Image of Thought in Proust, or The First Literary Machine (ca. 1964)

I began by proposing the modern relationship between philosophy and nonphilosophy as the relation between special kinds of machines that we have found to be incorporated into philosophical expression. According to Deleuze and Guattari, this machinic definition should be understood quite literally, particularly in defining the machines created by modern art and in determining how they work. In the 1970 edition of Proust and les signes, where he adds the chapter on “La Machine littéraire,” Deleuze writes that “the modern work of art is anything it may seem; it is even its very property of being whatever we like, from the moment it works: the modern work of art is a machine and functions as such.”1 What is painting, for example, but a special kind of machine for producing perception? What is called literature, but a generic term used to designate different kinds of writing machines, whether we are talking about a novel or a poem, or the specific machines invented by writers? Because of the sheer variety, “Literature” was invented only out of convenience to provide a generically definable order of machines made or invented for producing words and phrases according to a strictly determined logic, and employing a variety of forms and technical procedures. Likewise, we might say the same thing about reading, which is literally machinic: the eye follows the hand in tracing words across the page (left to right, or right to left depending on the recording apparatus that controls the reader’s gaze), unlocking signs that are not visible at first and yet are indistinguishable from the material signs that appear on the page. This mixture of visible and invisible signs, in turn, produces compounds that, in turn, produce a variety of effects like small explosions: images (of understanding, feeling, perception, judgment) that occur somewhere internal to the apparatus of reading machine, which are often described as taking place in the subject of the reader, even though this is a very imprecise way of describing things because the exact nature of the relationship between the subject who is reading and these little explosive events and images is unknown, particularly with regard to cause and effect, which is why the ultimate determination of what these images actually produce is so wildly inconsistent. Sometimes, in scanning the page, in discerning the visible and invisible signs, only certain compounds detonate; others fail to combine successfully. This produces a different outcome, a different “meaning”—again, a vague notion in determining the outcome of this process, much less in understanding what is actually produced or how this machine actually functions.

It is not by accident that cinema becomes the modern art form par excellence, since it externalizes a machinic relationship that remained virtual in painting and literature, making this relationship of movement internal to the apparatus of the camera-projector and incorporating the subjective form of the viewer as a purely virtual surface for recording the image. Nevertheless, I would not say, as others have, that this represents a synthesis; nor would I say, as still others have often said, that there is any teleological progression, since both literature (or the history of writing machines in general) and the artificial machines of the visual and plastic arts continue to remain autonomous and will evolve technically on their own—I would even say, more rapidly than before the invention of modern cinema, especially when one looks at how these machines have evolved over the last thirty years or so with the introduction of new digital technologies. Of course, the camera-projector apparatus does take over some the functions provided by the reader described above: it develops the image, causing it to explode autonomously and without the need of a kinesthetic process attached to the apparatus. Consequently, the subjective is reduced to occupying a purely virtual point of view that is projected in front of the image, “between the actual screen and the virtual brain.” This is the point where all of the images are targeted (as one also says in modern weaponry) in the process of being assembled, a machinic process that is echoed in Deleuze’s own account of the passage from the movement-image to the time-image in postwar cinema. With the gradual discovery of the techniques of framing and montage, the kinesthetic and affective possibilities of subjective variation are further refined and perfected, in effect, producing more possibilities for feeling and new subjective qualities than before, as if subjecting the receptive apparatus of the viewer to an intensive sensory training, or an “education of the senses” (Bildung).2 Thus, in a certain sense the cinematic machine requires new organs to function, even though this requirement is often mistaken for the formal and technical evolution of the movement-image (for example, from celluloid to digitally based media).

This highlights the differences between painting and cinema, particularly with regard to the subjective role of the viewer. Like the reader, the viewer of the painting introduces a wildly unpredictable moment into the total workings of the painting machine, which is one reason this machine has been vulnerable to so many breakdowns precisely from this perspective, which gradually constituted the viewer in a paranoid relationship that often determines the reception of the image in modern art. Consequently, contemporary art does not take what is merely demonstrable for its limit, but constantly seeks a presentiment of other conditions and new materials, and finer sensibilities of reception in its spectators or “public”; even though this has led to an increasingly “privatized public” composed of the artists themselves, as well as other art professionals or specialists (the critic, the curator, the dealer, the corporate consumer). In cinema, on the other hand, much less is left to the spectator with regard to the evolution of the art form itself. The director has greater control over the whole range of subjective responses to the image he or she produces, first of all because the viewer in cinema is trained to sit and quietly take the image in, to internalize the train of images in a more or less passive attitude, which is why cinema and video have become the favored art form of late-capitalist commodity culture. There is little time for thinking or random and highly subjective associations to disrupt this process; the viewer is “too busy” parrying the shocks, anticipating the next explosion, anxiously following the image as it thickens or dissipates, chasing after the image like a dog chasing a fly in the garden. Thinking occurs only at the end, if ever, and even then it takes the form of judgment. It is not like a book, which the reader can put down precisely at the moment where the image begins to unfold, preferring to defer the revelation and savor the promise of intensity; or a painting, which one can walk away from suddenly, and then choose to revisit years later, as if deciding one is now ready for the experience.

The difference between these two types of audiences might be characterized as the difference between learning and education. In viewing a painting, even upon repeated viewings, the viewer either learns to see the image or completely fails in seeing it. In cinema, the viewer is educated, she is forced to see whether or not she wants to, and what she sees is determined by whether she is a good or bad student. She can easily reject the entire experience, in which case the duration of the film becomes a “whole waste of time,” even though it is already too late for judgment at this point, since the experience now fatally belongs to the spectator’s life. Here we might ask: How much of our lives has been wasted on bad films and media productions, and is there some essential relationship between the commodity forms of late-capitalist culture and the quantity of wasted time that seems to increase exponentially with the invention of newer technologies and digital media? Moreover, at what point can we be led to say that “a life,” like a bad film, can also be subjected to this kind of judgment today—“the whole thing was simply a waste of time”? These will be questions that I will reserve for our discussion of Proust, and later on, concerning the relationship between modern cinema and the brain.

For now, let’s return to the question concerning the “presupposition of an image of thought.” As I said earlier, in place of presupposing the machines of natural perception and common sense (i.e., representation), modern philosophy begins to presuppose a plane of immanence that is both too near for intuition and too distant for external perception. More provocatively, in this new situation modern philosophy suddenly awakens to discover it now has “no eyes, no ears, and no mouth.” Therefore, in order to orient itself on this new plane, it must acquire different machines that will replace the organs of natural perception. Thus, the painting machine gives eyes to modern philosophy to see with; literary machines give it ears and a mouth. What does the cinematic machine provide? New compounds and a new series of images that will either allow philosophy to see farther and, at the same time, to draw nearer to inner experience; or negatively, provide a whole new set of transcendental illusions with which to contend. Is this situation any different from the case of modern science, which requires more refined technical machines to grasp the reality of the universe or the molecular composition of a living cell? In other words, perhaps the plan of immanence presupposed by modern philosophy requires no less than the modern sciences require: newer and more refined machines of perception, intuition, imagination, and expression. Of course, this is not necessarily perceived as progress at first, because the newer machines are difficult to get set up and working properly. The machine that replaces the mouth often stutters (like in a poem by Cummings or Celan), or utters purely paratactic speech (as in Artaud and Carrol), and even sometimes issues the unbearable cry of animals or sounds that are like dry leaves scraping against an open doorway (as in Kafka). The artificial eye machine produces visions that are blurred and fuzzy; objects are difficult to make out, because they are “certainly not clear and distinct,” and there is a zone of indetermination between background and the figure (as in a painting by Cézanne, whom Deleuze often privileges). The soft machines of cinema produce hallucinatory visions, uninhabitable and frozen landscapes, inhuman visages, new organs and images mixed with indistinguishable blocks of sensation and feeling (as in the cinema of Cronenberg), unbearable durations in which nothing happens (as in Tarkovsky). What kind of eyes and ears are these? What kind of mouth (as Artaud described it, “no teeth, no tongue, no larynx”)? Nevertheless, it is apparent that philosophy for some time has preferred these to natural organs, and even the most common language does not make much sense any more—not because it lacks understanding, but simply because it no longer has any ears! Perhaps, in this manner, we can understand why modern philosophy can be defined precisely as “a body without organs.” In fact, there may not even be the requirement of a subject, in the traditional sense.

How does thinking take place in this new situation? Thinking is not a machine, but rather is a subjective determination of an event that happens (or more often than not, actually fails to happen!) owed to the functioning of these new machines of perception, imagination, and feeling. Philosophy, on the other hand, is a machine that produces concepts from the raw materials provided by these other machines that it includes as functions of its total mechanism; if thinking is to occur at all, even if only rarely, then it occurs in the interrelationship between these different machines, something that is produced by this interrelationship, something that either works or does not work according to this interrelationship. Consequently, thinking does not always happen; the machinery breaks down and fails to produce the desired event. As Deleuze says in Difference and Repetition, in order to truly begin thinking, we must first learn how to forget what thinking is, and only in this way can we actually be receptive to the event of thinking within thought.3 In other words, modern philosophy begins to think only from a situation of stupidity—this will forecast the importance of Artaud’s example for Deleuze, which I will return to discuss below—a fundamental stupidity that, in turn, will strike against philosophy’s great ideas (the Self, the World, God), which Kant had already revealed to be transcendental illusions of Reason.

Once more, Heidegger provides us with a good illustration of this situation of fundamental stupidity that strikes against the innermost possibility (potentia) of thinking. The fact that many of his later works carry as their titles questions that could just as well be asked by children or idiots (e.g., Was ist das—die Philosophie?) should already indicate that the philosopher does not begin from a situation of knowledge. In one of his last works written in the period of the early 1950s, What Is Called Thinking?, Heidegger sets off to track down the image of thought in much the same way as Proust sets off in search of lost time. What results is a hilarious method of searching for an original image of thought, which ends with the philosopher contemplating what a child means when she says “Bow Wow Bad Bite” alongside the somewhat paratactic utterances of the pre-Socratics: “Needful: the saying also thinking too: being: to be.”4 If the comedy of this moment often goes unnoticed, it is because Heidegger’s philosophy continues to be read under a sanctimonious light (i.e., “the piety of Thinking”) in which the most childlike utterances are mistaken for sayings of an oracle, somewhat like the character of Chance Gardner from Jerzy Kosinski’s Being There. However, what if we considered the opposite? What if Heidegger was really attempting to learn how to speak as a child, to forget all previous language and signification, to become an apprentice in the things and concepts? This would not be a situation of an enforced naivete, or as in Descartes, an initial point of certainty arrived at through a rational method of doubt. This would be an altogether different image, which is why stupidity might be a better concept for determining what is presupposed in Heidegger’s final image of thought. Does this not illustrate a situation where the philosopher exposes himself to becoming stupid in order to learn again the signs that lead to thinking, almost in the sense of seeking to become an embryonic consciousness in order to experience the real sense of words?5

Around the question of an apprenticeship in signs to produce an original image of thought not presupposed by philosophy, in the traditional sense, we must now turn to Deleuze’s early reading of Proust, which exemplifies what I will call “the first literary machine.” The first edition of this study appears in 1964, under the title Marcel Proust et les signes, and concludes with the chapter called “The Image of Thought.” In light of the above discussion, however, it is important to note that the second part of this study, which is entitled the “Antilogos or the Literary Machine,” is not added until the 1970 French edition, which is around the time when Deleuze is just beginning to collaborate with Guattari on Anti-Oedipus.6 Consequently, it is Guattari’s concept of “machine,” which Deleuze first encounters reading Guattari’s “Machine et structure” in early summer of 1969, that becomes the fulcrum of his amended study of Proust’s Recherche that same year.7 It is important to point out that in 1964 Deleuze already found in Proust a means of what he called “anti-philosophy”—in this case, a means of posing a new concept of the idea without having to presuppose Plato. Again, this is a manner of forgetting what was already known beforehand in order to discover what remains to be thought, which must be achieved by a strict method of searching, and the image of reminiscence and memory offered by Proust serves as an intermediary and a screen that “filters out” the Platonic theory of ideas. But another change takes place in the passage from the 1964 to the 1970 editions, and this has to do with precise meaning of “the literary machine.”

First, let me quote different parts of the two editions. The first part is from the concluding section, “The Image of Thought,” from the 1964 edition:

As we have seen, this distinction between Proust and Platonism involved many more differences. There is no Logos; there are only hieroglyphs. To think is therefore to interpret is therefore to translate. The essences are at once the thing to be translated and the translation itself, the sign and the meaning. They are involved in the sign in order to force us to think; they develop in the meaning in order to be necessarily conceived.8

Now the passage that takes up the same subject in the 1970 edition:

We have now seen how Proust revived Platonic equivalence of creating/remembering. But this is because memory and creation are no more than two aspects of the same production—“interpreting,” “deciphering,” and “translating” being here the process of production itself. It is because the work of art is a form of production that it does not raise a special problem of meaning, but rather of use. Even the activity of thinking must be produced within thought.9

In the second passage, which can be understood as a gloss on the first passage, we note the appearance of the term “production,” which of course becomes a key term of Deleuze and Guattari’s major thesis in Anti-Oedipus. Thus, “all activity of thinking must be produced within thought.” Moreover, the activities of “interpreting, deciphering, and translating” are no longer related to the “sign,” and the term “hieroglyph” completely drops out of the second version and is replaced by the concept of the machine that produces . . . but produces what? Simply put, in this case, Deleuze is referring to a specific machine that produces memory, which is the Proustian literary machine. It is around this point that Deleuze makes a surprising discovery of what was merely intuited in his earlier work: to remember is to produce, just as to create is to produce; memory and creation are two aspects of the same process of production.10 Therefore, the problem of memory is no longer attributed to a sign that preexists the act of creation, it is the act of creation itself, and is attributed to a machine that produces a “spiritual equivalent” between memory and impression that defines the work of art. As Deleuze writes:

This is because the meaning (truth) is never in the impression nor even in memory, but is identified with the “spiritual equivalent” of the memory or of the impression produced by the involuntary machine of interpretation. It is this notion of a spiritual equivalent that establishes a new link between remembering and creating and establishes it in a process of production as a work of art.11

In this final passage we might confirm what was only a vague and indeterminate idea earlier on: the manner in which, in Proust’s work of art, “philosophy vies with nonphilosophy” (i.e., the Proustian image of memory vies with the Platonic theory of ideas), because it is only through the writing machine invented by Proust that philosophy obtains a better understanding of the function of the idea in reminiscence, that is, a better concept of how past is produced in relation to the sensible present in a manner that is essentially productive or creative. Yet, if the thinking subject must presuppose a plane of consistency in the form of a presupposed image of thought, is this plane the same for philosophy as it is for literature, or the writing machine? How does one distinguish between the ideas of memory and the ideas of thought, if it is true that in each case ideas do not preexist the impressions that are developed into signs? Is thinking just another species of remembering, like the material form of memory expressed by habit? Moreover, what does it mean to say that the idea of thinking must be produced, like memory, by something that functions as the spiritual equivalent of memory in matter? As Proust writes, “even memory, still too material, needs a spiritual equivalent,” in the sense that even memory needs a kind of machine to produce it.12 It is around this last question, in particular, that Deleuze finds in Proust, who is described as a pure subject of writing, a different presupposition operating at the basis of the work, and thus a different image of the body without organs.13 Therefore, a “body without organs” (BWO) is an anorganic, but nevertheless living, body that is fabricated or constructed to replace the image of the body presupposed by the empirical Ego, or Self. In this sense, all literature is replete with different kinds of bodies without organs.

But what does it mean to say that the subject who appears in the position of the narrator of the Recherche has no eyes, no ears, no mouth? First, it means he is neither completely naive nor stupid, but instead is “innocent of all experience,” or as Deleuze says, is probably quite mad. In a 1975 “Roundtable on Proust” that was later incorporated as the last chapter (“Presence and the Function of Madness”) of the 1976 edition of Proust and Signs, Deleuze confesses to having an impression that has only “very recently occurred to him” (and here I am only speculating that he means by this that this idea has occurred to him since the 1970 edition, which the audience would no doubt have already read):

I have the impression that there is in this book a very important, very troubling presence of madness. This does not mean that Proust was mad, of course, but that in the Search itself there is a very vivid, very widespread presence of madness. . . . Everyone knows who the ringleader is: the narrator. How is the narrator mad? He is a very bizarre narrator. Totally bizarre! He has no organs, he can’t see, he does not understand anything, he does not observe anything, he knows nothing; when something is shown to him, he looks but does not see; when someone makes him feel something, they say: but look how beautiful this is, he looks and when someone says: here, take a look—something echoes in his head, he thinks of something else, something that interests him, something that is not on the level of perception, not on the level of intellection. He has no organs, no sensations, no perceptions: he is nothing. He is like a naked body, a vast undifferentiated body.14

It is precisely the combination of innocence and madness that is ascribed to the narrator’s body, which feels impressions, memories, encounters, perceptions, and thoughts—but cannot understand them at first. It is safe to say that, whether due to madness or to innocence, the narrator is a being who exists without presuppositions. He presupposes absolutely nothing—not even his own sex!—not even the vaguest idea emerges at the basis of each impression or thought. Therefore, the narrator does not understand what he sees, what he feels, what is said by the characters, and this constitutes the basis of the search that Deleuze describes in his work as an apprenticeship of signs, since it is by developing and interpreting these impressions and signals that the Proustian experience of the world is woven together in the great web-body of signs.

Therefore, if it is not Marcel Proust who is mad, and the narrator is not made to be organically a representation of Proust, then there is the presence of something else, a presence of madness that is distributed across the characters of the Recherche, something that moreover cannot simply be reduced to language even though it is also something that cannot exist apart from language. Deleuze describes the subject of the narrator according to two different images: he is a spider; he is a vegetable. As a spider, he casts a web and perceives things only by signals, small vibrations. In another place, he is described as a kind of vegetal consciousness, which perceives or feels only what constitutes its immanent environment. “An orchid presents an image of the wasp drawn on its flower, with its antennae, and the insect comes to fertilize this image.”15 This is the manner by which the flower perceives the wasp that has no resemblance to empirical conscious, to the cogito, that Deleuze finds in Proust’s method, which he in turn seeks to apply to Philosophy as a means of liberating it from the traditional image of thought.

Here, in both images we are presented an image of the subject of writing, of a text that is composed by weaving together signs, in other words, by a language of flowers or grass. Deleuze describes this spider or plant consciousness that he finds immanent to every moment of the text, constituting the form of presence of the narrator, the presence of another consciousness that must be described not as the consciousness of the empirical self but instead as a consciousness that can exist only in writing, in literature. To call it madness is only to highlight the strangeness and otherness with regard to the empirical consciousness. What is the consciousness found in literature but something that is partly organic, but also partly vegetal? In other words, how does perception take place in the process of writing? When the narrator describes a scene or event, we often represent these signs as perception in analogy to the common senses. But this is not how they are made. Rather, they are woven into a text constituting a body that first captures these impressions and converts them into signs, a body that also senses and feels but is not organically composed.

It is absolutely a prejudice to ascribe the presence of conscious perception that appears in writing to empirical consciousness, which would be madness. This is a very simple point, but one that is often overlooked in the history of criticism. In Kafka, most famously in “The Metamorphosis,” Gregor Samsa is described from the perspective of waking up to discover that he is a beetle. But who is this “I” doing the describing, that is actually floating somewhere in Gregor’s room and observing the events from different vantage points, including Gregor’s own thoughts and physical experiences of pain. For example, Gregor suddenly falls from the bed to the floor. The narrator reports: “There was a loud thump, but it wasn’t really a loud noise.”16 From which perspective might we understand this statement? Literary criticism has been drawn to describe this perspective occupied by the written narrator from different points of view (omniscient, partial, subjective), but in no way can this consciousness be likened to the empirical self. To understand this fact, however, we must learn to become a little stupid, forgetful of our habits and prejudices, or even a little mad like Proust’s narrator. Yet the fact that this form of consciousness exists as a definite form of possible experience constituted by signs means that writing offers another possibility for human existence that is not possible for the simple empirical self, a means of occupying both objective and subjective points of view, another manner of perceiving, feeling, experiencing, and thinking.

The problem that Proust poses for us to consider is that the narrative consciousness is named in relationship to the empirical subject of Proust himself, as Marcel, and yet this would have to be another Marcel that is different from Proust, the author, one that offers Proust a method to distance himself from his own experience by passing through this strange consciousness born of writing. For example, every time I read a writer, and by this I mean when I read all their published works, their letters and diaries, their little notes to editors and lawyers, their grocery lists, I have the feeling that I am becoming a part of their consciousness, that I am inhabiting them and seeing things as if from inside their subjective perspective. I have the feeling of a familiarity that is bought by years of research. But this is just an illusion. Deleuze appears to offer a more accurate picture of this experience, which becomes the basis for the interpretation of the presence of the narrator in Proust and Signs: I am caught in their web, trapped in the intricate lacework of their signs. The only approximate representation of knowing their consciousness is the image of a spider dragging its heavy body toward me and in one terrifying moment wrapping me in its abdominal cord and drinking all my blood. This is a better representation of the experience of literature, or what a literary machine produces in me, which entails some kind of madness and obsession with a completely other consciousness, which is more likened to the consciousness of a spider or a plant.

Before turning to describe the components of the Proustian literary machine (what traditional literary critics usually call “the Proustian experience,” but this is strictly an anthropomorphism), let us briefly turn to the conclusion of part one, written at some point between 1962 and 1964, where we find one of the first mentions of “the image of thought” in Deleuze’s philosophy. Here we are given a more precise understanding of the problem of presupposition in philosophy, and why modern philosophy can no longer presuppose an image of thought that appears in natural consciousness. “Proust’s work,” Deleuze writes, “vies with philosophy. It sets up an image of thought in opposition to that of philosophy. He attacks what is most essential in a classical philosophy of a rationalist type: the presupposition of this philosophy. The philosopher readily presupposes the mind as mind, the thinker as thinker, and the subject as a being who wants the truth, and who naturally seeks the truth in consciousness. He assumes in advance the goodwill of thinking; all his investigation is based on a “premeditated decision.”17

In this passage Deleuze is referring to what he will later define, in the chapter “Image of Thought” in Difference and Repetition, as the “subjective presuppositions,” “under the double aspect of a goodwill on the part of the thinker and the upright nature of thought itself.”18 In the later chapter, Deleuze rigorously shows that this subjective presupposition, itself a form of prejudice, is implicitly borrowed from the idea of a common sense (cogitatio natura universalis). In other words, the classical philosopher sets off in search of something that is already presumed to be in agreement with his foreknowledge of it, something that is already accorded a nature that is disposable to being discovered or revealed, and, as Deleuze says, something that the classical philosopher claims to belong to him by a kind of right. Classical philosophy calls this something = x, “truth,” which it sets off in search of by means of a “premeditated decision” (or method). But this would be a lot like setting out on a search for buried treasure armed with the foreknowledge of its gradual discovery, if only because you already have an idea of where it is hidden because you first buried it there, and you drew a little map for yourself as a reminder. In both these chapters, of course, Descartes will constitute the prototype of the classical philosopher of the rationalist type who sets out on a search for what is called “Truth” by means of a premeditated decision. If we remember from the opening of the Meditations, this trip is planned out well ahead of time: he waited until he was mature enough to take the trip, being in “secure possession of leisure and in a peaceful retirement,” and even sets aside a full week (except Sunday, of course) to apply himself freely to “the overthrow of all his formerly held opinions.”19

What is wrong with this image? On the first day, the narrator of the Meditations, the cogito, is quickly able to strip away all “objective presuppositions”—and here it is important to point out that the philosophical cogito as a Body without Organs is first given by Descartes: it has no natural organs of perception (no eyes, no ears), no external limbs or body at all, no feelings other than those that are necessary for the search itself; no memory, and even no extension (because even the earth and the sky are vanished as fictions created by the mind). However, on the second day of the meditations, after getting a good night’s sleep having been exhausted by the previous day’s expeditions, Descartes’s narrator suddenly awakens to find that it is deep water, because it must now even doubt its own nature as a human being, or an animal rationale.

I suppose, accordingly, that all the things, which I see are false (fictitious); I believe that none of the objects which my fallacious memory represents ever existed; I suppose that I possess no senses; I believe that the body, figure, extension, motion, and place are merely fictions of the mind. What is there, then, that can be esteemed true? Perhaps this only, that there is absolutely nothing certain.20

It is here that the “Body without Organs” historically first appears as the being who says “I think,” but whose form of existence can never be proven and so must be submitted to a radical epokhē (as phenomenology will later label this method of suspending all objective presuppositions in order to finally begin as a purified cogito, that is, as a firm and immovable point from which thinking begins). To be fair to Descartes, everything cannot be said to have been predetermined in advance. He did not know that on the morning of the second day he would find himself in such deep water and suddenly have to deny his own human nature, all of his past memories, all sensation and feeling, all capacity to move around and occupy a place on the earth. Even the initial conditions of his search, his leisure and his peaceable retirement, seem absent on this morning. And yet, as Deleuze will show, this search was already betrayed from the start, even before it began.

Why? First of all, because one does not seek the truth only under the best conditions possible and in a pleasant state of retirement from the activities of the world (from all labor, need, desire, war, or natural disasters); there are other situations, more frequent, that are not so amenable, as when one is forced into thinking by something that is immanent and yet cannot be altogether thought. This might provide a better occasion for the event that causes thinking to occur; as Deleuze writes, “The act of thinking does not proceed from a simple natural possibility.”21 Secondly, Descartes is capable of removing all the objective presuppositions from thought, but even before he begins he does not place in question his own subjective presuppositions that thinking is by nature good and that the good nature of the thinker desires truth, even though he raises these subjective presuppositions at the end of the first day when he realizes that the path laid out for thinking is “arduous” and that the natural cogito so resists being filled with a feeling of indolence that it almost leads the narrator back to the ordinary course of life, to his leisure and his peaceable retirement: “just as the captive who perchance was enjoying in his dreams an imaginary liberty, when he begins to suspect that it was but a vision, dreads awakening, and [out of fear] conspires with the agreeable illusion that the deception may be prolonged.”22 In fact, Deleuze will highlight the nature of this fear and the conspiracy with illusion as the real subjective condition of thinking that must be presupposed in place of the essential good nature of the thinker and the upright nature of thought. Here we are given the essential image of stupidity and ill will as the preconditions of a modern image of thought:

Not an individual endowed with good will and a natural capacity for thought, but an individual full of ill will, one who does not manage to think, either naturally or conceptually. Only such an individual is without presuppositions. Only such an individual effectively begins and effectively repeats.23

Finally, although certainty, or the desire for certainty, is a representation of the truth under the conditions that everything can be doubted, it is already secondary to what causes us to seek the truth in the first place; moreover, it distorts the nature of this causality by already representing it as something that will extinguish this movement: it conceals or distorts the real causality of thinking by providing thought with an image of its end or its goal. Consequently it is the poet and the writer, and not the philosopher, who learns, rather than understands, that thought is nothing without something that forces and does violence to it, and this force will not be revealed as either uncertainty or simple doubt. Therefore, if we replace the certainty of truth with another object, the past, and another kind of certainty, real experience, then we have an idea of the Proustian search that Deleuze will say “vies with philosophy.” Here, the Proustian narrator still seeks the truth, but the truth is not represented in advance as an object that wants to be found. The search for “time lost” does not begin in such a premeditated fashion, nor is there any preguarantee that time will be found; most importantly, the only presupposition is the fact of having lost it—it has not purposefully been buried only to dig it up again later on, because it is not simple recollection or memory—and the forcefulness of this loss itself is the causality that sets the Proustian narrator on a search, which begins the adventure of thinking by tracing the signs and developing them.

Perhaps no better image of the conditions of the search can be provided than the following long passage from the third volume of the Recherche, which Deleuze quotes in its entirety as if to underline that it needs no further philosophical interpretation:

The truths that intelligence grasps directly in the open light of day have something less profound, less necessary about them than those that life has communicated to us in spite of ourselves in an impression, a material impression because it has reached us through our senses, but whose spirit we can extract. . . . I would have to interpret the sensations as the signs of so many laws and ideas, by attempting to think, that is, to bring out of the darkness what I had felt, and convert it into a spiritual equivalent, . . . Whether this was a matter of reminiscences of the kind that included the noise of the fork or the taste of Madeleine, or of those truths written with the help of figures whose meaning I was trying to discover in my mind, where, like steeples or weeds, they composed a complicated and elaborate herbal, their first character was that I was not free to choose them, that they were given to me as they were. And I felt that this must be the mark of their authenticity. I had not gone looking for the two cobblestones in the courtyard where I stumbled. But precisely the fortuitous, inevitable, way in which sensation had been encountered governed the truth of the past that it resuscitated, of the images that it released, because we feel its effort to rise toward the light, because we feel the reality of the joy of reality regained. . . . In order to read the inner book of these unknown signs . . . no one could help me by any rules, such reading consisting in an act of creation in which nothing can take our place or even collaborate with us. . . . The ideas formed by pure intelligence have only a logical truth, a possible truth, their choice is arbitrary. The book whose characters are figured, not traced by us, is our only book. Not that the ideas we form cannot be logically exact, but we do not know whether they are true. Only the impression, however paltry its substance seems, however unlikely its traces, is a criterion of truths, and on this account alone merits being apprehended by the mind, for only the impression is capable, if the mind can disengage the truth from it, of leading the mind to a greater perfection and of giving it a pure joy.24

It is important to notice the words and phrases that Deleuze emphasizes in reading this passage: necessary—in spite of ourselves—signs—herbal—I had not gone looking—fortuitous—inevitable—encountered—not traced by us. This is the same kind of paratactic utterance that appeared earlier in Heidegger, but we can take it as indicating the formula of Proust’s method, which is complicated and herbal (vegetable) and is composed of only those impressions and signs that his narrator “had not gone looking for,” and therefore is not created to include all impressions and signs. Hence, the book or the literary machine, as “an involuntary machine of interpretation,” will be a filter of sorts that first determines what impressions and signs are necessary to include and those that are merely possible and intelligible, and thus are not included—for example, the “two cobblestones” where he tripped are included because they are necessary and not merely possible perception, and constitute the signs of a real encounter or accident. The sign is created as the “spiritual equivalent” of the idea of the Past, beginning from impressions and developing these into signs that are “immanent” to the Past—not simply to the past of memory, but to the force of the past that causes each present to swerve toward it and often to become lost. Thus, each sign produced (remembered/created) by the machine will present what Deleuze later calls a “crystal image” of the moment when time is split into two unequal parts: one part rushing headlong toward the past where it becomes lost in relation to the memory; another part surging forward in relation to the past it will become in the sudden discovery or revelation of its impression that was hidden in some “material object.”

And so it is with our own past. It is a labor in vain to attempt to recapture it: all the efforts of our intellect must prove futile. The past is hidden somewhere outside the realm, beyond the reach [in some material object] . . . which we do not suspect [i.e., “had not gone looking for”]. And as for that object, it depends on chance whether we come upon it or not before we ourselves must die.25

The method that Proust devises for his search of this lost time, more a set of presuppositions for how the literary machine will work than a set of rules or axiomatics, can be contrasted to that of Descartes almost point for point. The method would proceed by interpreting the impressions of sensation as signs of laws and ideas; in contrast to Descartes, the writer or artist does not seek to bracket sensation first in order to determine the laws and ideas of the mind as pure cogito. Sensation is first and becomes an impression that is necessary and exists in spite of ourselves; thinking comes afterward, in working over these impressions to discover their law, that is, “to bring out of the darkness what I had felt”: the living being of sensation itself, but which can be really experienced only through an artificial and organic body created by the process of writing. Elsewhere, Proust will define the book, in purely machinic terms, as a telescope for looking into ourselves—however, through its apparatus this telescopic machine would cause the appearance of a sensation, a sensation that would not appear otherwise to natural perception.

To conclude our discussion of Proust’s literary machine, I will briefly turn to the second part of the 1970 French edition, where we are given both the definition of the literary machine as well as an image of the narrator as a Body without Organs that corresponds to what I have already described as the new image of thought: immanence. Here the Body without Organs becomes a being of pure sensation (no eyes, no ears, no memory, and above all, no thought), that is to say, a completely different Body without Organs than the one Descartes devised for philosophy, and therefore a completely different image of thought. In Anti-Oedipus, Deleuze and Guattari define a machine as anything that interrupts a flow. Here the machine (the book, the involuntary machine of interpretation) is a spiderweb—it is a vast and complicated partly animal and partly herbal web that interrupts the flows of signs and impressions that get caught up in it; and the narrator is the spider that drags its heavy body to the place of the interruption and spins by means of its abdominal fluid a cocoon around the impression it finds there (i.e., to develop the impression into a sign) in order to finally drink its blood (extract its essence, its spiritual idea). Is this being malevolent? Is the being of a spider malevolent for drinking the blood of the fly? In this question we find the image of thought divorced or separated from the idea, the prejudice actually, of a moral image of thought. In other words, it is in the Proustian image of thought, rather than the Cartesian, that we find a thinking that is capable of doing away with the subjective preconditions of good and common sense that still retain an anthropological prejudice of the cogitatio natura universalis. This is why Deleuze often refers to the being of the narrator as “in-human”—not in any monstrous sense of horror, but in the same way we would refer to the being of a spider as nonhuman without implying any judgment, except that it was a being of nature.

And yet the Proustian spider-narrator is not a being of nature, but a purely artificial being, a created being. Thus, in the conclusion of the second part, after diagramming the components of the literary machine and describing how it works, Deleuze concludes by defining the nature of the being that is found at its center, like a spider in an intricate web it has created:

The logos is a huge Animal whose parts unite in a whole and are unified under a principle or a leading idea; but the pathos is a vegetal realm consisting of cellular elements that communicate only indirectly . . . the bumblebee that constitutes the communication between flowers and loses its proper animal value becomes in relation to the latter merely a marginalized fragment, a disparate element in the apparatus of vegetal reproduction.26

Partly animal, partly vegetal or herbal, the process of writing and the being of the writer are compared to a bumblebee that loses its own proper animality by becoming part of the reproductive apparatus of the flower. In the same way, we can say that the writer also loses his or her proper human value (i.e., the attributes of a self or an individual) in becoming part of another reproductive apparatus, which is no longer that of the self or the individual, but rather the expression of an impersonal, and in some sense inhuman life. Is this the nature of the being that Deleuze proposes to replace the classical philosopher, who would become the marginalized and disparate element of a vegetal image of thought? In the next chapter, in order to respond to this question, I will continue to explore the nature of this being, and of the new image of thought, in the “rhizome.” Although the rhizome appears like a plant or a weed, in actuality it refers back to this partly vegetal and partly animal writing machine that was first created by Proust.