Incompletion, reversibility and fragmentary montage
Le multiple, il faut le faire…
Gilles Deleuze and Félix Guattari, A Thousand Plateaus
In 1798 Novalis famously wrote: ‘Poetry is the authentic absolute real. This is the core of my philosophy. The more poetic, the more true.’  This aphorism expresses what might be called the artistic destiny of Western metaphysics, or the poetic presentation of absolute truth. Other writings from the same period make this point clear. Friedrich Schlegel, for example, observes in one of his Critical Fragments: ‘The whole history of modern poetry is a running commentary on the folloeing brief philosophical text: all art should become science and all science art; poetry and philosophy should be made one.’  This union of poetry and philosophy appears as the poetic becoming of knowledge, ‘objectivized intellectual intuition’ in Schelling’s terms, or ‘transcendental intuition having become objective’, the overcoming of the Kantian prohibition of an archetypal form of understanding, such as Kant defined it in section 77 of the Critique of the Power of Judgment. Schelling, at the end of his System of Transcendental Idealism, writes that ‘art alone can make objective, in a universally valid way, what philosophy can only represent subjectively.’  In this view, poetry is the alpha and omega of knowledge: all science originates in poetry (mythic language) and in the end must return to that ‘universal ocean of poetry from which it arose’. And in fact, as Schelling says elsewhere, the philosophy of art is not a particular philosophy; it is philosophy itself in its demand for absolute truth. I am referring here to the philosophy of absolute idealism of Schelling and Hegel, rather than the critical and transcendental idealism of Fichte. Though in truth Romanticism is always oscillating between these two options, constantly moving back and forth from subjective idealism to the subjectivity of the absolute (Schelling’s notion of identity). Either way, Romantic aesthetics exhibits truth in the form of art, more specifically in the work of art.
I shall not dwell on this well-known link between early Romanticism and post-Kantian idealism. I simply want to emphasize the fact that this objectivized intellectual intuition of art took on many forms, particularly in German Romanticism, that together participate in what might be called an artistic ontology. The work expresses the absolute at the same time that it is itself absolute; that is to say, autonomous, autotelic (as K. Ph. Moritz would say) – a tautegorical symbol for Schelling, close in this respect to the Romantics, and not illustrative allegory (in the sense of the Ut pictura poesis). The Romantic work of art may be called absolute because it is a self-sufficient whole, a non-mimetic symbolic reality. Examples of Romantic ontology and artistic self-presentation abound: the brief total unity of Schlegel’s hedgehog fragment, ‘A fragment, like a miniature work of art has to be entirely isolated from the surrounding world and complete in itself like a hedgehog.’  Fragment 116, also from the Athenaeum, speaks of ‘Universal synthetic poetry’, which Walter Benjamin argued was a condensation of the absolute novelistic ideal of Romanticism (a synthesis of genres, forms, a synthesis of the author and his work, poetry and philosophy, art and life). And, furthermore, intuitions of the universe in landscape painting: certain of Caspar David Friedrich’s landscapes are simultaneously subjective and objective. Perspectives that open out onto the infinity of the natural world, they are at the same time mises en abîme of the observing subject, since the representation of the landscape here implies the spectator looking at it. Friedrich, it is well known, often places the observer in the foreground of his paintings. It is the observer who creates, if not the painting itself, as Marcel Duchamp would have it, at least the visible unity we call a landscape. ‘Landscape as such,’ writes A.W. Schlegel, ‘exists only in the eyes of the one who contemplates it.’  In addition to all these examples of the absolute work of art – literary, pictorial, novelistic or musical (the Wagnerian Gesamtkunstwerk is a direct descendant of synthetic Romantic art) – Novalis’s enigmatic fragment on the monological power of poetic language provides a dazzling synthesis of the main characteristics of Romantic ontology, particularly the intransitivity of poetic language, which, by affirming itself, independently of any message being transmitted, without any relation to an outside, expresses the whole (the world soul and the totality of its relations).  Only by creating a world does the artwork speak the world as universal totality.
It seems clear that an essential aspect of artistic modernity is founded on this importance of autotelic and absolute autonomy, which has led art to abandon mimetic constraints and to question its own essence. The metaphysics or Romantic theology of the work of art also made possible both a meditation on the spiritual in art – as Kandinsky would say, who, in addition to his theosophical sources, relied on the model of Romantic music – and a reflection on grasping the essence of the work of art. Among the descendents of the spiritual and ontological tradition of Romanticism, it is certain that the modernism of Barnett Newman, Mark Rothko, Clement Greenberg and Michael Fried borrowed (consciously or not) this conception of the auto-synthetic work: the idea of an absolute now. As Newman said, The sublime is now. According to Newman, the work of art is the image of a religious ‘revelation’, founded on ‘absolute emotions’ rather than on iconological referents taken from European history. Remember the ‘infinitely brief instant’ of fullness that Michael Fried mentions at the end of his famous 1967 essay ‘Art and Objecthood’.  This instant is emblematic of an art that exists outside of history, whole, nearly eternal, that Fried also conceives as a type of religious revelation. ‘Presentness is grace’, he writes at the very end of his text. As is well known, Fried makes an opposition between the total, self-sufficient presentness of the modernist work of art and the ambiguous practice of minimalist duration, which emphasizes the relation of works to the contingent space of the spectator (theatricality) and the fragmentary and partial character of seriality.
Recent studies which reinterpret early Romanticism suggest that we might do well to reconsider the theoretical contributions of this period and, consequently, also rethink the philosophical and artistic traditions that followed in its wake. Early Romanticism is not simply a foreshadowing of Hegelian thinking. Nor is it merely the introduction of metaphysics into art. Rather, it announces a mode of contemporary thinking that maintains a critical and ironic relationship to philosophical systems, even when it speaks the language of idealism. This contemporary reading of theoretical and artistic Romanticism stems, in part, from a Deleuzean approach to Romanticism, rather than from a Heideggerian and Blanchotian one, according to which art reveals and emphasizes the truth (Being itself), through self-sufficiency and non-mimetic procedures (the Absolute Book, coined by Blanchot, in reference to Novalis and Mallarmé).  This raises a second question, to which I will simply allude: namely, the supposed or possible Romanticism of Deleuze. I will limit myself here to three examples of the Romantic ‘liberation’ or reconsideration of the thesis of the artistic absolute: incompletion, reversibility and fragmentation.
The necessarily unfinished
While many works of Romantic art and literature have come down to us in an incomplete form, this is not due to any artistic failing on the part of their creators. The Romantic practice of incompletion is deliberate, conscious, the result of a rejection of definition, a strategy linked to a theory of becoming and an understanding of the relation between art and life. The celebrated Athenaeum Fragment 116 lays down the basis for this necessary incompletion by declaring that ‘the romantic kind of poetry is still in the state of becoming [im Werden]; indeed, this is its true essence, that it can only ever become and never reach completion.’  There are several ways, in theory and in artistic practice, of understanding this fundamental incompletion. First, Romanticism has especially highlighted an infinite, natural process, the process of natura naturans and poïetic power. So much so that pictorial or literary works that undertake to express this natural process, as opposed to copying appearances, deconstruct representative space, its narrative function (post-Aristotelian, then Albertian), as well as the structured organization of this narrative (whether pictorial or literary). Caspar David Friedrich, whose paintings stretch or abolish centralized perspective, refers to a ‘free and spiritual reproduction of nature, the true creation’.  Balzac, in Le Chef-d’œuvre inconnu, makes the expression of life the essence of any art that approaches the limits of figuration or of art. Paul Klee gives painting the task of exploring nature’s womb. As soon as infinite becoming takes over as the inaccessible object of the work of art, we have the appearance of an abstract aesthetic, or a figural space that approaches the limits of representation. Thus Novalis sees in the fairy tale (Märchen) a kind of dreamlike writing that resembles Kunstchaos – a chaos of natural forms and oneiric spaces – more than any teleological unfolding of a story.
Second, however, setting aside this great natural model, we could also point out that the Romantic practice of incompletion already calls into question modern intransitivity and modernist self-sufficiency. Friedrich Schlegel remarks in his Course of Transcendental Philosophy of 1801 that absolute truth cannot be known, much less presented in a work of art. From this follow three essential propositions of Romantic thinking: ‘All truth is relative’, ‘All Philosophy is infinite’ and ‘All knowledge is symbolic’.  Far from recording absolute truth or resolving the conflict between spirit and nature, the work of art is seen, in this perspective, as a partial form, a point of view, one object in a series, or an exercise in becoming. This opening up of the work confers on it the status of something at once provisional and experimental. Novalis said, not without irony, that the Laocoon Group is a ‘member of a series’, a ‘study’.  We could also take the example of Cézanne’s Grandes Baigneuses or his multiple versions of the Montagne St Victoire, which confirm this change in the status of the work of art, now something open and incomplete, like a sketch. The line between the sketch and the work becomes porous. Maurice Merleau-Ponty writes in his essay ‘Cézanne’s Doubt’: ‘Expressing what exists is an endless task.’  Picasso, for his part, said to André Verdet: ‘We make studies, we don’t make paintings. We approach them, we get closer and closer, but this never finishes.’  In short, this approach or approximation – in the Romantic sense of an endless approach or infinite aspiration – has shaped much of modernity, which grants specific artistic qualities, however contradictory, to the study, the sketch, the draft or the prelude. The transitory and undefined character of form became a new field of experimentation, a way of escaping from academic canons and predefined rules.
Contemporary artists too, following the Romantic decompartmentalization of the autotelic entity, have created fragmentary, partial works. Think of the use of repetitive or indefinite series in minimalist sculptures, of modular variations in Tony Smith’s and Donald Judd’s works, and of Robert Smithson’s series of sculptures arranged in order of increasing or decreasing size, such as Alogon #2. While these works do not belong to Romanticism, at least not in the usual historical and teleological sense invoked by this notion (from the end of the eighteenth to the mid-nineteenth centuries), they do take up one of the major operations of contemporary Romanticism: to leave form incomplete. In these works, form becomes a relative point of view on space and time, a module, an unstable segment in a total field that is impossible to cover entirely.
In this respect, the conceptual activity of On Kawara, totally and necessarily incomplete, consisting of painting on a small canvas the date on which that painting was made (Date Paintings and Today Series), might seem to be the continuation of the necessarily utopian incompleteness of Romanticism. A continuation understood here as an Aufhebung in Hegel’s sense (cancellation, preservation and overcoming). Here is what Lucy Lippard writes of On Kawara’s work:
the Today paintings are a series of fragments – more real, more like life. Seen from a microcosmic point of view, they emphasize the actual minutes in which they were made. From a macrocosmic point of view, they are a single incompleted work. Like life, therefore, they balance on an edge between continuity and discontinuity. … Literally speaking, continuity means nothing and discontinuity means existence. The Today series started and has not ended, so one could describe it as existing and not existing. 
To come back to the strictly theoretical dimension of Romanticism itself, we can observe that the many variations of interruption and fragmentary partiality destabilize the metaphysics of the intransitive auto-synthetic work. Two founding postulates of systematic thought and of intellectual intuition are thereby called into question by the early Romantics: (1) the unity of form and content; (2) the reconciliation of opposites within a synthetic unity. These two aspects of the Romantic critique of the system bring me to my second point: the role of structures of reversibility.
Structures of reversibility
I am going to rely for a moment here on the notebooks of Novalis known as the Fichte Studien. The first proposition essential to all knowledge, Fichte’s self-awareness or ‘act of positing’ (Tathandlung), becomes in Novalis a pseudo-proposition or an illusory proposition. Fichte’s Grundsatz turns into a Scheinsatz for Novalis: ‘The essence of identity can only be presented in an illusory proposition [Scheinsatz]. We abandon the identical in order to present it.’  According to Novalis, form and content are not identical, as they are in Fichte, for whom the form of the first proposition is itself the content, in other words an identity, and the content is the content of its form. On the other hand, with Novalis, Romanticism discovers that this self-awareness, in positing itself, reflects a self-consciousness more original than itself. Novalis speaks of Urhandlung, originary act, in order to distinguish it from Tathandlung, auto-positioning of the self. Thus intellectual intuition is not the absolute in action, but its reflection: it does not express unity, but rather the distance of absolute being, this originary act, also called ‘God’, ‘chaos’ or ‘nature’ – the terms vary. In consequence, as Novalis writes, the absolute, which is the ‘original ideal-real and real-ideal’, only appears to consciousness as ‘accident or halfway, then it must appear inverted – the unlimited becomes limited and vice versa. / Application to the original act.’  Towards the end of the manuscript, we can read: ‘This Absolute that is given to us can only be known negatively, insofar as we act and find that what we seek cannot be attained through action.’  The absolute remains therefore a postulate. Since the absolute is not only inaccessible as such, but only given incompletely to the subject, form does not merge with content, does not contain essence, but expresses disjunction and distance, lack of resolution. Hence the importance in Novalis of the play of opposition between contraries and figures of inversion, including the recurrent specular figures in the Fichte notebooks. His name for this fundamental structure of inversion of consciousness and its object is ordo inversus.
We must not forget that in this period ‘reflection’ refers specifically to specular access to identity (in Fichte and Schelling). Absolute identity is the self’s coinciding with itself through the medium of the mirror. To reflect is literally to recover, in the reflection, the power of the original image, the archetypal model. On the other hand, Romanticism was much more interested in the play of specular difference than it was in ‘mirrors of identity’. It sought in the mirror that which disturbed, upset or destabilized this identity, that which left it incomplete. The absolute thus presents itself as the impossible combination of two reversible sides, like the enantiomorphic structure (in Greek enantios-morphé, opposed forms) represented by our right and left hands, which are identical but non-superimposable, as Kant had already noted, since each occupies a specific region of space. ‘What is a sense?’, asks Novalis, ‘Homogeneity of nature with the person. The person in reverse is nature – Nature in reverse is the person.’  Or again: ‘The image is always the inversion of being. What is to the right of the person is to the left in the image.’  For the Romantic, in short, seeing oneself in the mirror is not only to rediscover oneself via the other, but above all to see one’s double, or to see oneself as an elusive, irretrievable double. This Romantic mirror stage is anti-Lacanian, since it undoes the self, leaves it incomplete, instead of allowing it to construct its identity through the gaze of the other and through imaginary recognition. The self-image, or image of self as other, is thus a radical ordeal for the self, opening the way to an encounter, real or ideal, with its Doppelgänger, a real or ghostly double, or to what the American mathematician Martin Gardner called ‘anti-worlds’. The enantiomorphic function revealed by the mirror makes apparent the asymmetry of the universe, the other side of the real, the hidden, inaccessible part of the world presented in the specular inversion. Lewis Carroll pays great attention to these fantastical inversions in Through the Looking Glass, and the question provides the substance for an enigmatic sculpture by the American artist Robert Smithson, who closely read Gardner’s book.
In 1966, Smithson created a false stereoscopic sculpture in two separate parts called Enantiomorphic Chambers, attached to the picture rail, like paintings. This mimimalist work is composed of two obliquely facing mirrors placed inside steel structures. The metal panels act as blinders for the person viewing the work. Smithson places these enantiomorphic optical chambers on the viewer’s left and right sides, so that the chambers ‘cancel out one’s reflected image, when one is directly between the two mirrors’.  This reflecting apparatus creates a kind of blind spot in the viewer’s vision, since he or she cannot see themselves directly in the mirror, but only partially, while at the same time sees what is behind him. The work upsets the pictorial laws of centralized perspective, those laws that make the spectator the structuring axis of the world presented. Similar to another work, The Eliminator – two mirrored panels, mounted at right angles, that reflect at intervals a blinking zigzag of neon light, which Smithson described as ‘a clock that doesn’t keep time, but loses it’  – Enantiomorphic Chambers is a disorienting work that plays on decentredness and ‘reversible structures’. The artist strives to pierce an opening to the beyond, or, rather, to create an extra-perceptive synthetic space, an hyperspace similar to the fourth dimension imagined by science and mysticism, a dimension in which oppositions and asymmetries are said to converge and be reconciled, as Jennifer Roberts notes in her book Mirror-Travels: Smithson and History. 
It is remarkable that the intuition of another reality in Novalis already occurs by means of enantiomorphic inversion. To the specular opposition of right and left, Novalis adds that of the inner world and the outer world, the world below and the world beyond, each being always the inverted image of the other, creating an impossible unity but a complex whole. In his Notes for a Romantic Encyclopaedia, Novalis writes:
The body is the interior realm in the inverse world, and the spirit is the exterior – the solid – etc. Fluctuating body – fluctuating spirit. All bodily operations are an inverse thinking. What is thinking, sensing etc. here – is burning, fermenting, thrusting etc. yonder. 
Thus, for Novalis, the beyond is not a pure utopian fiction. It is a change in the relationship between things and an inversion of the bonds between the spiritual and the corporeal: absolute unity being the reconciliation of opposites in a supreme reality (such as the hyperspace of the fourth dimension). Martin Gardner writes in The Ambidextrous Universe:
Our world and the anti-world could be identical just as our right hand and left hand are identical. We have only to take two imaginary leaps instead of one – a leap into a superior dimension’s space, and a leap into a superior dimension’s time. We who are prisoners of three spatial dimensions and one temporal dimension, let us see these two worlds in the form of mutual reflections evolving in opposite temporal directions. For a hyperspirit in a superior space–time, this world and that anti-world would appear to be identical. 
Though it already aspires to an ultimately conciliatory reality, Romanticism lays particular stress on its inaccessible character. In short, as all of these artists and philosophers claim – Novalis with his ordo inversus, Schelling in his dialogue on somnambulism, Clara, conceived as the inversion of the relationship between mind and body, and Smithson with his specular anti-worlds – the world beyond is the reverse image of the real, the world below flipped backwards or upside down. Considering the way he relates transcendence to perceptible immanence, we might wonder whether Novalis, in his own still mystical way, had not already conceptualized what Maurice Merleau-Ponty refers to in his late philosophy as ‘structures of reversibility’, structures that may be accessed only by the intermediary of the synergetic and reflexive body, the flesh. In Merleau-Ponty, the enantiomorphic function is just as important as it is in Novalis. It is what allows him to conceptualize the link between the visible and the invisible. The invisible does not refer to an ‘ulterior world’ (Nietzsche) but to the ‘total visible’ (or the ‘whole of the visible’), which is always ‘behind, after or between the aspects we see of it’.  As such, it is inaccessible and at the same time constitutive of the world that reveals itself to us, always entirely and incompletely. This reversible link between the visible and the invisible, which he calls the flesh of the world, is revealed by the experience of the reversal of the right and left hands: ‘My left hand is always on the verge of touching my right hand touching the things, but I never manage to make them coincide…’ 
This leads me to the third aspect of that contemporary Romanticism whose theoretical nature and possible artistic future I am attempting to determine: namely, thinking through montage (rather than system); that is to say, the construction and elaboration of concepts. This requires us to take up once again the question of the fragment, which has served as a common thread, whether in reference to incompletion or the reversal of opposites.
Montage versus system
Lacoue-Labarthe and Nancy in The Literary Absolute, along with Maurice Blanchot, distinguish the fragment, which they suspect of succeeding in recomposing a whole (a little whole or deliberate incompletion), from the fragmentary, the essential gesture of caesura, dissemination and dissolution that, by making the work impossible, make its problematic and philosophical character truly emerge. The fragmentary comes out of the more essential logic of ‘disaster’, that of the work of the absence of the work, as Maurice Blanchot puts it in The Writing of Disaster.  Pierre Boulez, for his part, has argued that every work since Romanticism, Mallarmé and Kafta is like a fragment ‘of an imaginary Whole’, the utopia of an immense work (masterpiece, total artwork…) whose completion is always postponed, deferred, approached yet necessarily impossible to achieve. Blanchot’s approach to a literature of absolute disaster, which underlies the reading of Romanticism in The Literary Absolute, must of course be taken with caution, particularly for its partiality and its insistence on a hermeneutical approach. It ignores, among other things, the poeïtic operations of Romanticism, the chance calculations, the role of drives and instincts, the involvement of a ‘synergetic’ and ‘reflexive’ body, to use Maurice Merleau-Ponty’s terms. Above all, this absolutization of the fragmentary negative overlooks Romanticism’s primary innovation in this regard.
That the fragment puts an end to a certain conception of the work of art, that it displaces it or profoundly modifies its nature, in no way implies ‘idleness’ or désœuvrement, nor that literary criticism is condemned to keep turning over this sort of negative theology. I also do not agree with Blanchot’s or with Lacoue-Labarthe and Nancy’s deeply ontological interpretation of the fragment, which does away with the necessary partiality and relativity of the Romantic work of art – the Romantic operation of deterritorialization – in order, finally, to re-absolutize the fragment by attaching it to a redemptive dialectic of incompleteness.
Between the absolute fragment and the critical dispersal of the whole, Romanticism in some sense imagined a third way, which has a lot to do with Deleuze and Guattari’s ‘rhizome’ and ‘plateau’: a sort of open and mobile system made of fragments, without any centre or stable position (auto-position), a paratactic composition (Adorno), a whole that is not absolute but variable and always in process. In The Infinite Conversation, Blanchot speaks of
a new type of arrangement not entailing harmony, concordance, or reconciliation, but that accepts disjunction or divergence as the infinite centre out of which, through speech, a relation must be created: an arrangement that does not compose but juxtaposes, that is, leaves each of the terms that come into relation outside one another, respecting and preserving this exteriority and this distance as the principle – always already undercut – of all signification. 
Blanchot defines very accurately here the kind of typically Romantic wavering between the binding and unbinding, the disjunction and juxtaposition of terms that are not absorbed into an all-inclusive whole but remain exterior to one another. It is not a matter of dissolving the distance between terms (subject and object, man and nature) but of investing and exploring this infinite distance. Post-Kantian Romanticism lies, from this point of view, in the unbridgeable abyss between nature and freedom that Kant alludes to in the Introduction to his Critique of Judgment-Power. I would suggest that this fragmentary disjunctive bond that Blanchot describes, this new organization that would take into account the outside, and so, by extension, a multitude of heterogeneous considerations, is a response to this Romantic project. On the literary and formal level, this also means that, as Roland Barthes points out, what is hors-texte (‘outside-the-text’) shapes the text, whenever it is a question of this fragmentary mode.
when fragments are strung together in a sequence, is organization impossible? Not at all: the fragment is like the musical idea in a song cycle (Bonne Chanson, Dichterliebe): each piece is self-sufficient, and yet it is never anything but the interstice of its neighbours: the work is made of nothing but what is outside the text, the hors-texte. 
Novalis and Das Allgemeine Brouillon
The very paradigm of this type of romantic paradoxical and fragmentary organization is offered by the four manuscripts of Novalis’s Notes for a Romantic Encyclopaedia (1798–99). The text is a strange, captivating object whose form remains difficult to define: at once a project for a concrete encyclopaedia of learned knowledge, complete with section headings, a treatise on Romantic methodology (Novalis here lays down the basic rules governing the combination of heterogeneous categories of thought) and a poetic metamorphosis of abstract knowledge. As Derrida made clear in La Dissemination, Novalis’s text is at once ‘organicist and tabular, germinal and analytic.’  It seems to prefigure Hegel’s Outline of an Encyclopaedia of Philosophic Sciences, but it diverges on one major point: the systematic and architectonic construction of the whole. Novalis combines at least two distinct ways of thinking: an empirical reflection, directly concerned with nature and the human body (the manuscripts are filled with physiological and medical observations) and a speculative and structural logic of the combinatorial imagination. The joining of disorderly nature and rigorous structure is one of the central, fascinating motifs of this inexhaustible text that constantly leaps from one register to another, from one speed to another, from one plateau to another, as Deleuze and Guattari would say. In this twofold perspective, which crosses Naturphilosophie and the logic of the Idea, the fragmentary apparatus of the Notes for a Romantic Encyclopaedia plays a central role.
The combinatorial analysis of concepts sketched out by Novalis, which makes constant use of permutation and displacement, also refers back to the ‘lesson of disorganization’ proposed by Friedrich Schlegel as a theoretical and aesthetic programme. Grotesques, dissonances, philosophical chaos, such combinations teach us, to quote Schlegel, ‘how to construct confusion with method and symmetry’.  Thus, going beyond the aphoristic form and the caesura, Novalis’s encyclopaedic system is plural and open to constant changes and metamorphosis. It breaks up and reconnects its parts without, however, retotalizing them by indexing them to a whole. This central text of Romanticism associates the profusion of objects and inexhaustible possibilities of combination and configuration, even on a grammatical and syntactical level. These two features replace the systematic concerns of idealism with the modern question of montage, of which Novalis provides us here with the first great theoretical and experimental model. And we could add that the fragment, understood as the interval between the whole and the anti-system, is a major component of this theoretical apparatus.
The meanings produced through montage are as numerous as the forms this type of association has taken on: plastic montage (Dadaist, Cubist and Surrealist collage), theoretical montage – if we think of Aby Warburg’s Mnemosyne atlas, his project for a non-chronological, visual history of art, making use of arrangements and combinations of disparate sources and discontinuous temporalities – and, of course, film montage (Eisenstein, Koulechov). These various forms of montage seem to have followed, rediscovered or expanded on Novalis’s ‘encyclopaedic’ project of creating an intermediary zone between the architectonic organization of the system, which assigns to each part a place in a pre-existing structure, and the empirical domain, which spells out and imitates the indefinite nature of the world, in the manner of a cabinet of curiosities or random collection of objects. Rather than dissolving the distance between opposites, Novalis invests the space between them and modulates these intervals through analogies and connections, creating sense and not system, since meaning emerges from a relation and signification is a matter of proximity and distance. This distancing from the types of totalizing organization promoted by philosophical idealism has opened up a particularly fruitful area of research in contemporary thought. Georges Didi-Huberman is correct in comparing the difference between Novalis’s Romantic encyclopaedia and Hegel’s Encyclopaedia of Philosophical Sciences to the distance that separates Warburg’s Mnemosyne atlas and Cassirer’s Philosophy of Symbolic Forms. In the Notes for a Romantic Encyclopaedia, Didi-Huberman writes: ‘It is not the unity of each domain but rather the circulation of relations between them that appears above all.’  Deleuze is probably the thinker who has come closest to this Romantic third way, without however naming it as such. This is what he wrote on the work of the American poet Walt Whitman, in Essays Critical and Clinical:
For Whitman, fragmentary writing is not defined by the aphorism or through separation, but by a particular type of sentence that modulates the interval. It is as if the syntax that composes the sentence, which makes it a totality capable of referring back to itself, tends to disappear by setting free an infinite asyntactic sentence, which prolongs itself or sprouts dashes in order to create spatiotemporal intervals. … Yet Whitman sometimes places the Idea of the Whole beforehand, invoking a cosmos that beckons us to a kind of fusion; in a particularly ‘convulsive’ meditation, he calls himself a ‘Hegelian’… But when Whitman speaks in his own manner and his own style, it turns out that a kind of whole must be constructed, a whole that is all the more paradoxical in that it only comes after the fragments and leaves them intact, making to attempt to totalize them. 
For Deleuze, the essential thing is to maintain a rhythm of thought that authorizes copula and additions along the way, a nomad logic of and (et), distinct from the sedentary and immobile ontologic of is (est), as suggested in Mille Plateaux. 
The necessarily ambiguous character of fragmentary thought, in Deleuze’s (and not Blanchot’s) sense of fragmentation, should be emphasized here, since it is a question – bringing us back to the Romantic project – of making connections rather than thinking them.
Deleuze the romantic, or philosophy as constructivism
The question of a contemporary romanticism can primarily be posed if Romanticism’s aesthetic theory and philosophical bases are reconsidered in a new light, by revealing, through Novalis’s work in particular, an entire theory of montage as well as an associative and process-based thinking. This theory refutes in advance the thesis of modernist purity and the autonomous work, which romanticism is supposed to epitomize. This rereading of the romantic question also owes much to the Pop-analysis of Deleuze and Guattari, whose gesture of deterrioritalization has reopened the closed field of thought and art. Since the 1960s, art has thrown the modernist autotelic view into crisis, represented by the ‘purity’ of Greenberg and the intransitive formalism of Fried, condemned in its time by Robert Smithson.  It is now characterized by the juxtaposition and the combination of art and non-art (Fluxus, Duchamp), by the mixture of high and low, of popular and high culture, of the central and the marginal, and by the many ways the work of art has been decompartmentalized and opened up to everything, initiated by the first German Romantics.
This being said, the relation of Deleuze and Guattari to Romanticism is complex and it will have to suffice, as a conclusion, to link the hybrid and pluri-disciplinary undertaking of romanticism – which, through Novalis and Friedrich Schlegel, unites poetry and philosophy – to the rhizomatic ‘pragmatics’ of Deleuze. What is at play in both is simultaneously the transmutation of understanding into a poetic and artistic act, and the creation of a new geography of thought, operating by shifts and branchings from one field to another (crossing disciplines), in order to give birth to the plural. ‘The spirit [Witz] is creative’, Novalis noted; ‘it makes [macht] similarities.’  This configures a plurality of analogies instead of a unitary taxonomy of knowledge. Deleuze and Guattari consider that the problem is not to replace the one with the plural, to turn ontology on its head (just as Nietzsche did not put the body, the ‘grand reason of the body’, in the place of the soul): rather one should have a method of thinking which makes its object instead of contemplating and reflecting upon it. ‘To attain the multiple, one must have a method that effectively constructs it; no typographical cleverness, no lexical agility, no blending or creation of words, no syntactical boldness can substitute for it.’  Deleuze and Guattari would go on to qualify this method, in a persistent echo of Romantic poetry-philosophy, as ‘constructivism’ or the creation of concepts: the juxtaposition of plural and heterogeneous conceptual planes (the other, for example, is a combination of several concepts, such as: reality–virtuality, face, language).  In this regard, it seems that the extraordinary theoretical and practical matrices found in the first Romanticism’s collections of fragments (published in the Athenaeum and in the manuscripts of Novalis’s Notes for a Romantic Encyclopedia) are more related to the Deleuzean deterritorialization and reterritorialization of concepts – as the link of the thought with the outside or the non-book – than with Mallarmé’s and Hegel’s myth of an absolute and substantial book as a microcosm of the world: ‘the classical and romantic book’, Deleuze and Guattari remarked, ‘constituted by the interiority of a substance or subject.’ 
Early Romanticism – and herein is found its genius – enucleated or removed the subject from the very heart of subjective idealism, by shifting the self towards an exteriority and a radical alterity that decentre both thought and creation. Contemporary art, which joins art with what is not art, perhaps began when Novalis noted in a flash of his usual brilliance: ‘Ego = non-ego – the supreme principle of all science and all art.’ 
This article is an output from the AHRC-funded project ‘Transdisciplinarity and the Humanities: Problems, Methods, Histories, Concepts’ (AH/1004378/1), 2011–2013, located within the Centre for Research in Modern European Philosophy (CRMEP), Kingston University London
1. ^ Novalis, Werke, Tagebücher und Briefe Friedrich von Hardenberg’s, hsgb Hans-Joachim Mähl and Richard Samuel, t. II, Das philosophisch-theoretische Werk, Carl Hanser Verlag, 1978, p. 420.
2. ^ Friedrich Schlegel, ‘Critical Fragment 115’, in Philosophical Fragments, trans. Peter Firchow, University of Minnesota Press, Minneapolis and Oxford, 1991, p.14.
3. ^ F.W.J. Schelling, System of Transcendental Idealism , trans. Peter Heath, University of Virginia Press, 1993.
4. ^ Friedrich Schlegel, Athenaeum Fragment 206, in Philosophical Fragments, p.45, trans. amended.
5. ^ August Wilhelm Schlegel, Die Kunstlehre: La Doctrine de l’art. Conférences sur les belles lettres et l’art, trans. Marc Géraud and Marc Jimenez, Klincksieck, Paris, 2009, p. 169.
6. ^ Novalis, ‘Monolog’, Werke, Tagebücher und Briefe Friedrich von Hardenberg’s, t. II, pp. 438–9.
7. ^ Michael Fried, ‘Art and Objecthood’, in Art and Objecthood: Essays and Reviews, University of Chicago Press, Chicago, 1998, p. 168.
8. ^ See Martin Heidegger, ‘The Origin of the Work of Art’, in Basic Writings, edited by David Farrel Krell, Routledge. London, 1978, p. 162: ‘In the work of art the truth of beings has set itself to work’; and Maurice Blanchot, The Infinite Conversation, trans. Susan Hanson, University of Minnesota Press, Minneapolis and London, 1993, p. 423. About Novalis and Mallarmé, see Olivier Schefer, Mélanges romantiques. Hérésies, rêves et fragments, Félin, Paris, 2013, pp. 117-29.
9. ^ Friedrich Schlegel, Athenaeum Fragment 116, in Philosophical Fragments, p. 32, trans. amended.
10. ^ Caspar David Friedrich, En contemplant une collection de peintures, ed. Laure Cahen-Maurel, José Corti, Paris, 2001, p. 54.
11. ^ Friedrich Schlegel, Transcendentalphilosophie, Félix Meiner, Hamburg, 1991, pp. 92–3.
12. ^ Novalis, Notes for a Romantic Encyclopaedia, trans. David Wood, SUNY Press, New York, 2007, no. 745, p. 137.
13. ^ Maurice Merleau-Ponty, ‘Cézanne’s Doubt’, in The Merleau-Ponty Aesthetics Reader: Philosophy and Painting, Northwestern University Press, Evanston, Illinois, 1993, p. 66.
14. ^ Quoted by Jean-Claude Lebensztejn, Études cézaniennes, Flammarion, Paris, 2006, p. 72.
15. ^ Lucy Lippard, ‘Just in Time: On Kawara’ (1977), in On Kawara: Whole and Parts, 1964–1995, Presses du Réel, Dijon, 1996, pp. 360–61.
16. ^ Novalis, Fichte Studies, ed. Jane Kneller, Cambridge University Press, Cambridge, 2003, p. 3.
17. ^ Ibid., p. 13.
18. ^ Ibid., p. 168.
19. ^ Ibid., p. 60.
20. ^ Ibid., p. 40.
21. ^ Robert Smithson, ‘Interpolation of the Enantiomorphic Chambers’, (1966), The Collected Writings, ed. Jack Flam, University of California Press, Berkeley and London, 1996, p. 39.
22. ^ Ibid., p. 327.
23. ^ Jennifer L. Roberts, Mirror-Travels: Robert Smithson and History, Yale University Press, New Haven CT and London, 2004, pp. 52–3.
24. ^ Novalis, Notes for a Romantic Encyclopaedia, no. 147,
25. ^ Martin Gardner, The Ambidextrous Universe: Left, Right, and the End of Parity, Basic Books, New York, 1964, ch. 29.
26. ^ Maurice-Merleau Ponty, The Visible and the Invisible, trans. Alphonso Lingis, Northwestern University Press, Evanston IL, 1968, p. 136.
27. ^ Ibid., p. 147, trans. amended.
28. ^ Maurice Blanchot, The Writing of the Disaster, University of Nebraska Press, Lincoln, trans. Anne Smock, 1995.
29. ^ Blanchot, The Infinite Conversation, p. 308, trans. amended.
30. ^ Roland Barthes by Roland Barthes, trans. Richard Howard, Macmillan, London and Basingstoke, 1977, p. 94, trans. amended.
31. ^ Jacques Derrida, Dissemination, University of Chicago Press, Chicago, 1981, p. 50.
32. ^ Friedrich Schlegel, Athenaeum Fragment 389, in Philosophical Fragments, p. 79, trans. amended.
33. ^ Georges Didi-Huberman, L’Image survivante. Histoire de l’art et temps des fantômes selon Aby Warburg, Paris, Minuit, 2002, p. 443.
34. ^ Gilles Deleuze, ‘Whitman’, in Essays Critical and Clinical, University of Minnesota Press, Minneapolis and Oxford, 1997, p. 57.
35. ^ Gilles Deleuze and Félix Guattari, A Thousand Plateaus: Capitalism and Schizophrenia, trans. Brian Massumi, University of Minnesota Press, Minneapolis and Oxford, 1987, p. 25.
36. ^ See, for example, Robert Smithson, ‘Letter to the Editor’, in The Collected Writings, p. 66: ‘Fried, the orthodox modernist, the keeper of the Gospel of Clement Greenberg, has been “struck by Tony Smith”, agent of endlessness.’
37. ^ Novalis, Werke, t. II, no. 732, p. 649.
38. ^ Deleuze and Guattari, A Thousand Plateaus, p. 22.
39. ^ Gilles Deleuze and Félix Guattari, What is Philosophy?, trans. Graham Burchell and High Tomlinson, Verso, London and New York, 1994, Ch. 1.
40. ^ Deleuze and Guattari, A Thousand Plateaus, p. 9.
41. Novalis, Werke, no. 83, p. 331.