which shows us Chardin ‘pouring out’ the plates – that is, transforming the representation of the table into a gesture of projection which renders the acts of spreading colour and laying the table equivalent. Next it consists in the whirlwind of adjectives and metaphors that succeed in articulating two contradictory operations. They transform the qualities of the fruits represented into substantial states of matter. The amber, the rime, the mist, the wood, or the moss of some living matter take the place of the grape, the plums, the hazel nuts, and the table of the represented still life. At the same time, however, they systematically blur the identities of objects and the boundaries between realms. Thus, the guipure of the melon, the wrinkles of the orange, or the copperas of the apple endow the plants with the features of human faces or works, whereas moss, coolness, and film transform solid elements into liquid ones. Both operations conduce to the same result. Linguistic tropes change the status of the pictorial elements. They transform representations of fruits into tropes of matter.
This transformation is much more than an aesthete’s re-reading. The Goncourts simultaneously register and configure a new visibility of the pictorial phenomenon, an aesthetic-type visibility where a relationship of coalescence between the density of the pictorial matter and the materiality of the painter’s gesture is imposed in place of the representative privilege of the form that organized and cancelled matter. They elaborate the new regime of visibility that makes a new pictorial practice possible. To do this, they do not need to appreciate the new painting. It has often been remarked that, with reference to Chardin, Rubens or Watteau, the Goncourts elaborate the visibility of Impressionist canvases. But no law of necessary similarity obliges them to accommodate on the canvases of innovators the viewing machine thus constructed. For them pictorial novelty is already realized, already present in the present woven by the interweaving of their linguistic tropes with Chardin’s brush-strokes and figures. When innovators want to make the physical play of light and the hachure of colour directly equivalent, they short-circuit the labour of metaphor. One might say, in Deleuzian terms, that they make diagrams which remain diagrams. But if Deleuze helps us understand why Edmond de Goncourt cannot see the paintings he has rendered visible, the converse is also perhaps true: Goncourt helps us understand what Deleuze, in order to preserve the idea of painting as a labour of sensation on sensation, tries not to see – the pictorial diagram only makes visible if its labour is rendered equivalent to that of metaphor, if words construct such equivalence.
To construct that equivalence is to create solidarity between a practice and a form of visibility. But this solidarity is not a necessary contemporaneity. On the contrary, it is asserted through an interplay of temporal distances that remove pictorial presence from any epiphany of the present. The Goncourts see Impressionism already realized in Chardin. They see it because they have produced its visibility through a labour of de-figuration. De-figuration sees novelty in the past. But it constitutes the discursive space that renders novelty visible, which constructs a gaze for it in the very discrepancy of temporalities. Accordingly, the discrepancy is as much prospective as retrospective. It not only sees novelty in the past. It can also see as yet unrealized possibilities of painting in the present work.
This is what is revealed by another critical text – the one Albert Aurier devoted in 1890 to Gaugin’s Vision du sermon (also known as La Lutte de Jacob avec l’ange). This text is a manifesto for a new kind of painting, a painting that no longer represents reality but translates ideas into symbols. Yet this manifesto does not proceed by a polemical argument. It too proceeds by a de-figurative description, which uses certain devices from the mystery story. It plays on the discrepancy between what is seen and what is not seen in order to establish a new status of the visible in painting:
Far off, very far off, on a mythical hill whose soil is a rutilant vermilion in appearance, we have the biblical struggle between Jacob and the Angel.
While these two legendary giants, transformed into pygmies by the distance, fight their fearsome fight, some women are watching. Concerned and naïve, they doubtless understand little of what is going on over there, on that fabulous crimson hill. They are peasants. And from the size of their headdresses spread like seagulls’ wings, the typical mixed colours of their scarves, the form of their dresses or blouses, we can tell that they come from Brittany. They have the respectful attitudes and wide-eyed expressions of simple creatures listening to rather supernatural, extraordinary tales from a mouth that is above reproach and revered. So silently attentive are they, so contemplative, bowed and devout is their bearing, that one would say they were in a church. One would say they were in a church and that a vague scent of incense and prayer was wafting about the white wings of their headdresses and that the respected voice of an old priest was hovering about their heads … Yes, no doubt, in a church, in some poor church in some poor, small Breton village … But if so, where are the mouldy green pillars? Where are the milky-white walls with the tiny monolithographic Stations of the Cross? Where is the wooden pulpit? Where is the old parish priest who is preaching? … And why over there, far off, very far off, is that fabulous hill, whose soil seems to be of a rutilant vermilion, looming up? …
Ah! It is because the mouldy green pillars, and the milky-white walls, and the little chromolithographic Stations of the Cross, and the wooden pulpit, and the old parish priest who is preaching have long since been wiped out, no longer exist for the eyes and souls of the good Breton peasants! … What wonderfully touching accents, what luminous hypotyposis, strangely appropriate to the crude ears of his oafish audience, has this mumbling village Bossuet encountered? All surrounding materialities have vanished in smoke, have disappeared. The one calling out has himself faded away and it is now his voice, his poor old pitiful, mumbling Voice that has become visible, imperiously visible. And it is his Voice that is contemplated with naïve, rapt attention by these peasants with white headdresses; and it is his Voice this rustically fantastic vision, looming up over there, very far off; his Voice this legendary hill whose soil is vermilion-coloured, this childish dreamland, where the two biblical giants, transformed into pygmies by the distance, are fighting their hard, fearsome fight!6
This description is constructed through a mechanism of mystery-making and substitutions, placing three paintings in one. There is a first painting: some peasant women in a meadow who are watching the fighters in the distance. But this appearance condemns itself as incoherent and calls for a second painting: to be dressed thus and have these attitudes, the peasants should not be in a meadow; they should be in a church. Therewith Aurier evokes what a painting of this church would normally be: a genre painting with miserable décor and grotesque characters. But this second painting, which would impart a certain context to the contemplative bodies of the peasants – the context of a realist, regional painting of social customs – is not there. The painting we do see precisely constitutes its refutation. In and through this refutation, we must therefore see a third painting – that is, see Gaugin’s painting from a new angle. The spectacle it presents us with has no real location. It is purely ideal. The peasants do not witness any realistic scene of sermon and struggle. They – and we – see the Voice of the preacher: the words of the Word that pass via this voice. These words tell of the legendary fight of Jacob with the Angel, of terrestrial materiality with celestial ideality.
Thus, the description is a substitution. It replaces one scene of speech by another. It does away with the story with which the representative painting was in harmony; and it does away with the scene of speech to which the spatial depth was adjusted. It replaces them by a different ‘living word’: the words of Scripture. And the painting thus appears as the site of a conversion. What we see, Aurier tells us, is not some scene of peasant life, but an ideal pure surface where certain ideas are expressed by certain signs, making figurative forms the letters of an alphabet peculiar to painting. Description then makes way for a neo-Platonic discourse that shows us in Gaugin’s painting the novelty of an abstract art, where visible forms are merely signs of the invisible idea: an art breaking with the realist tradition and its latest novelty – Impressionism. Removing the genre painting that should have been there, Aurier replaces it by a correspondence between the ‘conceptual’ purity of the abstract painting and the beatific vision of the ‘naïve’ listener. For the representative relationship he substitutes the expressive relationship between the abstract ideality of form and the expression of a content of collective consciousness. This spiritualism of pure form is the counterpart of the ‘matterism’ of the pictorial gesture exemplified by the Goncourts. Unquestionably, the contrast runs throughout the nineteenth century: Raphael and an Italian purity of form against Rembrandt/Rubens and a Dutch epiphany of physical matter. Yet it merely repeats the old controversy over drawing and colour. The very controversy is caught up in the elaboration of a new visibility of painting. ‘Ideaism’ and ‘matterism’ contribute equally to forming the visibility of an ‘abstract’ painting – not necessarily painting without representation, but painting that oscillates between pure realization of the metamorphoses of matter and translation of the pure force of ‘internal necessity’ into lines and colours.
It might readily be objected to Aurier’s demonstration that what we see on the canvas is not signs but easily identifiable figurative forms. The peasants’ faces and poses are schematized. But this very schematism assimilates them not so much to the Platonic Idea as to the advertising images which still embellish Pont-Aven biscuits today. The scene of the struggle is at an uncertain distance, but the relation of vision to the semi-circle of peasants remains ordered in accordance with a consistent representative logic. And the cloisonné spaces of the painting remain connected by a visual logic that is consistent with the narrative logic. In order to assert a radical break between the old ‘materialist’ painting and a new conceptual painting, Aurier must go significantly beyond what we see on the canvas. Through thought he must liberate the coloured spots that are still coordinated by a narrative logic and transform schematized images into abstract schemas. In the space of visibility that his text constructs for it, Gaugin’s painting is already a painting of the sort Kandinsky would paint and justify: a surface where lines and colours have become expressive signs obeying the single constraint of ‘internal necessity’.
The objection simply boils down to confirming this: the ‘internal necessity’ of the abstract canvas is itself only constructed in the device whereby words work the painted surface so as to construct a different plane of intelligibility for it. This comes down to saying that the flat surface of the painting is something quite different from the self-evidence of the law of a medium that has finally been conquered. It is a surface of dissociation and de-figuration. Aurier’s text establishes in advance a peculiarity of painting, an ‘abstract’ painting. But it also revokes in advance any identification of this ‘peculiarity’ with the law of a surface or material. The dismissal of representative logic is not the straightforward assertion of the physical materiality of the painting, refusing any subservience to discourse. It is a new mode of the correspondence, the ‘like’ that linked painting to poetry, visual figures to the order of discourse. Words no longer prescribe, as story or doctrine, what images should be. They make themselves images so as to shift the figures of the painting, to construct this surface of conversion, this surface of forms-signs which is the real medium of painting – a medium that is not identified with the propriety of any support or any material. The form-signs that Aurier’s text makes visible on the surface of Gaugin’s painting are open to being refigured in various ways – in the pure flatness of the abstract ‘language of forms’, but also in all the combinations of the visual and the linguistic which will be presented by Cubist or Dadaist collages, the appropriations of Pop Art, the décollages of the new realists, or the plain writing of Conceptual Art. The ideal plane of the painting is a theatre of de-figuration, a space of conversion where the relationship between words and visual forms anticipates visual de-figurations still to come.
I have spoken of theatre. This is not a ‘mere metaphor’. The arrangement in a circle of peasant women with their backs to the viewer, and absorbed by a distant spectacle, obviously puts us in mind of the ingenious analyses of Michael Fried, inventing a pictorial modernity conceived as anti-theatre, as an inversion of the motion of actors towards the audience. The obvious paradox is that this anti-theatre itself comes directly from the theatre – very precisely from the naturalist theory of the ‘fourth wall’ invented by a contemporary of Gaugin and Aurier: the theory of a dramatic action that would pretend to be invisible, to be viewed by no audience, to be nothing but life in its pure similarity to itself. But what need would life in its pure similarity, life ‘not looked at’, not made into a spectacle, have of speaking? The ‘formalist’ dream of a kind of painting that turns its back on the spectator in order to close in on itself, in order to adhere to the surface that is peculiar to it, could well be nothing but the other side of the same identitarian dream. A pure painting, clearly separate from ‘spectacle’, would be required. But theatre is not primarily ‘spectacle’, is not the ‘interactive’ site calling upon the audience to finish the work denounced by Fried. Theatre is first and foremost the space of visibility of speech, the space of problematic translations of what is said into what is seen. Accordingly, it is quite true, it is the site of expression of the impurity of art, the ‘medium’ which clearly shows that there is no peculiarity of art or of any art; that forms do not proceed without the words that install them in visibility. The ‘theatrical’ arrangement of Gaugin’s peasant women establishes the ‘flatness’ of the painting only at the cost of making this surface an interface that transfers the images into the text and the text into the images. The surface is not wordless, is not without ‘interpretations’ that pictorialize it. In a way, this was already the lesson of Hegel and the meaning of the ‘end of art’. When the surface is no longer split in two, when it is nothing more than a site for the projection of pigments, Hegel taught, there is no longer any art. Today, this thesis is commonly interpreted in a nihilistic sense. In advance, Hegel supposedly condemned art for art’s sake to the fate of ‘anything goes’ or showed that now, rather than works of art, there are ‘only interpretations’. The thesis seems to me to require a different reading. It is indeed true that Hegel personally turned the page on art, put art on his page – that of the book which tells the past the mode of its presence. But that does not mean that he turned the page for us in advance. Instead, he alerted us to this: the present of art is always in the past and in the future. Its presence is always in two places at once. He tells us in sum that art is alive as long as it is outside itself, as long as it does something different from itself, as long as it moves on a stage of visibility which is always a stage of defiguration. What he discourages in advance is not art, but the dream of its purity. It is the modernity that claims to vouchsafe each art its autonomy and painting its peculiar surface. Here indeed is something to fuel resentment against philosophers who ‘talk too much’.