A Crystal Matter: Virtual Structures and the Search for Belief
 Requesting more "structure" might seem peculiar for a number of reasons. Not only do Deleuze's cinema books move towards films liberated from linear structure, they are also rigorously opposed to cinematic theories informed by structuralism. Replacing the traditional tendency to reduce films to preexisting and signifying "structures" – linguistic, psychoanalytic, etc. – Deleuze attempts to understand cinematic movement on its own terms: as blocks of space/time that create an open-ended array of a-signifying and non-linguistic signs, images, and thoughts. In this way, it creates new structures of other sorts ("signaletic matter"). Theorizing cinematic movement could then no more be the applying of already existing outside models, but rather the act of creating concepts and an open semiotics extracted from and corresponding with the autonomous structure of the films themselves.
 Something similar goes for the philosophical understanding of what is generated and composed by the world itself.  These compositions, which do not follow any predetermined or teleological paths, take place on an immanent transcendental field where intensities, or "pure" differences, self-organize. This field is what Deleuze usually refers to as the virtual (although this is an elusive and at times inclusive concept). It is the genetic foundation – the "groundless ground" – of the world and everything that moves in it.  As esoteric as this may sound, it nonetheless regards unfixed but rigorous structures that are determined. " The reality of the virtual," Deleuze writes in Difference and Repetition, "is structure. [T]he virtual is completely determined."  However, this does not imply determination along the logical lines of propositions, essences, the general/particular, or any non-temporal organization built on negative and fixed difference – that is, a traditional logic of representation that only, and only to some extent, applies to actual state of affairs. (In this sense the virtual is completely undetermined). The virtual is rather constituted by "problematic Ideas" (defined multiplicities), which lack negative determinations and are non-localizable. Instead they are "positively" or "progressively" determined as reciprocal relations of forces in mobile systems. More precisely, they consist of differentiated elements or "emissions of [pre-individual] singularities" that "possess a mobile, immanent principle of auto-unification through nomadic distribution, radically distinct from fixed and sedimentary distributions [...]"  But these nomadic systems or problematic Ideas also possess the capacity to be actualized – they seek "solutions" – through intensive processes of differenciation: forming qualities or organical and non-organical life/material extension. 
 The epistemological capacity for determination here lies in a "transcendental sensibility" and a painstaking "transcendental empiricism" that seeks to extract these (always preserved) virtual events from actual state of affairs (LS, 118). This means abandoning preexisting schemas in favor of a creation (or recreation) of concepts that follow a logic of multiplicities, extracted from or corresponding with relating strands of contemporary physics, biology or mathematics, rather than standard logic.  But as this discussion of the virtual concerns the relation between the world and philosophical thought, what about cinematic thought? The aim of this article is to pose questions and speculate on how to understand the relation between virtual structures generating matter and those cinematic images that, according to Deleuze, present virtual time directly.
 Whilst there are no general aesthetics in Deleuze, What Is Philosophy? proclaims that art is about the creation of percepts and affects or "blocks of sensation." But in fact, an even more general definition is often invoked: art exposes a direct relation with the virtual. "Life alone," state Deleuze and Guattari, "creates such zones [of indetermination] where living beings whirl around, and only art can reach and penetrate them in its enterprise of co-creation." 
 What seems to unite Deleuze's different readings of art is art's creative relation to the virtual in some form or another. "Percepts" and "affects" differ from perceptions and affections in that they do not belong to actual state of affairs. The definition of percepts is to "make perceptible the imperceptible forces that populate the world, affect us, and make us become" (WP, 182, 164). This notion of being able to "reach and penetrate" the virtual, in some form, is at work in Deleuze's Proust and Signs as well as in his two books on cinema, where art and signs in a sense are described in relation to its ability to bring forth and connect directly with the virtual. But the virtual in Deleuze's philosophy at large takes on several forms – or is perhaps presented through slightly different images of thought – with different levels of complexity. Different virtual structures (Ideas) also actualize different things – organical and non-organical matter, the present, affects, qualities etc. (DR, 232, 235). But in what sense do virtual structures come forth in art/cinema? In regards to the complex different/ciations of material Ideas, are there certain limitations? If so, can they be transgressed by art/cinema itself?
The cinema and the virtual
 Among the arts, the automatic movement of the cinema bares an ideal capacity for producing sensations that connects thought with virtual intensity. But this is a potential realized only by some images. Cinema 1: The Movement-Image discusses images with an indirect relation to virtual time, where Cinema 2: The Time-Image deals with images with a more direct relation.  (These categories are in some sense generalities, covering an immense array of different types of images and signs, and mixed formations and combinations of images and signs.) The movements of matter (image=matter) in the movement-image bases itself on a sensory-motor logic. Therefore, although not spatializing time (which Bergson described cinema as doing), they still reduce time to the measuring of movement in space. As such, they foremost concern mobility in the actual.  In these images, moving parts will relate to an open whole akin to a notion of the one and the many forming a rational totality. Consequently, they abide by some sense of truths already established. Although virtual time is a force kept on the outside by the sensory-motor schema, there are many exceptions where the virtual is allowed to enter more or less directly (more on this below). But in these cases, virtual intensity tends to be rechanneled into the closed set of a dogmatic image of thought, which totalizes the moving parts and re-subordinates time to it. Even the disruptive movement-images of dialectic montage (Eisenstein) are in the end synthesized into a rational whole where change is retroactively calculable and predictable as a teleological unfolding, reducing time to the measuring of change in the whole as such.
 The separate parts of the movement-image relate to an open whole of matter-movement which is also an open whole of virtual time/past. That is, not just extensive space as a whole but the world in its entirety as change and durée. This "whole" equals the virtual in a somewhat simple and organical form: a general "plane of immanence" which includes all the co-existing levels of contraction and expansion in the whole of time (where also the actual images not shown are "virtual" in the sense that most actual images in the whole are only "virtually" present for us). But as time is subordinated to movement, this whole is only presented indirectly. However, the movement-image has a tendency to deterritorialize and expand towards a beyond of the sensory-motor schema. In this way it might even come to include ("relaxed") elements of direct time in the form of "recollection-images," "dream-images," or "world-images" – more or less neatly separated from the actual.  But as the movement-image keeps inside, or its exceptions are rechanneled into, a senso-motoric logic and/or a pre-existing organical totality/truth, the virtual as transcendental time will never really be released as a direct image. 
 Time-images, on the other hand, show transcendental time directly. They cease to base themselves on the sensory-motor (or the image of thought that subordinates time to movement), and will thereby reveal time as ontologically primary. Instead of expanding sets, the images first of all start to follow the direction of contraction in relation to the virtual. This is what the lacunary moments in Italian neo-realism start to indicate. With the full force of the crystal-images, or the hyalosigns, which intensify the contraction of the virtual and the actual, we are given a direct relation not just with empty time but with the "fundamental operation of time," which reveals the present as a constant split between the actual present (which flows to the future) and its co-existing past (which it flows back to).  The actual in the image is opened up to its virtual underpinning, its own virtual "double," which forms close circuits of exchange that are so contracted that their respective status as actual and virtual become indiscernible (although in no way indistinct). With the even more complicated and disruptive chronosigns, time is further revealed as intensive force. The crystal-images and the chronosigns, in a sense, still relate to a whole of virtual time as co-existence. But the virtual, especially in the latter images, is no longer the "open whole" of durée but an "outside" of intensive difference. The outside is let in directly through irrational intervals between images – involving, or sometimes even consisting of, the image/sound relation – which have now become primary (TI, 179-81). In addition, the virtual past tends to turn into the coexistence of incompossible pasts and incommensurable becomings.
 But some of these images (for instance the notion of the actual image's "own" virtual double), might also open up for questions less general than both intensive becoming per se and the co-existing non-organical whole of time contracting/actualizing/individuating present material objects: virtual Ideas. How are we to understand the relation between the crystalline regime and virtual problematic structures actualizing matter? Concepts like "Riemannian spaces" are mentioned and referred to here and there, but never really explicated as determined or determinable virtual multiplicities. And apart from the Nietzschean exceeding of true/false and time as non-linear series, Cinema 2 mainly presents us with the general schema of transcendental time's relation to the actualizations (or subversions) of the present, and the consequences of this being presented – as thought-images – for thought. But what about the concreteness of more specific virtual structures that (through the present) actualize organic and non-organic life? The "whole of durée" as a gigantic multiplicity is rather imprecise in view of the rigorousness of different/ciation. 
Virtual determinations and art
 As stated above, we get different descriptions of the virtual throughout Deleuze's oeuvre. There is an elusive spectrum that perhaps can be regarded as spanning from the Bergsonian durée in general, as all the images in the open whole, or the plane of immanence (in that sense), to the determined genetic structures of virtual Ideas, and the complex processes of different/ciation. In the latter notion, the virtual is the reality of co-existing multiplicities that are determined and differentiated, up to the point of being explicitly mathematical. 
 In a way, Deleuze's conception of art can be seen as an exceeding of the actual into the intensive. On the simplest and most general level, this conception of art as revealing a genetic plane of reality from which the forms of the world springs forth could actually be argued, as also Miguel de Beistegui seem to suggest, to correspond quite well with for instance the phenomenology of art in Heidegger and Merleau-Ponty (de Beistegui, 16, 18-19). There are of course important differences, and Deleuze proceeds from a very different philosophical perspective, from which he describes phenomenology's relation to art as about transcendent ideals like Humanity and the Beautiful, which in the end only recreates doxa, incapable of forming concepts that can handle non-human percepts and affects beyond cliché (WP, p. 149-50, 178-9). But even if Deleuze in some ways has the reversed view of ontological ground compared to the Heidegger of foundations, and, alluding to Merleau-Ponty, pronounces that "the flesh is too tender" and too catholic/religious, and that the body without organs in the end is something completely different and more intense, etc., what in comparison seem explicitly original with the virtual in Deleuze's ontology, following a Bergsonian inspiration, is its structures and their correspondence with – and affirmation of – parts of modern mathematics and science.
 But what, then, is the relation between this philosophical-mathematical conception of the virtual, and the virtual we are shown through art in general and the cinematic crystalline regime in particular? The "scientific" and "mathematical" dimension of the virtual – multiplicities, distributions of singularities, differentiation etc. – is not a central part, if present at all, of the discussion of the virtual in the crystalline regime. Some of these concepts do appear, but in an including or general sense. The occasional use of terms like "differential" or "differentiation" in the cinema books does not have to relate in any rigorous sense to the notions of differentiation, differenciation and different/ciation in Difference and Repetition. "Deleuze uses the term differenciation," comments Ronald Bogue, "in what appears to be a broadly inclusive sense that does not draw on his earlier meditations [in Difference and Repetition] on the word."  This is, however, not quite the case – depending on how it is interpreted – with Deleuze's conception of art in Difference and Repetition itself:
When it is claimed that works of art are immersed in virtuality, what is being invoked is not some confused determination but the completely determined structure formed by its genetic differential elements, its 'virtual' or 'embryonic' elements. The elements, varieties of relations and singular points coexist in the work or the object, in the virtual part of the work or object [...]." (DR, 260)
Although slightly altered, this notion of art and determination is still there in What Is Philosophy? The expressiveness of both nature and art are described as "relationships of counterpoint" that "join planes together, form compounds of sensations and blocs, and determine becomings." (WP, 185, italics mine). But the problem here is in what sense these genetic virtual determinations, which art itself holds, coincide with the self-organizing virtual structures that actualize matter – and not just affects or percepts. Before coming back to this issue of virtual determinations of the world and of the work of art itself, there is an important aspect of Deleuze's own notion of virtual time that needs to be discussed in relation to what is invoked by the cinema books: memory and the past in itself.
Overcoming memory and exceeding the past
 Although I am not sure how to fully understand the relation between Bergson and Deleuze on this point, it is safe to say that the full scope of the virtual is developed and changed in the latter. In any respect, virtual time is not just the "past in itself" (which preexists and co-exists with the present).  Deleuze posits a virtual that equals not just the in itself of the past as being, as presented in Matter and Memory, but an "empty" form of time, constituted by a transcendental field of pure difference beyond memory. Also, in Bergson, the past, although it is itself productively divergent and aimed towards the future, can never be reached by us directly (with the exceptions of intuition, dreams, and vague sensations like déjà vu) other than as a representing recollection-image, which then secures a relation to the pure past as mere sedimentary memory and a repetition of the same. Resemblance in Bergson is what allows something in the past to be actualized for us – either if it is called forth as part of habitual action or as recollection-images (MM, 213, 220, 313, 316, 323). However divergent (and differential) the virtual is in Bergson's notion of creative evolution, the virtual is "not a domain of the lived".
 "[I]t it is from Proust," writes Keith Ansell Pearson, "that Deleuze gets his crucial definition of the virtual."  Deleuze himself describes Proust's conception of time as "extremely different" from Bergson's (Bergsonism, p. 126, n 16). In the former, the pure past can be lived and experienced. But also the notion of the "past in itself" will, in a sense, be exceeded (but not discarded) by Deleuze. This will include a reading of Proust guided by Nietzsche and the eternal return: "The Proustian formula 'a little time in its pure state' refers first to the pure past, the in-itself of the past or the erotic synthesis of time, but more profoundly to the pure and empty form of time, the ultimate synthesis, that of the death instinct which leads to the eternity of the return in time." (DR, p. 149, italics mine; see also 153.)  Virtual past as the preservation of the same will then be firmly overcome, not least concerning our possibility to experience it directly.
 The movement to overcome memory is what the book on Proust shows us. The Search, as Deleuze's thesis goes, is not about memory, but the search for truth, and the orientation not to the past but to the future.  A first step is that a different form of recollection is introduced: involuntary memory. Involuntary memory, triggered by the "sensuous signs" like the Madeleine, can open up to the empty and pure past directly (DR, 106 ff). What is then invoked is not something once actual/present that is stored as such in the past, and actualized in the present as a resemblance, but a pure "essence" that rises up in its "internalized difference."  In the resonance between them the actual present (the tea-dipped Madeleine) sweeps up the past present (Combray) as a fragment of an "empty" past, making them share a more fundamental identity of pure difference. The invoked Combray then rises up as something "absolutely new" (PS, 60-1, 56 ff, DR, 153).
 However, this is not the end of the Search. Involuntary memory is in fact still too "bound to the actual," and can only "grant us essences in a slackened, secondary state and so obscurely that we are incapable of understanding the gift we are given and the joy we experience" (PS, 65). As the final stage of the search, art enters. This means a more determined grasp of essences through an overcoming of memory also in its involuntary form. Deleuze: "art is beyond memory [...] What art regains for us is time as it is coiled within essence, as it is born in the world enveloped by essence, identical to eternity. [...] there is only the work of art that lets us regain time" (PS, 45). 
 There is another important factor in Deleuze's rereading of the whole of time (of particular importance to some of the chronosigns in the cinematic crystalline regime): Bergson's cone is in Cinema 2 superimposed with Leibniz's pyramid of incompossible worlds. Consequently, Deleuze, in contrast with Leibniz but in agreement with Borges, does not regard incompossible worlds as mutually exclusive, but as co-existing in the same world (the whole pyramid). The past in itself, in this sense, therefore consists not just of everything that has been, but of everything that could have been, but in a way such as the past, repeating only difference, is opened up for the co-existence of many incompossible worlds within an immanence that never forms a linear totality (TI, 130 f. Rodowick, 98-9, 204, LS, p. 130-1).
 The cinema books, with some exceptions in the movement-image, follow a development towards more and more direct connection with the virtual, which, with some imagination (and possibly some violence), could be read along with the progression ending up in the stage of art in Proust & Signs, where art is described as having the ability to "regain" or connect directly with time in its pure state, beyond memory (which will then reveal how the "essences" were already incarnated in all the other signs and objects). Whereas the organical regime of the movement-image corresponds with the Bergsonian notion of "recollection-images," which only repeats the same, the crystalline regime will exceed the past as such and relate directly to a pure time rather through "the disturbances of memory and the failure of recognition" (TI, 55).
 This notion, of a beyond memory and representation, will start to clear up the virtual for being constituted by difference, multiplicities and becomings.  Such virtual becomings and events are also beyond Bergsonian durée: when "Bergson says that there is always time between two instants, however close to each other they may be," Deleuze and Guattari write, "he has still not left the domain of functions and introduces only a little of the lived into it. But when we ascend towards the virtual," they continue, "when we turn ourselves towards the virtuality that is actualized in the state of affairs, we discover a completely different reality [...]" (WP, 157, 158).
 The question then becomes, to what extent does "modern" cinema actually venture into this "completely different reality"? Without doubt, Deleuze's notion of a crystalline regime becomes Proustean and Nietzschean in the sense of a presented virtuality beyond memory and the subversion of already existing truths. But could there be images showing the determined structures of the virtual of material immanence, without turning them into – which is the conduct of royal science – something actual?
The past and the chronosigns
 While the narrative might take the route through memory, as in the case with In Search of Lost Time, memory is only a means to an end. Crystalline images do not just actualize, in a shattered and paradoxical manner, memory that is only "virtually" present for us. They rather re-create the past into parts in virtual events, beyond the actual of history, which eludes the empirical present in de-actualizing it or in no longer tolerating "the separation or the distinction of before and after, or of past and future" (LS, 3). The virtual event also enfolds the past, present and future as co-existing in becomings, or as parts of incompossible "wholes." This is brought about in different ways (and to different degrees). The "hyalosigns" shows variations of the present/past or the virtual/actual as indiscernible from each other. The different "chronosigns" complicate matters even more through an accentuated refusal to unify the parts into wholes, paradoxical and incompossible compositions of pasts and presents that profoundly subverts successive and linear time (either in the form of co-existing "peaks of presents" or "sheets of past"), power of the false/un-linear series (becomings), etc.  All these compositions entail some form of irrational cuts, where the interstice or conjunction itself becomes primary (causing the virtual "outside" to enter – "topologically" merging inside/outside – which forces us to think the "unthought within thought"). In different ways, then, the virtual past is transformed into something in itself different.
 But as complex as this difference can be, how much virtual structure are we presented with, and in what sense? In a way, the "fundamental operation of time" shown in the crystal-image is general and schematic. Including the virtual events shown by the chronosigns, and even the "readable" thought-images, the virtual presented in the crystalline regime is in a sense undetermined, not fully corresponding with the notion of self-organizing virtual Ideas. That is, not merely the virtual Idea that the work itself constitutes (see below), but that it shows us (as the crystal-image shows us "the fundamental operation of time") how material extension is a divergent "solution" for complex problematic structures on the virtual plane of the world.
Intensity in movement-images and limitations in the crystal
 If what is ultimately at stake here is the relation between the cinematic virtual and the complexes of material immanence, how are we then to understand the many instances among the movement-images that transcend actual state of affairs and even become intensive or "genetic"? Let us discuss some examples.
 For instance, among the "perception-images" there are "liquid," "molecular," and finally even "gaseous" images. They produce intensive "matter-images" beyond human perception, and might reach "the genetic element of all perception" (MI, 84-5). Is this then the virtual of matter? It certainly regards matter somewhat beyond actual state of affairs – for instance, advancing Bergson's general notion of perception directly inside matter into an eclectic non-human perception (seeing from all perspectives at once), as the case with Vertov, followed by American experimental cinema. But these types of images are still categorized inside the logic of movement-images, where time is still subordinate. They may transcend human perception, but are still subjected to the temporality and some of the logic of the actual – only a much more intense, varied, microscopic, non-human, or shattered conception of it (not all that far from royal science using technologies like satellite systems or microscopes showing the movements of molecules). In the case of Vertov's matter-images, they are "not subject to time, which has 'conquered' time, which reaches the 'negative of time', and which knows no other whole than the material universe and extension" (MI, 81). Although the gaseous "camera-consciousness" of American experimental cinema is described as raising "itself to a determination which is no longer formal or material, but genetic and differential," it does so with regard to the formation or determination of the subtracted perceptions we form as humans. In other words, it does not necessarily have to do with genetic structures in the virtual that actualize the matter there for us to perceive, but could be interpreted as merely complicated states of actual matter more or less only "virtually" present for us, or caught in the act of becoming perceived. And also, their "differential" aspect, which more or less regards actual matter, lacks any further or explicated determinations. To the extent, then, that these images reveal any form of genetic side of matter, it is perhaps in line more with for instance a phenomenological conception of art (see above). And last but not least, all their intense variations tend to be totalized into the closed set of some kind of established image of thought – communism (Vertov), drug utopia (American experimental cinema), etc.
 A similar case pertains to some of the affection-images and their more determined virtual elements. These images manage to extract themselves from being defined by their part in the sensory-motor chain (although on a larger scale they still tend to fall in line with a whole that re-channels its force). They might take place as close-ups of faces, but more subtly in spaces that themselves transcend metric determination – a sort of "any-space-whatever" (espace quelconque), which becomes the "genetic element of the affection-image." As such, they hold other types of determinations: "ideal singularities" and "virtual conjunctions." However, these strands of rigorous virtual structure in the affection-image deal with the constitution of pure affective qualities (the "glistening" of a leaf, the "sharpness" of a knife) – that is, not the formations of matter itself. The singularities are of a fairly actual kind, and the notion of "virtual conjunctions" regards connections, outside actual state of affairs, between qualities, rather than the self-organizing structures of matter described in Difference and Repetition (MI, 102-10).
 The "dynamic sublime" of German expressionism might be a more difficult case, and its "any-space-whatevers" seems to hold both almost Nietzschean and crystal-like aspects.  By means of shadow and light, "infinite" intensive non-organic forces are at play behind the formation of material objects, turning space into an indeterminate and "spiritual" space of pure powers, qualities, and affections. However, this seems to be more of an unclear classical vitalism than Deleuze's own rigorous and structured conception of the virtual and the actual.
 In comparison, the images in the crystalline regime are founded on another conception of time. They thereby give us direct images of virtual time in ways that are more unrestrained, and which cease to totalize into more or less commonsensical wholes. But as they penetrate the virtual directly, these images seem to risk a disconnection with the concreteness and specificity of matter and material genesis.  Even as the crystalline images intricately connect with un-linear structures of time, becomings, and thought, they are limited in regard to the explication of the virtual Ideas of non-actualized matter. In this sense, they do not yet show us the full "reality of the virtual."
 In any event, the problem of structural determination in art can be challenged from a different perspective: the work of art, as well as the concept, constitutes a form of virtual determination in itself.
Creation and capturing: Three different virtual determinations
 One aspect of the "obscurity" of involuntary memory lies in not being able to seize its own virtuality: it is not captured. The creations that are philosophical concepts and works of art on the other hand preserve themselves as determined virtual expressions that are created. "[T]he power of the false," states Deleuze, "is [...] the most general principle that determines all the relationships in the direct time-image" (TI, 131, italics mine). In a sense structurally similar to the virtual Ideas of the world itself, the concept, which has extracted the event from the state of affairs, relates to the actual from the position of virtual determination. The concept contracts heterogeneous elements, made up of "inseparable intensive variations," which it (re-)determines. The concept, as well as the work of art, is a sort of virtual multiplicity created by the philosopher/artist, whereas, for instance, a virtual idea of an organism is a self-organized creation of the world. They share a similar kind of structure but are not the same thing. 
 Also art, concepts, and the virtual Ideas of the world are all potentially generative in some sense – they all create effects. But while the virtual structures of the world, in themselves, hold the dynamic capacity for self-creation or "dramatization" of their Ideas (spawning organic and non-organic life), the virtual structure of art and concepts bares only the potential to make us come in contact with the virtual forces of the world, which then makes us, or make us want to, become in some sense (WP, 182).  Art and concepts can create new thought, or make us connect with the unthought (virtual) in thought. The concept in this sense points to the future and "is the contour, the configuration, the constellation of an event to come" (WP, 32-3). 
 But the philosophical work behind the formation of the concept also entails thought to co-creatively retrace the virtual structures of the world that it first encounters as a chaos. But the event is not retraced in the same form it was encountered in. Philosophical thought will end up in the creation of a concept that determines the virtual process as a form of consistency (which does not mean that it is turned into something actual; the consistency keeps the event as a virtual co-existence, a "quasi-cause", with the actual beyond linearity). "[I]f we go back up in the opposite direction," write Deleuze and Guattari, "from states of affairs to the virtual, the line is not the same, because it is not the same virtual [...] The virtual is no longer the chaotic virtual but rather virtuality that has become consistent [...]" (WP, 156).
 The virtual of the world itself is only chaotic from the perspective of the actual. But the philosophical concept, which is extracted from actual state of affairs, "apprehends the event, its becomings, its inseparable variations [...]" (WP, 158). Through the concept, then, we become aware of how chaos itself in fact is "characterized less by the absence of determinations [...]" (WP, 42). But also the work of art has more to do with determining than a mere making-chaotic of actual or linear structures per se: "A work of chaos is certainly no better than a work of opinion; art is no more made of chaos than it is of opinion." "Art," they continue, "constitutes [...] a Chaosmos, a composed chaos – neither foreseen nor preconceived" (WP, 204).
 When becomings break with "the form of the true to replace it by the powers of life [...]," they leave behind the "foreseen" or "preconceived" and the organical regimes reference to what already exists as a truth, and which is therefore discovered (TI, 135). The crystalline notions of fabulation and falseness mean the creation of the singular and the new holding a truth which is not preexistent but created.  But as with virtual Ideas in nature the virtual of these new creations nonetheless holds determinations. As such they encounter us, more or less as chaos, and we can then try to "retrace" their virtuality and create some kind of corresponding conceptual determinations ("the intelligence come after"). But again the question is: to what capacity do these created virtual determinations in art itself, and more specifically the crystalline regime, correspond with the complexity of the genetic structures of the matter – and not just affects and percepts – which they make us come in contact with?
 The sensory expression determined in a work of art bears a sort of structural congruity with the world's processes of expressions, and apart from the former's inability to self-create, they are both formed on a kind of "plane of composition" (WP, 183-6). This could imply that art at least bears the potential to correspond more fully with the complexity of the world's processes of expression. In the cinema books, at least in view of their ending, cinematic expression seems to be mostly connected with the provoking of new thought. In relation to thinking, modern cinema/art constitutes a sign from the outside that "disorders the senses" and forces us to think the unthought and create new concepts (and, on a biological plane, possibly even create new circuits in the brain). This means that it is up to thought as philosophy or (minor) science, after the encounter of art, to explore the virtual "empirically" in all its complexity.
 But what kind of correspondence with such complexity already encounters thought in the crystalline images themselves? We are dealing with relations between at least three different planes here: the concepts (crystalline images) as virtual determinations (on a philosophical plane of immanence), extracted from the block of space-time in the image (art's plane of composition), in which we in some sense might see the determined virtual structures in the material processes of the world (the world's plane of composition). Apart from some references and notions like non-linear series and "Riemannian spaces," Deleuze's mathematical-philosophical conception of the virtual are toned down in the discussion of the crystalline regime. And Deleuze never really speaks of virtual structures in the rigorous sense of different/ciation (with its constitutive parts). Un-linear and non-localizable space, whether in the form of "Riemannian-," "Quantum-," "topological-," or "crystallized" spaces, are mainly a more general matter of "direct presentations of time" (TI, 129).
 The "indiscernability" of the crystal-image has a specific significance in how it seems to apply for art in general: "[A] zone of indetermination, of indiscernability," state Deleuze and Guattari, "as if things [...] endlessly reach that point that immediately precedes their natural differentiation" (WP, 173, 182 f.). "Indiscernability" between the actual and the virtual is not a philosophical apprehension of determined virtual structures of non-actualized matter, but a general showing of how art, as well as nature, "combines these two living elements [virtual and actual] in every way [...]" (WP, 186). 
How fixed are the limits of art?
 The somewhat indeterminate constitution of the virtual in the crystalline regime – along also with it in many ways is still too "bound to the actual" – shows that it does not fully correspond with the complexity of "the depth of things." In conclusion then, maybe this general display of the forces of virtual becoming in and of actual things really does show us a sort of limitation in what art can do. "Deleuze's thought on the arts," as Ronald Bogue states, "constitutes only one dimension of his philosophy [...]" (Deleuze on Music, Painting and the Arts, 195).
 Limitations in the crystalline regime might be further illuminated by Deleuze's earlier notion of "counter-actualization" in art. One of the forms of the crystal-image relates to the circuits of indiscernability of the actor and the role. In The Logic of Sense, Deleuze describes the relation between the actor and the event like this: "The actor [...] actualizes the event but in a way that is entirely different from the actualization of the event in the depth of things. Or rather, the actor redoubles this cosmic, or physical actualization, in his own way, which is singularly superficial [...] and keeps from the event only its contour and its splendor, becoming thereby the actor of its own event – a counter-actualization " (LS, 171).
 As stated, it is after being confronted with art that we have to become philosophers and from there explore the virtual more fully and empirically. "[A]rt," writes Manuel DeLanda, "is concerned with making perceptible the usually hidden realm of the intensive. Similarly, philosophy must make the virtual intelligible. Philosophy must go beyond the centers of convergence where the larval subjects of percepts and affects undergo intensive becomings, to reach the virtual in its full divergence and difference [...]"(ISVP, 217). But how far can we imagine art, even future forms of art, going, in itself, in this direction, and therefore in making us see that there is something more than just intensive becomings in general in the world for us to explore? If film is itself a form of thought – more or less – can we imagine thought-images that, themselves, resonate deeper with the complexity of Deleuze's own philosophy of matter?
 "The definition of the percept itself," according to Deleuze and Guattari, is "to make perceptible the imperceptible forces that populate the world" (WP, p. 182). But as all virtual structures are "non-localizable" and in a sense "incorporeal," they might also imply the impossibility of being perceived in a more explicitly determined form by something as simultaneously actual and fleeting as a moving image. The cinematic moving image might only be capable of showing virtual time on a more general note, and to provoke thought which then has to think procedures like different/ciation through other means. But just as the cinema has brought along totally new possibilities of visualizing intensive levels of the world, future technologies might present us with possibilities not yet thought of in this regard. With the ongoing advancements in technology and science, who knows what will come? 
 Deleuze himself stated that we are (were) in the middle of the formation of a third wave of images, based on "yet unknown aspects of the time-image," taking over after the "modern" regime. But in this new regime, which includes the electronic image (video, television, digital images), "information [is] replacing nature" (TI, 265-6, 269). This can be interpreted as "a postmodern phase of cinematic 'mannerism,' such that the proliferation of images threatens to extinguish nature itself."  Clearly, such destructive deterritorializations, which only further disconnect us from the world, would be the opposite of the sort of virtual "reterritorialization" this paper is arguing for – and of the general task of restoring a belief in this world that Deleuze – although he doubts that cinema is sufficient for this – ascribes modern cinema after the fall of grand truth. But the future holds endless possibilities and new forms of time-images are to be expected in a future that is always open.
 "One gets the sense," states D.N. Rodowick, "that, for Deleuze, the cinema of the movement-image has been fully realized while that of the time-image is emergent" (Rodowick, 89). But our current audiovisual "society of control" poses a great amount of challenge, with its ever expanding capacity to swallow and re-channel also resisting aesthetical strategies into totalities of communication, information or at least the axioms of money. Avoiding a sad Baudrillardian direction, maybe resistance has to be thought along more parameters than readable time-images – along new parameters? Maybe a request for the affirmation of virtual-material structure has become more subversive than comparably undetermined becomings more easily reterritorializeable in a post-modern regime?  Cinema came along with a new capacity for art to put movement into thought. Maybe new technologies – possibly more directly intertwined with organical matter – can go even further? In this sense, could there be moving images, cinematic or otherwise, that – in order for us, in this day of age, to really be able to believe in this world – have a more explicit resonance with the self-composing structures of the world?
 In relation to philosophy, Deleuze regards the history of cinema as a sort of parallel history of thinking that, apart from the more general percepts and affects, creates its own autonomous thought consisting of sounds and moving images. But different regimes produce different kinds of thought, and different dispositions to the world. This also means that different images have very different relations to Deleuze's own philosophy of matter. If film with Godard became thought-structure with the capacity to equal Spinoza's Ethics, might it in the future become possible to think the Ideas of matter in the fourth and fifth chapters of Difference and Repetition? 
 Maybe the spiritual space of modern cinema is too "catholic"? In relation to a Nietzschean problematic, something essential is at stake throughout Cinema 2. How is modern cinema, after the fall of grand Truth, to restore a belief in this world? (TI, 181-2, 201-2). The time-images break with a dogmatic image of thought and may produce a shocking awareness of how "we are not yet thinking," and a will for us to become. But the virtuality of the time-image also in some sense leaves the discussion of the preciseness of immanent material genesis in favor of new thought and virtual becomings in general. Cinema 2 bends so much towards the merging relation between thought-images and thought (noosphere) – instead of the virtual structures of nature per se – that it finally suggests that "we no longer ask ourselves, 'What is cinema?' but 'What is philosophy?'" But maybe another question could have been posed at the end: "How do we truly reconnect thought with the world?" In any event, What Is Philosophy? was to be followed by a philosophy of nature... In order for the time-image to really restore a belief in the world, perhaps it has to be more firmly reconnected with the complexity of nature in ways not yet thought of?
 Speaking about the world itself in this sense can first and foremost be attributed, as Manuel DeLanda has suggested, to how Deleuze – with regard to perhaps the most fundamentally "untimely" aspect of his ontology – is a form of modern "realist." DeLanda: "Deleuze's main contribution to philosophy [...] is to have rescued realism (as an ontological stance) from the oblivion in which it has been for a century or more. In some philosophical circles to say that the world exists independently of our minds is tantamount to a capital crime. Non-realist philosophers (from positivists to phenomenologists) have created a straw man to kick around: the naive realist, who thinks we have unmediated access to the external world and who holds a correspondence theory of truth. So the key move here was to create a viable alternative form of realism to deprive non-realists of that easy way out. Similarly, when it comes to defend [sic] the autonomy of non-human entities (atoms, molecules, cells, species) the crucial maneuver is to account for their mind-independent identity without bringing essences into the picture." (Deleuzian Interrogations: A Conversation with Manuel DeLanda, John Protevi and Torkild Thanem; see also Intensive Science and Virtual Philosophy, s. 4ff. [Hearafter ISVP]). Deleuze has an important predecessor here in Bergson, although the latter tries to formulate a third way between idealism and a more traditional notion of "realism." (Matter and Memory, p. viii ff, 38, 45, 236, 240-1 & 271 [Hereafter MM]).
 "Intensity equals difference," "the condition of that which appears," which "is not space and time, but the Unequal in itself, disparateness as it is determined and compromised in difference of intensity, in intensity as difference" (Difference & Repetition, p. 281 [Hereafter DR]).
 DR, p. 260. On the matter of a potential change in Deleuze's relation to "structures" after Difference and Repetition, see note 28 below.
 The Logic of Sense, p. 118 (Hereafter LS).
 There is determination of different types on two different levels. The virtual structure is a differentiation, and its actualization is a differenciation, together they form processes of different/ciation (which are discontinuous, abrupt, asymmetric and pertaining to specific temporal rhythms on the virtual plane) (DR, p. 258, 261-2). More explicitly concerning organisms and material extension ibid. p. 262-74, 305-319. ("The entire world is an egg.") However, these processes are not one-directional. Most explicitly in the organism, the dynamic processes proceed through complex interaction (and the capacity to affect or be affected) with challenges posed by actual environments and material contexts (Ibid. p. 269. DeLanda, ISVP, p. 83-88). Regarding the notion of "Ideas": In Deleuze's ontology (skipping over Kant here), Plato is not so much overthrown as intricately reversed. The grounding "Ideas" (now completely immanent) are themselves made up of pure differences. "[T]he pure concept or Idea of difference within overturned Platonism [...]" (DR, p. 81). That is, in a reversal of traditional foundations, pure time and difference comes first (ontologically, not linearly), and self-identities are mere surface effects produced by dynamic processes – in immanence. But it is not an Aristotelian immanence. Virtual actualizations are not the same as actualizations of preconceived potentials. Not merely in the open non-teleology of the virtual, or the notion of the virtual as being fully real, but in how the circuit virtual/actual is not one-directional and in how the relation is asymmetric, they do not resemble one another (DR, p. 349, 260 ff.). The virtual is also never depleted in being actualized, it is always preserved on its own plane, intertwined with, and leaving traces inside, actual state of affairs.
 Apart from the general notion of philosophy as the creation of concepts, on an equally created plane of immanence, most of what Deleuze does boils down to the production of (or adherence to) an ontology or metaphysics of difference corresponding with the parts of contemporary science that do not belong solely to exact applications in actual state of affairs (such concepts would be mere metaphors if used by philosophers) but to related notions that science cannot do without, which are "inexact but rigorous," and which belong equally to scientists, philosophers and artists, allowing scientific ideas to "enter into resonance" with a philosophy of difference (a resonance which in Deleuze goes both ways since he has the ambition, following Bergson, to provide modern science with the metaphysics it lacks ) (John Marks, Deleuze and Science, p. 1-4). The same applies to mathematics (DR, p. 239). "The entire Idea is caught up in the mathematico-biological system of different/ciation" (Ibid. p. 273). However, as philosophy, science and mathematics are different procedures, Deleuze's concepts will always involve some sort of "philosophical transformation of the mathematical concepts involved" (DeLanda, ISVP, p. 78).
 Deleuze & Guattari, What Is Philosophy?, p. 173. (Hereafter WP.)
 Hereafter MI and TI. These two overarching regimes are not complemented by a "matter-image," since matter, accentuated in the more Bergsonian analysis of the first regime, is always implied in the very idea of an image, since matter itself – on a general level – is an aggregate of (moving) images. The main point of this Bergsonian notion of matter as image is that there is no difference in kind between the images we make of matter and matter itself. As "image," matter is then also equaled, in Deleuze interpretation, with light. Or rather, matter is a form of vibration (with different levels of contraction and relaxation) in which light is one of its aspects. Human consciousness is itself a form image (interrupting a flow of other images), and conversely all matter has degrees of consciousness. But even if our images of matter and matter itself coincide, our everyday consciousness cannot grasp matter in all its complexity. The images that we normally perceive are a form of subtraction, reducing matter to what we need in order to act and respond in sensory-motor situations. The three cinematic movement-images described by Deleuze, which equal the movements of matter only to the extent that human consciousness' images of matter are of the same kind but not the same degree as matter in all its complexity. Deleuze: "No doubt there can be more in matter than the image we have of it, but there cannot be anything else in it, of a different kind" (Bergsonism, p. 41).
 That is, the level of matter that has been actualized. The intensive process of actualization, in the present, is the most contracted and intensive form of time and matter, whereas matter as the extensive, the already actualized, is the most relaxed form of matter and time. On the level of actualized extensions and qualities, intensive difference is largely "cancelled" in so far as it "appears outside itself and hidden", which in fact is what happens objectively on the level of material extension as a form of "objective illusion" (DR, p. 287, 300; DeLanda, ISVP, p. 74-5). The level of the actual in matter, then, is a product of the world itself, and not a mere figment of specifically human dispositions.
 "Cinema" as the larger than life projection of "bigger and bigger circuits which would unite an actual image with recollection-images, dream-images and world-images" (TI, p. 68). These images belong to the less intensive "broader trajectories" of the cone (Bergsons's notion of the whole of time, with the tip of the cone as its most contracted part: the present) where time is more "slackened" – as such they can be organically linked relatively to the actual (or an individual consciousness) as a center (Ibid. 273).
 D.N. Rodowick, Gilles Deleuze's Time Machine, 48 f, 142-3.
 The virtual past coexists with itself in all its levels of contraction and relaxation, with its most contracted point being the present, which is a process that constantly splits up the present in two dissymmetric flows: back to the virtual and forward to the future (TI, 81-82, MM, p. 193 ff.).
 The co-existence of "memory" or "past" is at the end of Bergsonism presented as the co-existence of a virtual multiplicity as a whole, actualizing material objects (p. 113).
 DeLanda, ISVP, p. 9-41. This includes a correspondence with more or less recent theoretical developments in science like quantum mechanics, non-linear dynamics, and complexity theory (Miguel de Beistegui, Truth & Genesis: Philosophy as Differential Ontology, p. 15-6).
 Deleuze on Cinema, p. 206, n 17.
 Bergson posits the important idea of a past autonomously existing in itself: "[M]emory, always present in its entirety to itself [...]" (MM, p. 224). It is therefore not dependent on being contained by something actual, like a brain or a mnemotechnology. The Past is primary Being, consequently, the past does not follow the present, as a passive gathering of presents that has been, but something that already is in itself, and that creatively co-exists with the flows of the present (which is the most contracted point of this already present Past/Memory=Being) (Bergsonism, p. 74).
 Keith Ansell Pearson, "Deleuze on the Overcoming of Memory."
 In Difference and Repetition, Deleuze schematizes three syntheses of time. The first concerns habit, linearity, and bare repetition. The second concerns Bergsonian notions of a whole of time/past/memory which coexists with the present, causing it to pass. But even if it shows us the split between the actual present, and the a priori past which was never present, this second synthesis does not fully move beyond representation since the present resembles the past that grounds it. The third synthesis reveals the difference which grounds the two other syntheses, which are now both, the present as well as the past, subordinated to the future which holds the "royal repetition" of difference (DR, p. 117).
 Deleuze, Proust & Signs, p. 4, 26. (Hereafter PS). See also WP, p. 167-8.
 "[N]o longer the stable essence, the seen ideality that unites the world into a whole [---] but a kind of superior viewpoint [...]" (PS, p. 110). A viewpoint, which in itself is difference, into life as difference (PS, p. 41 f.). The essences or Ideas, then, belong to the future, and the sensations are not so much a discovery as a production or a setting forth.
 The notion of "eternity" in this quote refers to certain strands of Neoplatonism. Deleuze: "Eternity did not seem to [certain Neoplatonists] the absence of change, not even the extension of a limitless existence, but the complicated state of time itself. [---] [E]ssence is the birth of Time itself" (Ibid).
 This do not mean that the multiplicities populating Deleuze's virtual are without a sort of non-actual history – they are of course all but timeless in any direction. But here they also share a sort of complex co-existence with, and a "doubling" of, actual history (DeLanda, ISVP, p. 188).
 "The direct time-image here does not appear in the order of co-existences or simultaneities, but in a becoming [...] as series of powers." In which, as with the incompossible pasts, "the past is not necessarily true" (TI, p. 274-5).
 Rodowick: "[T]he dynamic and mathematical sublime of German Expressionism and French Impressionism [...] the episodic nature of the small form of the action-image: all are harbingers of time [...] " (Deleuze's Time Machine, p. 73). "[T]he time-image [...] has been haunting cinema from the beginning" (Bogue, Deleuze on the Cinema, p. 105).
 Although they, on one level, obviously never leaves material immanence in neither their own materiality (image=matter) nor the material objects presented: "The crystal image and its virtual image thus constitute the smallest circuit, ultimately a peak or a point, but a physical point [...]." (IT, 70, italics mine).
 As conscious human productions with a virtual structure, art, philosophical concepts, as well as strands of science are all "real without being actual, ideal without being abstract." Or as the ending of What Is Philosophy? asserts: "It is here that [...] philosophy, art and science become indiscernible, as if they shared the same shadow that extends itself across their different nature and constantly accompanies them" (p. 218). But as concepts and art are defined in "almost the same terms," they belong to different "planes" (of immanence/composition) – although these planes can slide into one another – and their becomings are not quite the same." Conceptual becoming is heterogeneity grasped in an absolute form; sensory becoming is otherness caught in a matter of expression" (Ibid, p. 177). Consequently, their types of virtual structure differ accordingly: "Beings of sensations are varieties, just as the concept's beings are variations [...]" (Ibid, p. 175). The precepts and affects of art can also be seen as the (needed) non-philosophy of philosophy, which is always both inside and outside of concepts – and sometimes they converge, especially through style (Bogue, Deleuze on Music, Painting and the Arts, p. 191-3).
 Deleuze: "How does actualization occur in things themselves? [---] Beneath the actual qualities and extensities, species and parts, there are spatio-temporal dynamisms. These are the actualizing, differenciating agencies" (DR, 266). And although Deleuze here refers to other types of concepts:"[w]hat remains outside the concept refers more profoundly to what is inside the Idea" (Ibid, p. 273). "A concept alone is completely incapable of specifying or dividing itself [...]" (Ibid, p. 270).
 "Art and philosophy converge at this point: the constitution of an earth and a people that are lacking as the correlate of creation" (WP, p. 108; see also, p. 32-1). Regarding cinematic art, see TI, p. 217. This view of the concept might be somewhat accentuated by what Eric Alliez has identified as a change from Deleuze's earlier affirmation of structuralism, which occurs from the crack in the third series in Logic of Sense and more fully from Anti-Oidepus and onwards. The concept gets a more intense, physical and direct relation to the world, which it can now itself directly affect, or in other words "to invest the created from the point of view of creation" (Eric Alliez, The BwO Condition or, The Politics of Sensation & Anti-Oedipus Thirty Years On). However, Alliez's main argument which is that Deleuze will break with his earlier affirmation of structuralism seems more open for discussion. What kind of structuralism? Although there seems to have been a change (at least concerning the idea of depth, which Deleuze himself openly states ("Nomadic Thought," Desert Islands and Other Texts, p. 261), Alliez downplays the fact that the "structuralism" of earlier Deleuze followed a different, and more ontologically inclusive, logic of sense than language centered "structuralism" (DR, p. 255-6, 232-4), and that it could perhaps be more of an example of his "free indirect discourse." The change in Deleuze's thinking here seems to mainly concern the affective potency in the concept and its relation to the world, rather than the world itself. And wherever the change in Deleuze's thinking lies, it does not in any event mean that virtual multiplicities of the world suddenly became thought of as no longer corresponding with strands of modern mathematics and physics (A Thousand Plateaus, p. 506).
 TI, 146. "[T]ruth is always created" (PS, p. 147). "[A]lways a truth of time [...]" (Ibid, p. 17; see also Negotiations, p. 126).
 Although, on the level of the virtual itself, the different multiplicities are not distinguished from each other clearly like essences co-existing side by side, but has relations that are "indiscernible". They constitute "complexes of co-existence" (DR, 235-6). While still being distinct in nature, they, as Manuel DeLanda writes, relate to each other as "meshed together into a continuum" (plane of consistency). In this way they create "zones of indiscernability where they blend into each other" (Intensive Science and Virtual Philosophy, p. 21). But multiplicity is also formed in the very process between the virtual and the actual. That is, the intensity driving the process of actualization which intertwines the virtual and the actual. There are two types of multiplicity, intensive multiplicities on the virtual plane and extensive multiplicities on the actual, but avoiding a simple dichotomy, "multiplicity is precisely what happens in between the two" (WP, p. 152). This "in between" underlines how the indiscernability of becoming between the virtual/actual forms a type of multiplicity in itself.
 It should be stated, however, that this has nothing to do with technological determinism. In Deleuze, an aesthetic or an image of thought (or a regime or a dispositif), in some sense or another, precede mechanical technologies (TI, p. 267; see also Deleuze & Guattari, Anti-Oedipus, London, New York: Continuum, 2004, p. 34"). The point here is that technologies nonetheless can play a part in providing new means and possibilities that did not exist before.
 Gregory Flaxman, The Brain Is the Screen: Deleuze and the Philosophy of Cinema, p. 34.
 Deleuze himself, of course, insists that that the battle against cliché has to entail that time-images not only constitute intensive forces or a break with the sensory-motor, but that they become readable thought-images (TI, p. 23; Bogue, Deleuze on the Cinema, p. 111). The movement-images do have their own "noosigns." But it is only with the time-image's noosigns, where the thought-image becomes readable (lectosigns) in a new sense, that cinema really starts to constitute thought and even hold "concepts."
 Since this mainly concerns the virtual Ideas of organical and non-organical matter, I am here excluding large portions of very complex discussions in these chapters on various interrelating subjects concerning difference and intensity. (In the fourth chapter, the discussion of virtual Ideas and different/ciation also pertains to the workings of virtual Ideas in thought, linguistic structures etc.)
Alliez, Eric, The BwO
Condition or, The Politics of Sensation,
—. Anti-Oedipus Thirty Years On, Multitudes, 2004
Ansell Pearson, Keith,
"Deleuze on the Overcoming of Memory", in S. Radstone & B. Schwarz (eds.), Mapping
Memory, Fordham University Press,
Beistegui, Miguel de, Truth & Genesis: Philosophy as Differential Ontology, Bloomington : Indiana University Press, 2004.
Bergson, Henri, Matter and Memory, New York : Cosimo, 2007.
Bogue, Ronald, Deleuze on cinema, New York : Routledge, 2003.
—. Deleuze on Music, Painting and the Arts, London : Routledge, 2003.
DeLanda, Manuel, Intensive Science and Virtual Philosophy, New York : Continuum, 2007.
—. Deleuzian Interrogations: A Conversation with Manuel DeLanda, John Protevi and Torkild Thanem, http://www.dif-ferance.org/Delanda-Protevi.pdf
Deleuze, Gilles, Bergsonism , New York : Urzone, Inc., 1988.
—. Cinema 1: The Movement-Image, Minneapolis : University of Minnesota Press, 2003.
—. Cinema 2: The Time-Image, Minneapolis : University of Minnesota Press, 2003.
—. Desert Island and Other Texts: 1953-1974, New York : Semiotext(e), 2004.
—. Difference and Repetition, London, New York : Continuum, 2004.
—. The Logic of Sense, London, New York : Continuum, 2003.
—. Negotiations, New York : Colombia University Press, 1995.
—. Proust & Signs, Minneapolis : University of Minnesota Press, 2000.
Deleuze, Gilles & Félix Guattari, Anti-Oedipus, London: Continuum, 2004.
—. A Thousand Plateaus: Capitalism and Schizophrenia, Minneapolis: University of Minnesota Press, 1987.
—. What Is Philosophy?, New York : Columbia University Press, 1994.
Flaxman, Gregory, The Brain Is the Screen: Deleuze and the philosophy of cinema, Minneapolis : University of Minnesota Press, 2000.
Marks, John, Deleuze and Science, Paragraph: A Journal of Modern Critical Theory, vol. 29, number 2, Eastbourne : Edinburgh University Press, 2006.
Rodowick, D.N., Gilles Deleuze's Time Machine, Durham, N.C. : Duke Univ. Press, 1997.