May 25, 1982
But what does “time-image” mean? This implies for us, once again, [that] time can only release an image from itself, only if it does not hold onto the form of succession which, on the contrary, refers completely to a succession of “movement-images”, that is, a form of succession, even if this form of succession includes accelerations, slows down, displacements, flash backs, etc. So this time-image … would be related to the movement-image, but in such a way that the movement-image, in a way, would be an indicator to designate it and that it would be necessary to pass from the movement-image to this deeper time-image, if we know that we cannot it to reduce to a succession of images. That means for us that, in a certain way, it is indeed “image” itself, it is not a succession of images.
Seminar Introduction
In the first year of Deleuze’s consideration of cinema and philosophy, he develops an alternative to the psychoanalytic and semiological approaches to film studies by drawing from Bergson’s theses on perception and C.S. Peirce’s classification of images and signs. While he devotes this first year predominantly to what he considers to be the primary characteristic of cinema of the first half of the 20th century, the movement-image, he finishes the year by emphasizing the importance of the post-World War II shift toward the domination of the time-image in cinema.
For archival purposes, the English translations are based on the original transcripts from Paris 8, all of which have been revised with reference to the BNF recordings available thanks to Hidenobu Suzuki, and with the generous assistance of Marc Haas.
English Translation

Antonin Artaud
Having ended the previous session on Georges Comtesse’s query about “Last Year in Marienbad”, Deleuze opens the session with Comtesse’s detailed follow-up to Deleuze’s discussion on Robbe-Grillet’s novels in relation to cinema. Then, as he does in the final chapter of The Movement-Image, Deleuze links up the long discussion of the different facets of the movement-image with the second major cinematographic concept, the time-image. This development brings him to the real question, as it were: “Is there a specific relation between cinema-image and thought?” Deleuze considers the two paths developed to answer this, on one hand, the path of the cinematographic Idea and, on the other hand, the suspension of the sensory-motor image and the emergence of the pure sensorial image. In this path, Deleuze outlines four modes of image corresponding to the pure sensorial (imaginary, didactic, critical, and “attractional”, that will subsequently be called temporal), and then he develops six sub-modes of the imaginary. In the process of considering the relation of these modes to the general question of “cinema-thought”, Deleuze develops an important connection with Artaud’s understanding of cinema, particularly through his little-known scenarios. The session concludes with Deleuze’s summary of these four modes linked to various sub-modes and works by certain filmmakers.
Gilles Deleuze
Seminar on Cinema: The Movement-Image
Lecture 20, 25 May 1982
Transcribed by Li-Ting Hung (Part 1) and Yu Yue Xia (Part 2); additional revisions and time stamp, Charles J. Stivale
English translation by Billy Dean Goehring
Part 1
… Second: I don’t know where we’re meeting, since the theater is going to be renovating and setting up this room. I’ll see if I can find out whether I have a room later, during the break; if not, I’ll put a note up on the door next week with the location. It might be a small room, but since it’ll be our last meeting, there will be so few of you that it shouldn’t be a problem. [Laughs. A student says something inaudible.] Anyway, well… [1:00] Yes, most importantly, there’s certainly more to be said, but I have to wrap it up because the theater … [Deleuze makes a noise] No, I wouldn’t dare do anything so obscene. [Pause]
Third, no, we’ll keep going, but before we do—[There’s a noise outside] What’s that racket?—before we do, I believe George Comtesse had an intervention he wanted to make, if he’s here? Are you there? Alright.
Comtesse: I wanted to chime in regarding the four constraints you proposed on the New Novel, your four propositions on the New Novel and their possible correlates in film, however problematic. The four propositions being: first, the optical gaze, a gaze shedding light on the unbearable or intolerable; second, there’s an optical[1] description of an isolated element which is what furnishes this [intolerable] element; third, with Robbe-Grillet you have absolute subjectivity, the main trajectory of his writing; and fourth—a point I won’t get into—concerns how Robbe-Grillet’s distinctive disavowals sometimes line up with societal changes or transformations. I’m more interested in the first three points, [and] more so in the context of Robbe-Grillet’s first four novels than in any of his theoretical writings. Because I don’t think what Robbe-Grillet calls total subjectivity fully bears out in his actual writing.
For example, [Jean] Ricardou once pointed out that there was a discrepancy between Robbe-Grillet’s fiction and his theoretical work. I think those who analyze Robbe-Grillet’s novels, such as Bruce Morrissette, give up on the notion of Robbe-Grillet’s total subjectivity and instead replace it with what [Morrissette] calls the “I” in Robbe-Grillet’s writing, an “absent-I”. The absent-I is completely different—this is my first point—from total subjectivity, because this “absent-I” is filled in neither by a prior substance-to-be-expressed, nor by something which might originate some future creation. The “absent-I” is, on the contrary, an “I” devoid of being, substance, subjectivity, or time. For example, the problem that came up on the subject of time—the conclusion Bruce Morrissette comes to in his examination of Robbe-Grillet’s fiction—is that there might be another temporality, one that doesn’t adhere to a conventional temporal structure, a temporality of repetition—and that’s what we find at work throughout Robbe-Grillet work. That’s my first point.
Second, this “absent-I,” which defies total subjectivity, is the intersection of two things: it’s the intersection of two things. There’s a gaze, but not an optical gaze initially. With Robbe-Grillet, the gaze has more to do with immobility, the weight of silence, the “absent-I.” It’s the intersection of both the gaze… [Someone makes a noise near Deleuze, and he says, “Shh! Shh! I can’t hear!”] I was saying: it’s the gaze, immobility, and the narrative role of silence. In other words, it’s a gaze that narrows in on fragmentary elements, insofar as what a gaze does is isolate elements; the voice itself, the narrative voice, as it lingers on an element, keeps simultaneously both erasing and integrating, and then ultimately dissolves it. Put another way, when it comes to the optical gaze in Robbe-Grillet, even with extremely precise topographical, geometric, architectural descriptions—the purpose of the optical gaze is never to reveal anything, to reveal, say, the intolerable because, in a way, the optical gaze itself is what the novelist finds intolerable. And because the optical gaze doesn’t reveal anything, it isn’t a perspective that uses the “I” to signify space, nor one that uses space to signify an “I,” because the narrative voice is one that constantly empties out whatever might fill in the gaze.
So, contrary to what other literary critics might have said, Robbe-Grillet’s work doesn’t privilege space; it’s not about the literary expression of space. Rather, it’s a “lack” of space, as well as a “lack” of time. In other words, far from it being an optical description which brings things to light, such description turns out to be—in Robbe-Grillet’s New Novel, at least—an endless optional illusion, endless because, in its delusion, the optical gaze always disregards it. And that’s what accounts for its illusion, its illusive mastery; the optical gaze has constant contempt for the gaze, fascinated or dazzled by silent immobility, such as we find, for example, in Robbe-Grillet’s film, like the protagonist with his white, mummified, cadaverous [word unclear] at the start of the film, or in Last Year at Marienbad, when we meet the guy, and he says, “You’re scared, you’re frozen, shut off, absent.”[2]
In other words, it’s through this immobility, this particular kind of “I.” It’s neither mobile in relation to an immobile space, nor is it immobile in relation to a mobile space—precisely because its immobility is the source of the illusion of movement in space, a movement, in other words, which constantly suggests immobility or crushing confinement. It’s why, for example, in The Erasers, in Robbe-Grillet’s first book, freely walking, Wallas’s stroll through the city, a circular walk in space, meaning he’s always headed back to where he started, and that’s because this walk—hence the illusion of movement in space—because this walk attempted to break free from the gaze’s immobility, to bring about a mobile displacement in the immobile space of the lines or rows of brick houses. For example, Robbe-Grillet writes, quite ironically, “[For it is Wallas] who is advancing; it is to his own body that this movement belongs, not to the backcloth some stagehand might be unrolling […] It is of his own free will that he is walking toward an inevitable and perfect future.”[3] And then he multiplies his assurance…
Deleuze: What a quote! Where is it from?
Comtesse: It’s from The Erasers, about his walk.
Deleuze: From The Erasers?
Comtesse: Yes, The Erasers—Wallas’s walk through the city.
Deleuze: Do you have the page numbers?
Comtesse: Ha, no, I don’t remember what page it was on! The movement belongs to his own body. Yet that’s nothing but action that’s been deteriorated by the power of the false. It’s a humorist power of the false, since he makes it seem like the movement is his, that it belongs to him. It’s a self-assured march. But he comes right back to where he started, and that itself is the problem: the illusory movement in space in Robbe-Grillet.
In other words, when it comes to Robbe-Grillet’s novels, space is either a vain attempt to ward off immobility, as in The Erasers, or else there’s this sort of assessment of a troubling spatial distance that renders desire impossible. Matthias, for example, in the café scene in The Voyeur. So, in the New Novel—Robbe-Grillet in particular—increasingly, the protagonist, who initially appeared to walk—in The Voyeur, The Erasers, etc.—increasingly, the protagonist comes up against an immobility, for example, the jealous man writing behind his slats [in Jealousy] or the enclosed, contiguous space of the room in In the Labyrinth, where [we find the line] “I’m alone here, under cover. Outside it is raining.”
In other words, Robbe-Grillet offers a sort of immobility that becomes palpable in his novels, one that unravels both the optical illusion, the illusion of the optical gaze, as well as the illusion of moving in space—leading us where? Again and again, a fascinated, dazzled gaze, which might take the form of a montage, or a single cut-away on its own—a montage, a jump cut—a dazed or fascinated gaze, a gaze that seems to have an object. I say “seems” to have an object, which is the opening sentence of La Maison de Rendez-vous: “Woman’s flesh has always played, no doubt, a great part in my dreams.”[4] In other words, this fascinated or dazzled gaze “no doubt,” he says, seems to have an object. In other words, the ambiguity of this “no doubt” immediately introduces a humorist uncertainty of a “likely object” to the ironic certainty of a possible object.
I’ll wrap up by saying that, ultimately, for Robbe-Grillet, space is described or imagined—the space of fantasy, space composed of snippets of perceived space, for example, as a smash-up or a faded perception of space superimposed onto another visible space—whether described, imagined, or fantasized space, he constantly deforms and differentiates space, like the labyrinthian space of repetition he tries to bring out with Alain Resnais in The Last Year at Marienbad—the labyrinthian space of repetition, i.e., the repetition of a perspective fascinated and dazzled by mute immobility, where the protagonist’s space—which we often think of when we talk about Robbe-Grillet’s fiction—doesn’t necessarily coincide with the space of the novel. There is a discrepancy rather than coincidence.
However, you might say, if this repetition continues throughout the whole space—the entire space, which constantly exceeds the labyrinthian space of fiction—this labyrinthian space itself is distinct from the space of repetition as such, i.e., the space of the language games that makes it such that something like the gaze can be repeated. Put another way, it’s supported by the space of language, or of language games, and I’ll end with a quote from Bruce Morrissette’s laudable reading of Robbe-Grillet, which reads: “there is no longer even a ‘absent-I,’ nor a ‘he’”—for example, Kafka’s “he,” as opposed to Joyce or Proust’s “one”—“There isn’t even a ‘absent-I’ or a ‘he’; only an ‘I’ and ‘he’ that verge on disappearing, bleeding together in a text that attempts to coincide with the space of language.”[5]
Deleuze: Well, I’m rather intrigued by your comments. Just one question: you alluded to a few films throughout your remarks. That being the case, seeing as you’re ostensibly focusing on Robbe-Grillet’s novels, would you say the same holds for cinema?
Comtesse: Of course not!
Deleuze: Really?
Comtesse: No, because at several points Robbe-Grillet points out that there’s a difference, that it’s not at all the same thing. [Deleuze: Okay] “I’m talking about a novel, which is not at all the same as when I’m working with Resnais or what I’m doing in The Immortal One [the film], The Man who Lies,” etc. [Deleuze: Okay] And, for example, he stresses the fact that in his novels, “protagonists are completed isolated and mute. And the writer’s voice is by no means a speaking voice.” [Deleuze: Okay] It’s a different situation. Of course, there might be some sort of correspondence, in terms of the operation itself, but whether it’s writing or in film, it makes a difference.
Deleuze: I see. I believe one resource—this is a slight departure from what you were talking about, Comtesse—one resource, it doesn’t explain Robbe-Grillet, but it’s a film that I really want to see, but I think it’s pretty difficult to watch, is Beckett’s Film. Have any of you seen Beckett’s Film, with Buster Keaton? You’ve seen it? [A student says “yes”] I feel like there’s… like that could be a pretty important resource for, that it would be very, very important. Anyway, it should be possible to watch this film. I don’t believe it’s very long, right?
A student: Right.
Another student: I saw a show about Robbe-Grillet directed by Rea Saba, and like George said, Robbe-Grillet put together…
Deleuze: TV shows? Well, well, well, now we have a third Robbe-Grillet, via broadcast. With broadcasts, it’s… it’s… [Deleuze doesn’t finish his sentence]
Okay, well, then, anyway, let’s move on. We absolutely have to wrap things up. And so what I’m giving you is more like a program that you’ll have to finish on your own, and it will be different for everyone. But if I find it necessary to go back to where we started, it’s because, in effect, things are starting to come full circle, and they’re coming full circle in two different ways, following two paths. And we currently find ourselves traveling along both of them. Or we ought to, we ought to follow… Would you shut the door? We ought to take both paths.
From the beginning, when we set out classifying different instances of movement-images, right, our intuition told us that movement-images wouldn’t be the only type of cinematic image. What’s more, we had our Bergsonian hypothesis. Which is what I mean by our first path. Our Bergsonian hypothesis: that the movement-image is a slice or a temporal perspective of—of what? Of a duration. Let’s tentatively say: the movement-image is a slice or temporal perspective of a time-image, of a time-image.
However, see, now that our examination of movement-images has started to circle back around, if we have several avenues to choose from, it’s because there is certainly more than one way in which movement-images nudge us away and point towards other images. It’s not even that these other images are immobile; they might be immobile, but not necessarily. The point isn’t that they’re the opposite of movement-images; it’s that they’re totally different—in other words, it’s that we cannot approach them as we would movement-images. And indeed, if a movement-image is the deepest cross-section of a time-image, that ought to highlight the sheer volume of said time-image, the fact that it’s “voluminous.” Thus, not only does it have depth [e.g., of field], it is temporal—the time-image!
But what exactly does “time-image” mean? For us, that implies, again, that time can only be depicted if its image is not in the form of a sequence—that, on the contrary, would only amount to a series of movement-images, i.e., the form of a sequence, even if this form involves speeding up, slowing down, cutaways, flashbacks, etc. The time-image is only related to the movement-image in that movement-images are in a way signposts for [time-images]; we have to go from movement-images to these deeper time-images once we recognize that the latter can’t be reduced to a sequence of images, which means that in a way, [the time-image] is really the “image” itself—it’s not a sequence of images.
But what on earth does that mean, that it’s “the image itself”? Again, our definition was: the movement-image is a cross-section or temporal perspective of a duration. We saw what that meant for Bergson. Duration is what changes at every moment. It’s what constantly changes and varies; it’s a totality, properly speaking—that of being open, the concept of open whole that we spent so much time on. And in the end, wouldn’t the time-image ultimately be the “whole” of a film? What Eisenstein tends to call the Idea, “Idea” with a capital “I”. We’d have cinematic ideas, i.e., images whose character is different from that of movement-images, cinematic Ideas. Thus, we would go from the movement-image to the Idea. But this Idea isn’t a series of images; it’s the time-image, or the Whole, as open totality. So on and so forth, right. But it’s the Whole, right, the Whole.
If you like, this other sort of image—the one I’m looking for—is this other type of movement-image the Whole? Yes, it’s the Whole. It’s the Whole of the film, in Eisenstein’s terms. That’s what we’re calling the Idea. Okay, but then it’s not any particular image; if it’s the Whole, it isn’t a particular image. And yet it’s that, too! We’ve painted ourselves into a corner, for better or for worse. We have to sort it out. We have to maintain both. Yes, it is the Whole of the film. And yet, it is a certain type of image. How do we reconcile that? It being a certain type of image means that it’s one type of image among others. It’s a distinct type of image, distinct from movement-images; it’s something alongside movement-images. At the same time, I contend, it is the Whole of the images of the film. Will this be a headache to deal with later? Maybe not.
We shouldn’t be too hasty in calling this a contradiction. It reminds me of an author in a different context, a literary author: Proust. In one of his last books, the last volume, Time Regained, Proust spends pages and pages saying that there’s a Whole in his book, In Search of Lost Time. There is a Whole! And it’s true—there is a Whole! Only it’s a very special Whole, because it itself is a part among parts. So it’s a very special Whole, all the more since a Whole ostensibly has no parts, as it is a Whole of parts. It doesn’t have any parts itself, but that doesn’t stop it from having “aspects.” It’s a Whole under one aspect or another. And not only is it a Whole under one aspect or another but, at the same time, we should add that it’s a part among other parts.
Well, then, let’s keep looking. This direction gives us the first route. When I brought up—but we don’t have to get hung up on it—I briefly mentioned depth-of-field, and that’s one type of image. Right. That’s only one type of image. In Welles, for example: depth-of-field images, as opposed to movement-images with no depth-of-field. Yet, in a way, it’s also true that these depth-of-field images can lend themselves to open totalization, can be Open Wholes under a certain aspect. And true, in the end these depth-of-field images, oddly enough, appeared to have two functions whereby they fundamentally constitute time-images, i.e., they aren’t content to merely suggest an image’s volume; they introduce a fourth dimension, time, as twofold: as contraction time—as we discussed—and a form of time that seemed to be almost the opposite of that, a sheet or circuit, time-sheets or time-circuits. And I felt like that corresponded perfectly to the two principal forms of Bergsonian memory: memory-as-contraction and memory-as-sheet or circuit.
But having said that, we shouldn’t get too attached to depth-of-field because if we start to think, “depth-of-field is the time-image”—no, no, no! It could be, but there’s no reason to go there. Besides, there are some very talented directors, on the subject of Time, who never or almost never used depth-of-field. Take Fellini, for example. Or Visconti—at root, his approach to time and cinematic time does not depend on depth-of-field or anything. So, we shouldn’t say that depth-of-field is required for time-images. No. You can—You can use depth-of-field to get at how time operates, either time as contraction, where time is weighted by the form of its contraction, or time as layered, circuited time. Sure, it’s possible, but it isn’t necessary. It’s optional.
However, what is it about then? See, as for this first path, when it comes to the relationship between movement-images and time-images that could just as easily be one between movement-images and Ideas-with-a-capital-“I”—the Whole—what’s happening, exactly? It comes down to our ultimate problem. And fittingly, as we come to the end, the very end of the year, as this has been our problem from the beginning, namely—obviously, and that’s how I got into this subject—it comes down to the question: is cinema capable, even hypothetically—I’m not asking whether it’s succeeded—is it capable of offering us a new way of thinking, that is, a new form of thought. And by the same token—this is inseparable—[is it capable] of making us think in a new way? Are cinema-images uniquely tied to thought?
So, that was our problem, in effect. And you can sense that I’m homing in on some necessary relationship linking a movement-image to either—perhaps—either time-images or cinematic Ideas (with a capital “I”), which is both a film’s Whole as well as one type of image among others; if I find that, that will settle the issue—as far as I’m concerned, at least—insofar as we’re dealing with a relationship between cinematic images and thought. Which reminds me of a particularly striking passage from Alexandre Astruc on the subject of depth-of-field, but obviously we have to put it in context. Suffice it to say that depth-of-field is a crucial technique. You shouldn’t use it—obviously, you shouldn’t use it arbitrarily; that would be pointless. It’s a technique that should only be used when there’s something to gain from it. And if you don’t use it, there are plenty of other techniques. Besides, technique isn’t everything. But I’m reminded how Alexandre Astruc says that depth-of-field, such as you might find, for example, in Jean Renoir, before Welles, which Welles mastered and brought to a new level—well, these images have changed a lot, he says, in terms of the function of thought in cinema. I love this passage because it’s still a bit mysterious; he doesn’t really spell it out. He writes, “Before, ultimately, the relationship between thought and image, if the film-image had any bearing on thought, it was in the form of metaphor.” [6]
And indeed, previously in film—I’m thinking of [Jean] Epstein, when he asks how cinema is related to thought—the relationship [between cinema and thought] always boils down to the idea that film presents thought in an incredibly powerful metaphorical way.[7] And Astruc writes, “With depth-of-field, thought ceases to function metaphorically in relation to cinema; it becomes”[8]—and now things get mysterious, but the writing is wonderful—“it becomes a ‘theorem.’” It becomes a theorem; we’ve gone from one state of thought to… what could he mean?
It isn’t very clear, because he goes on, writing, “What impression do we get from depth-of-field?” It’s as though, he claims—and here he’s thinking of certain depth-of-field shots in Renoir—it’s as though the camera were “a snow plow.”[9] “[Renoir] directs with a snow plow, crushing the extras against the sides of the camera.” Yes, “crushing” whatever no longer matters in the shot, a sort of temporal breakthrough—right away, things are getting interesting—a temporal breakthrough where, insofar as it feels like the camera is pressing on, things get forced off the right and left margins of the screen.
I really like an image that comes up much later than Astruc’s book, from Fassbinder’s film, Lily Marleen [1981]—I think? Is that what it’s called? Lily Marleen. For those who’ve [not] seen it, just to catch you up to speed, it perfectly illustrates the “snow plow” effect Alexandre Astruc is describing. An altercation breaks out in the back of a café; they’re really going at it, and the depth-of-field is enormous. And people are scared off, customers are afraid, the café’s customers are frightened by the fighting. They run towards the audience; they run up to the foreground, see, and it makes you feel just like someone might when entering the café, pushed back by the frightened crowd, while the fight carries on in the back. And so it’s a wonderful image, a textbook depth-of-field shot where you can clearly see how the camera functions as a snow plow; it feels like the camera tosses aside useless extras, the guys running, the ones running towards us. A beautiful “snow plow” scene.
So, that would be more of a “theorematic” method rather than a “metaphorical” one. Eventually you might wonder whether movement-images, if there were only movement-images, wouldn’t that bring us back, for example, to what the French School—[Abel] Gance, Epstein, etc—focused on: basically, a “metaphoric way of thinking”?[10] Maybe once you produce something other than a movement-image, something else emerges, i.e., the possibility for thought. Well, let’s stick to Astruc’s terminology for the time being—[the possibility for] “theorematic thought.” But what does “theorematic thought” refer to in cinema? Is it a coincidence that one of Pasolini’s major films is called Theorem [Teorema, 1968]? What is that film about? It’s a funny film! Okay, well, that’s that.
Our second path, reaching the same conclusion—I just showed how the movement-image itself [pause—the sound of an engine outside], because it’s a cross-section or a temporal perspective, hints at another type of image. The second route would be what we’ve focused on so intently the last few times; I’ll summarize it again. This time it’s like a suspension, a calling into question of movement-images, which all the more suggests another type of image, another type of image. This time we broach the question directly by interrogating the movement-image itself. I’d argue that our first path was indirect. Here, putting movement-images into question—among them, action-images in particular—calling action-images into question directly leads us to another sort of image that I might call a time-image, a thought-image, or a cinematic Idea.
That’s another route, and we saw how certain trends in contemporary cinema have taken this path, namely—to jog your memory, since it’s been a while—Italian Neorealism, the French New Wave, the so-called New York School, etc. Right. And we were very interested in that since now it’s like sensorimotor images have been put on hold, sensorimotor images have been suspended for the sake of bringing forth a quote-unquote “pure” sensorial image, a “pure” sensorial image, i.e., what I referred to as a pure image of sight, or of sound. Yet it is this pure sensorial image, cut off from, or at least out of phase with, its normal motility, that makes contact with our other type of image [the sound of something falling or of Deleuze bumping into something] Ow!
And so at this stage, with this other type of image, I’m still hanging onto our hypothesis, if you recall—whether it’s good or bad, it’s way too late to say—if it’s true that the pure optical image, what I’m now calling—now that they’re synced up, there’s no reason to privilege the optical, since it applies to sound as well; the same goes for Italian Neorealism, incidentally; there have been new developments in technique, but it doesn’t change anything—If the pure sensorial image is cut off—I’m saying “cut off” as shorthand; you know it’s more complicated—cut off from its conventional, customary motor extension, well, then, by definition we’re immediately led to another type of image, which we’re currently looking for, what we’re calling time-images, thought-images, which, again, is both a particular type of image and at the same time the Whole of the film’s images, the Whole under one aspect or another. And it’s because it’s always the Whole of the film under one aspect or another that I can say that it’s both the Whole—oh, but watch—it’s also one special sort of image among others.
Where does that leave us? Well, it means—okay, now I’m expanding on where I left off—Well, yeah! We need, since our purely optical, sound, sensorial images are no longer connected to traditional motility [pause]—traditional motility meaning the action-image as we’ve discussed it, whether action-images in the form SAS or ASA—since so-called pure sensorial images are no longer connected to [pause] movement-images, they become connected to or can—they have the capacity—become connected to or connect us to another type of image, right, to the Whole. But this Whole is also a part. What I’m describing is the Whole, Wholes, aspects of the Whole, which is also a part alongside other parts; in other words, again, it’s a type of image different from that of movement-images. Allow me to call it a “mode.” A “mode.”
Why call it a “mode”? Because the term, “mode,” does a good job of expressing the payoff for this other type of image, which as we saw, is thought. I contend that there are as many modes of thought as there are aspects of the Whole, or particular images other than movement-images. Thus, the role—this is abstract, but I believe this abstraction might help us track what I have left to cover—all I’m saying is that pure sensorial images are no longer extended in the motility of action-images, but are now extended in what will be “modes” of thought-images. Maybe there will be lots of modes, but what modes all have in common is their being modes of thought. Hence my question: what is the pure sensorial image—which I’ve used as a premise, which I’ve tried to make the premise of all our previous meetings—what modes, what modes of thought are pure sensorial images primarily connected to?
Alright, I think that up to this point, and this is obviously not an exhaustive list, they’re tied to four main modes. [pause] They’re connected to four main modes. And here, I’ll summarize what we have left to cover, but we could spend a whole term on it, and there are just two meetings left before we’re done. Four main modes. But, again, there are five, six, n modes; it’ll be…it’ll be up to some new filmmaker to find others. But by my count, there are four. And I use a single mode to group really different people.
For the sake of convenience, I’ll call the first the “imaginary mode,” referring to a certain type of image that I’d like to call—I’m anticipating the classification of signs we’ll go over next time—let’s call them “scenes,” scenes like you might see on stage. Scenes. So, the first major mode would be the imaginary, and… [Interruption of the recording] [46:33]
Part 2
… to get into more detail, I might point to Fellini as an example. Okay, that would be one mode.
The second mode. It’s very different, what I’d call a “didactic” mode. [pause] In this case, pure sensorial images wouldn’t be tied to scenes, in the imaginary mode, but to a certain mode of thought we might refer to as—it’s a rather fitting term, actually—“didactic,” which means what, exactly? Here I’d include—I’ll have to explain why in a moment—I’d include Rossellini’s later work: his Socrates, his Taking Power, etc.[11] And for reasons I’ll try to explain, I’d also put [Jean-Marie] Straub in this category. [Pause]
Third is what I’d call the “critical” mode. In this case, the purely sensorial image is fundamentally linked to and links us to thought understood as a positive critical activity and not as a didactic activity. And for this kind of positive criticism—I’m not so confident on this point—I could only think of one example, but it’s a very important one: Godard.
The fourth mode—you’ll understand why I’m going over the list of modes beforehand—the fourth mode is, to use a complicated, philosophical term, the “transcendental” mode. This mode corresponds to when the purely sensorial image is linked and links us directly to time-images as a mode of thought, that is, to time as the time not of things but of thought. Because strangely—not so strangely, in the end—thought takes time. I mean, thought isn’t something eternal. And the idea of time as a mode of thought seems to be one of most fundamental problems, perhaps common to both cinema and philosophy. And for an example of this transcendental mode, I could bring up any of what I think—you’re welcome to rank them however you like—any one of what I think are the greatest directors of all time, such as—it’s a short list, really—whether it’s Resnais, Visconti, [pause; someone makes an announcement outside over a loudspeaker] Pierre Perrault in Canada, someone whose, to use Comtesse’s terminology, whose temporal structure is eminently paradoxical—from what vantage point are they paradoxical? Because we’re really dealing with time as a mode of thought.
Well, I’ve gone over the list, as imprecise as it may be, but I’ll go over it again anyway. You can tell right away that the pure sensorial image, which breaks with normal motility, is no longer extended into movement, as is the case with action-images, once it extends into one of these modes. But clearly, if I’ve run through the list—you might have others to add, if you see something I’ve overlooked, I’d be interested to hear it—I went through the list just so that you can see how these distinctions are still rather blurry. Because obviously all of these modes sort of bleed into one another.
All the same, clearly, I can’t just say that Fellini uses the imaginary mode and that’s that. It’s obvious that Fellini has a fundamental relationship, that Fellini’s approach has something fundamental to do with time. That doesn’t mean that—I’ll elaborate—his relationship to time doesn’t involve the imaginary mode. Right. But in Visconti, meanwhile, it’s clear that Visconti—or Resnais—his work had to do with the imaginary; sometimes the imaginary is even the focus of his films. As I see it, that’s not his… it’s sort of a gut feeling, but our guts might all be different. In Resnais’s case, I think that wasn’t really his problem: he only got to the imaginary via what for him was a more pressing issue, the transcendental route, i.e., that of time, of the time-image.
Why does it all blur together? In other words, why are there bonds linking these four modes together, which I have arbitrarily isolated? As I said before, this is the thought-image in film, in any case. In cinema, the thought-image is the hyphen-image. It’s thought, which is both the Whole of a film and one particular type of image. So, as a particular image, there very well may be four modes, sure, but it’s also the Whole of the film—that is, in any case, it’s about the relationship between the pure sensory images and thought; how do pure sensory images incite thought?
For me, that’s what’s at issue between cinema and philosophy. At the end of the day, is it possible that so-called philosophical thought and so-called cinematic thought share a common cause? And it’s there, then… I can’t delve into particular modes yet, the four modes I just defined. Because I want to emphasize that, if I’m calling them “modes,” it’s because they do indeed share a common root or substance, i.e., in thought, and that brings us to what’s really at stake in cinema.
Hence one of my concerns. It’s that everyone’s already suggested this. It’s always been hinted at it, everything I just went over. Yeah, yeah… But there’s one, there’s one who did more than simply gesture at it, someone who wound up in a catastrophic situation—true, film directors always find themselves in catastrophic situations, so you shouldn’t read too much into that—and I’d like to say a few words on the matter. It’s Artaud, Antonin Artaud. [Pause] Because he got in a bit of a pickle that, I think, isn’t widely talked about, but when it comes to Artaud, what is widely discussed? Artaud thinks, for better or for worse, that he has something to say about cinema and about how to make film. He wrote a few scripts himself—he wrote some. But then certain events unfolded that I’m still not clear on—I don’t know much about the state of his collected writings… there are certainly texts I don’t know about; I should ask Artaud’s editor—but as far as I can tell, it’s a real mess. There’s one scenario he wrote, the only one that made it to film: The Seashell and the Clergyman [1928]. Well, it was directed by Germaine Dulac, who was otherwise an excellent filmmaker, a very good filmmaker. Anyway.
Things get murky from here. Did Artaud participate in the filming? Did he even participate in—if not in the filming itself—did he actively contribute to the scenario’s adaptation, or to the editing? Some say yes, I believe, and others say no. Some of what Artaud has written about Germaine Dulac is very laudatory; elsewhere Artaud is horrible and treats poor Germaine Dulac like a dog. He claims that she doesn’t understand anything. Right. And the situation gets even more confusing if you consider the fact that Artaud writes, “Everyone’s been stealing from me.” Not just that they mutilated, tampered with, or distorted his film, but that it was actually the first surrealist film. He really resents Buñuel and Cocteau; he’ll go on to claim that they profited off him. But they failed to capture what was at the heart of The Seashell—or what ought to have been at the heart of The Seashell and the Clergyman. Anyway, let’s move on.
What does he mean? He’s referring to the fact that they’ve made dreamlike films. The situation then gets even more complicated, because when you see The Seashell and the Clergyman, it actually does look like the first surrealist film because it’s a purely dreamlike film. And he calls Germaine Dulac a villain because she completely misunderstood his scenario, his film; she turned it into a dream. Alright! So, he both says that it’s the first surrealist film, but at the same time, he claims that it was anything but a dreamlike film, or it wasn’t supposed to be, at least. What a mess! Then what… Because what exactly is he complaining about?
Dreamlike films are already accounted for in our categories, right? See, that gives us some headway. The dreamlike film would be one case, for our purposes at least, of what I’ve been referring to as the imaginary mode. And the imaginary mode can assume a dreamlike form, in dreamlike images, dream images which, after all, are very different from movement-images—even when they move. Well, in the dream image, you can really see how this mode, the dreamlike imaginary mode in this case—“imaginary” includes dreams, but it covers much more besides—well, you can see what makes the dream image a mode of thought. But the imaginary mode might be the most ambiguous of my four modes, the most dangerous, the most dubious. I bring it up because it is, dare I say, the one I’m least fond of. I’m just saying, I don’t like it… so… But that’s where I’m coming from; may you do like it.
At any rate, I think that’s how it gets you, because it’s… How is it a trap? You can already tell. It’s because it is manifestly the mode which is the easiest to pull off using, quite frankly, cheap techniques. It’s not a coincidence that in the early days of cinema, when everyone was still trembling with excitement, like young folks discovering things for the first time—and they were!—the relationship between cinema and thought wasn’t all that difficult, allegedly: cinema opens a window into dreams. Cinematic thought was explicitly assimilated—see Jean Epstein’s work, for example—it was explicitly relegated to dreamwork. I think it’s very… I’m referring to someone who certainly wasn’t a surrealist author—as far as the Surrealists are concerned—Epstein is anything but a surrealist author, but he contends that one of cinema’s key aspects is its ability to simulate dreamwork. In what way? With condensations, super-impressions, ruptures in logic, plane ruptures, all of film’s technical methods, it has the potential to be an excellent experiment in dreamwork.
Yet I wonder—isn’t this imaginary mode ultimately the most dangerous? We’ll see; it’s an open question. But that would somewhat explain Artaud’s reaction. He goes out of his way to specify that The Seashell and the Clergyman should resemble a dream, but that it shouldn’t be one. His stance isn’t clear; it’s rather complicated. In any case, for those who haven’t seen it, I promise that if you watch The Seashell and the Clergyman: it is a dreamlike film. Germaine Dulac made it a dreamlike film. Indeed, it is the first surrealist film. Well, that’s not what Artaud wanted. But what did he want? What was Artaud expecting?
Well, I found his thesis, his practical thesis, rather interesting. He argues that, for him, that what’s really at stake in cinema is thought, a question of thought and not of dreams. Alright.
It’s a question of thought. The second he suspects that it’s no longer an approach to the problem of thought, he abandons cinema. Then he thinks that cinema is worthless. He believed in cinema insofar as he believed that the problem of cinema was a problem of thought. Only what does that mean? His position is even more complicated than I’m making it seem, because when he says that cinema’s problem is one of thought, he does mention dreaming, but he rejects it at the same time. I’ll refer us to a text—everything I’m quoting from him comes from the third volume of his Complete Works.[12]
“This scenario [Seashell] is not the reproduction of a dream and must not be regarded as such. I shall not try to excuse the apparent inconsistency by the facile subterfuge of dreams.”[13] “Facile subterfuge”—I like that a lot; that’s great. “Dreams have more than their logic. They have their life where nothing but a sombre truth appears.” Artaud’s style is unmistakable. “This script searches for the sombre truth of the mind in images which emerge exclusively from themselves.” Alright. But he doesn’t deny that dreams are involved. Moreover, he says—on page 76, I think[14]—the scenario (still The Seashell and the Clergyman) might “resemble”—see, he doesn’t deny it—it might “resemble a dream and ally itself with the mechanics of a dream”—that is, to dream work, to what psychoanalysts call dream work. It might “resemble and ally itself with the mechanics of a dream, without really being a dream itself.” See, he’s sort of painted himself into a corner. “It is to show how far it restores the pure work of thought.” I’m going out of my way to demonstrate how ambiguous his position is in order… see, what matters how cinematic images are related to thought, and properly understood, dreaming is a mode of thought.
So, [for Artaud] the relationship between thought and cinematic images stems from the allure, stems from the dreamlike mode, but this isn’t the final word on the matter, despite appearances. It’s a complicated position. Personally, I think this sort of mess is inevitable for anyone making imaginary films: they wind up in a sticky situation where they know that their goal lies elsewhere, but they get caught up fiddling with this imaginary stuff, and they’re not able to get out of it. Which is why, of all four methods, the imaginary is the only one that’s really suspect. But you be the judge. Whether it’s the best or whether it’s the least interesting. At the very least, you know, it’s… it’s like, so much to say that… Anyway, it’s not my favorite.
But let’s move on. What was Artaud trying to do? Given how complex his position was—see, it’s complicated, so I went over all of that because I didn’t want to say anything rash—well, it’s funny, because what exactly does Artaud want? From what I can tell, no one has picked up on certain moments in his work—pointing them out makes me seem pretty smart—which only stand out because of everything that’s happened in modern cinema, completely independently of Artaud.
Because think about what Artaud’s always been saying: Artaud takes issue with a dualism he sees in cinema, in the French cinema of his day, which tended towards either abstraction or towards narrative. Indeed, everyone had a part in abstract kinetic cinema[15]—Grémillon, Dulac, and so on—everyone was making abstract films. Their approach to film was as a reflection on visual rhythms, as purely an examination of visual rhythm. Between that an narrative cinema, Artaud claims, we need a third option. Otherwise, cinema will perish—it will drown in either platitudes or in abstraction. Right. But then what is his solution? Well, this is what he proposes.
First, what is narrative film? He says that as a form of cinema, it’s textual. It depends “solely on the text [to the exclusion of the images].” It’s a plot-driven kind of film. In other words, it’s what we’re referring to when we talk about film through the lens of action-images, S-A-S or A-S-A. It’s a form of cinema that tells a story; it’s narrative film. I’m not saying that the narration is all that’s there; beautiful things have come from these sorts of film. But it’s narrative film, the cinema of action-images. It has to do with what we’ve been calling sensorimotor images; it’s precisely about the sensorimotor side of images: the action-image. Anyway.
Yet he tells us on page 76–oh, geez, my page numbers are all wrong—no, it’s page 20, I hope. Here’s the passage that tipped me off. He says, alright, that’s what I don’t like. Then he writes, “We must find a film”—“We must find a film with purely visual sensations the dramatic force of which springs from a shock on the eyes, drawn, one might say, from the very substance of the eye, and not from psychological circumlocutions of a discursive nature which are nothing but visual interpretations of a text.”[16] What a beautiful passage. See, “a film with purely visual sensations” doesn’t imply vision in the abstract. It’s a question of form, since the context concerns his denouncing abstract kinetic cinema. It’s not about purely visual movements; it’s about… I’m not happy with this wording, but I’m just putting it in the terms of the quote—it’s about “situations,” not abstractions but purely visual situations, as opposed to narrative films made up of optico-motor, sensorimotor situations.[17] “Purely visual sensations the dramatic force of which springs from a shock on the eyes.” Alright. Anyway.
A second quote. And it’s here that maybe… This is where he adds that it “is not the reproduction of a dream and must not be regarded as such.” A purely visual situation. Right. He says that there is drama, but it’s no longer a narrative drama, or a drama of actions. It’s another sort of drama. Let’s think, what would that be? Is there another passage where he elaborates on this? Yes! Page 61. Artaud writes [Pause], “The clash of objects and movement”—there’s that word again, clash[18]—“The clash of objects and movements produces psychic situations”—note that he just rejected psychology—“psychic situations which wedge the mind in and force it to find some subtle means of escape.”[19] So Artaud is calling for a form of cinema that uses purely visual situations to produce purely psychic situations. Right.
Did he have an idea for how to pull it off, for how to achieve that? I think, if I didn’t say that this was Artaud; if I had told you that this was written by Godard or written by Rivette… or even written by Rossellini—the differences between all these directors notwithstanding—I don’t think any of them would have disagreed with a single word of it. Let’s move film away from optical situations…—if I were to summarize Artaud’s manifesto, rewording it—let’s detach action-images from narrative cinema, that is, let’s get away from sensorimotor images in order to establish a direct link between pure optical situations and just-as-pure psychic situations, yeah, I think that’s what we’ve seen from the beginning, since I’ve tried to pin down the common thread running from Italian Neorealism onward.
But I’m certainly not saying that Artaud anticipated that, since again, it isn’t just the direction of Germaine Dulac. Read the scenario for The Seashell and the Clergyman itself; the only way—and this might also be attributable to surrealism, which is fine by me—because at this point, he was still hung up on surrealism, the only way Artaud understood his project was that it was a film, despite its dreamlike aspects. In other words, he took the hardest, most ambiguous route—he went about his project in the most ambiguous way, using the only method incapable of realizing his project, the imaginary. Alas… he couldn’t do it any other way.
That’s why I don’t claim that modern cinema hinges on Artaud—not at all. Other methods had to be forged for something like Artaud’s project to be viable, using purely optical situations and purely psychic situations. All I’ll say, for those interested in Artaud’s problem, is that this much is true of The Seashell and the Clergyman: I think he’s pushing it. It’s the reproduction of a dream. Read the scenario. It’s—watch the film, which occasionally makes its way to theaters, but read the scenario; it’s about a dream. I have a hard time… indeed, its optical situations are dreamlike; they aren’t optical situations. Anyway.
On the other hand, two of his other screenplays are worth mentioning, ones that didn’t make it to film. I recommend that you look them over if you’re curious. There’s a scenario called “Flights,” which is a drama, but the drama seems incredibly boring. “Flights” is about a young lawyer; a young woman comes into his office with some papers, some document that’ll win him his case. So the lawyer declares, “With this we are sure to win.”[20] Then there’s something of a spark between the lawyer and the woman. Then some conniving rat—you can tell right away that they’re up to no good—breaks into the lawyer’s office under some pretext, snatches the file, and then slinks away. The lawyer returns and discovers that the file is missing—he pulls at his hair, the woman starts crying, and so on, the plot’s off to the races. But that’s only a part of the story; that’s how it starts. A cab ride—not very dreamlike—a cab ride where the guy is trying to track down the thief. Has his work cut out for him, trying to spot him from a cab, eh? We get a description of a long… with optical situations. Here, too, things occasionally veer toward surrealism, but it feels like the setting isn’t right for surrealism, like this is more of what Artaud had in mind. I think the real Artaud is more present in this earlier scenario, “Flights,” than in The Seashell and the Clergyman. Later, [the lawyer] gets on a plane after his cab pulls up just in time to see the thief catch a flight to the oil fields out east—the trial is about oil. So the thief takes the Orient Express, and the lawyer catches a plane. There are two journeys; both voyages are treated as purely optical situations. In the end, obviously, he gets the files back, strangles the thief, everyone’s happy. [Laughter]
And then something really interesting. The second scenario is “Eighteen Seconds,” eighteen seconds of a man’s life—the film is an hour and a half long, but this hour and a half is a time-image of 18 seconds. What’s it about? It’s about Artaud’s own drama, such as how Artaud describes his approach to drama, namely, there’s something within thought that hinders the exercise of thought, or if you prefer, an inability to think at work in the heart of thought. An inability to think at the heart of thought—how does Artaud intend to depict that in film? Once again, it’s through a series of visual situations, clashing visual images and the failure to form thoughts for which images can serve as mode. After 18 seconds, the guy fires his revolver and—bang!—kills himself.
My point is that these two screenplays—I’m not calling them modern, which would be nonsense—but they have modern tendencies that I find lacking in The Seashell and the Clergyman. What I’m getting at, then—this quote from Artaud was only meant to lay out my problem—I’m not trying to say that Artaud had it all figured out; that’s not at all what I had in mind. I’m saying that I think it’s interesting how Artaud paired these terms—optical situations and visual situations—in reference to psychic situations, “psychic situations” meaning the painful exercise of thought. Right. So, where were we.
Now I’ve got my bearings… we’ve gone from pure sensorial images to thought, a uniquely cinematic form of thought, so much so that I’ll call them thought-images, with a hyphen—a relationship actualized under four possible modes—again, this isn’t an exhaustive list: the imaginary mode, [which is] extremely suspicious, not to be trusted [laughs]; the didactic mode; the critical mode; the transcendental mode—and whatever other modes filmmakers might come up with.
Now, let’s take a closer look at the first mode. Well… Let’s take a break because I have to go to… Yes, someone wanted to say something?
A student: [Inaudible]
Deleuze: What about Kurosawa in The Idiot [1951]?
The student: [Inaudible]
Deleuze: The time-image? Well, now, listen, what I want to do next time is…
The student: [Inaudible]
Deleuze: It’s fine if you want to add more to the list. The Idiot isn’t fresh in my mind, so I don’t have an answer for you, but that sounds great; what you’re saying is close to what I have in mind. If he didn’t have another way of [pulling it off], you might have other directors in mind I hadn’t thought of. If you can pin down how Kurosawa was able to achieve time-images, I’m all ears… I should… I should make a quick note of that… [Interruption in the recording] [Pause]
… But in the end, these four modes might fall short. [Pause—the sound of students coming in]. So, see, that’s what’s left for us to do, and then we’ll be done: an examination of our four modes. These are modes whereby, once taken up by the sensorial image, result in thought-images, since they are modes of thought. That much should be clear. So, this is how we’ll proceed: I’m only going to cover the bad mode, the first one, and then we’ll breeze through the others since they’re the good ones [laughs]. Because, I mean—all the same, I’m calling the first one “bad,” but not really—on the contrary: it’s great. So… It’s complicated, at any rate. Each individual mode covers a wide range of different things, different sub-modes. You see, our main payoff will be having completed a table of signs; next time we’ll get into a classification of signs. We’ll have a whole series of signs… Nothing like Kant’s paltry twelve categories—we’ll have eighty! Alright, then.
I call it the “imaginary” mode, which is fitting because it covers extremely different things. First off, it covers what we might call the “fairytale” mode.[21] What do I have in mind by “fairytale mode”? The sort of fairytale mode you find in [De] Sica. Why am I circling back to De Sica? I’m thinking along the lines of Miracle in Milan. Confronted with the sort of fantastical elements found in Miracle in Milan, many critics said, while De Sica was so integral to Neorealism, here he’s shrinking away from Neorealism’s demands.[22] You can tell that, following our line of reasoning, we’re in a position to say that that isn’t true. In the same way, when Fellini develops a so-called “imaginary” form of cinema, they say, oh no, he’s leaving Neorealism behind. For us, on the contrary, this is no longer a problem, since by our standard neorealism is a method of bringing out pure sensorial and optical images; it’s not at all surprising that Neorealism might associate such images no longer with ordinary, conventional movements but with a mode… a mode of thought—the imaginary mode, for example. For us, such a development with someone like Fellini, or in films like De Sica’s Miracle in Milan, is no cause for alarm. There’s nothing stopping us from saying, well, that lines up with… with… with what Neorealism is, at its core. So, that’s what I mean by De Sica’s fairytale mode.
And that brings us to something very different: the “dreamlike” mode. And here I’m using the word, dreamlike, in a strict sense: images are displayed as if they were from a dream, whether in surrealism films or in early Buñuel. You might come up with plenty of other directors who use dream images. Sure, but we can’t go on forever; you’ll have to flesh these things out on your own time.
Our third sub-mode, I think, is pretty distinct. Let’s call it… well, none of these terms are set in stone—a “phantasmic” mode.[23] A phantasm is very different from a dream, but for some reason, that doesn’t surprise us. We might recall, from the little psychoanalysis we do remember, that phantasms are fundamentally tied to visual and auditory sources whose motility, whose extension into movement, is cancelled out; the scene of the phantasm takes place in this nullification—“scene” in the strict sense.
The phantasmic mode is, for example, what we see in late Buñuel. I mean Buñuel à la Belle de jour [1967] or Ghost of Liberty [1974], but I think one of the first Buñuel films that heralded this later trajectory—I’m not saying the very first, but one of the first—is Belle de jour. So, then, let’s put late Buñuel in the phantasmic mode, along with the sort of repetitive structures in Robbe-Grillet which Comtesse alluded to. As well as—among other things I think the two have in common—as well as Robbe-Grillet’s sort of films. Except for Marienbad, which I don’t consider to be Robbe-Grillet’s alone.
The fourth sub-mode—you can see how many there are—is the “theatrical” mode, but theatrical in a properly cinematic way. I’m not about to drag out the same old question about the relationship between theater and cinema for the hundredth time; I’m referring to what might be called little theater[24] in film. The use of little theater or chamber theater in cinema. Such little theater or chamber theater in cinema is above all characteristic of [Jean] Renoir’s work. Little theater has always been a crucial, pervasive cinematic point of reference for Renoir. One of his well-known films is dedicated to it, The Golden Coach [1953], but throughout all of Renoir’s work, especially in The Rules of the Game [1939]—The Rules of the Game is a more famous example—you see references to little theater, such as the well-known scene in The Rules of the Game where the guests play out a sort of dark comedy, right. Chamber theater, references to theater, over other modes, more in Renoir than with other modes. There are references to little theater in Rivette’s Paris Belongs to Us [1961], in the hotel in Last Year at Marienbad, and it comes up rather frequently. But if we had to pin this sub-mode on anyone, this theatrical sub-mode, it ought to be Renoir—I think Renoir ought to be the shield-bearer, even though you can also find it in plenty other directors.
The fifth sub-mode—I’m starting to lose track, at this point—is what I’d call the “attractional” mode. There’s the fairytale, the dreamlike mode, the phantasmic mode, the theatrical mode, and then the attractional mode, as in “attraction.”
Because this gets us into an integral problem in the history of film, one that harks back to [Sergei] Eisenstein. It’s no secret that attraction plays a role in the theory of montage, as important as it is for Eisenstein, an aspect which Jean Mitry did an especially good job of highlighting. In his critique, Mitry tried to demonstrate how important this aspect was for Eisenstein’s montage. What Eisenstein called the montage of attraction—and the montage of attraction is what I — [End of the recording][25] [1:32:51]
Notes
[1] In some instances, “visual” might make for more comfortable English turns of phrase than “optical,” but I chose the latter for the sake of continuity with Deleuze’s discussion of “optics” and the “optical” elsewhere.
[2] This is a paraphrase.
[3] Comtesse doesn’t have the page numbers, but the relevant passage is from Alain Robbe-Grillet, The Erasers, trans. Richard Howard (Grove Press: New York, 1964).
[4] Translation is Richard Howard’s, in Alain Robbe-Grillet, La Maison de Rendez-vous, trans. Richard Howard (Grove: New York, 1966).
[5] As the translator did not have access to the Morrissette text, these “quotes” represent a rough estimate based on Deleuze’s wording.
[6] Deleuze seems to be paraphrasing. Versions of these claims appear in Astruc’s “Du stylo à la camera et de la camera au stylo” and “L’Avenir du cinéma,” which were both cited in Pierre Lherminier, L’Art du cinéma (Paris: Seghers, 1960). In Cinema 2, Deleuze cites Lherminier’s book but does not offer page numbers, nor does he distinguish between Astruc’s essays. Although the translation includes quotation marks, the reader should not understand these passages to be quotes. Deleuze cites Astruc in this regard in Cinema 2, p. 226. It concerns a text in Pierre Lherminier, L’Art du cinéma (Paris: Seghers, 1960).
[7] Transcript: See references to Epstein in Cinema 2, p. 165. It’s the text, Ecrits sur le cinema (Paris: Seghers, 1975).
[8] Context for Deleuze’s gloss: Astruc writes that the “fundamental problem of the cinema is how to express thought.” Early, silent film tried to do this through “symbolic association[s]” present “within the image itself,” whereas, because its “primary function is to move, i.e., to take place in time,” film has become “a theorem.” Following the translation in Alexandre Astruc, “The Birth of a New Avant-Garde: La Camera-Stylo,” in The French New Wave,” eds. Ginette Vincendeau and Peter Graham (London: Palgrave Macmillan, 2009).
[9] I’m following Adrian Martin’s translation. Alexandre Astruc, “The Future of Cinema,” trans. Adrian Martin, in The Third Rail, issue 9 (http://thirdrailquarterly.org/alexandre-astruc-the-future-of-cinema/)
[10] For English’s sake, I’ve sometimes translated pensée as “form of thought” or “way of thinking.” I hope the reader is not misled by the “form” or “way” in either of these choices.
[11] On the didactic in Rossellini, see Cinema 2, pp. 253-4.
[12] Deleuze makes this same reference on this same subject in Cinema 2, p. 319 n. 18. The translation follows the Hamilton translation in Antonin Artaud, Collected Works: Volume 3, trans. Alastair Hamilton (London: Calder & Boyars, 1972). Any pagination refers to this translation.
[13] Artaud, p.20
[14] Artaud, p. 63.
[15] So-called cinéma pur
[16] Artaud, 20.
[17] Deleuze says “situation” where the Artaud translation has “sensation.”
[18] The “shock” on the eyes, per the Artaud translation.
[19] Artaud, 61.
[20] Artaud, 43.
[21] La mode féerique
[22] See for example Dario Tomasi, “Vittorio De Sica e Cesare Zavattini, verso la svolta,” in Storia del cinema italiano VIII 1949-1953, ed. Luciano De Giusti (Venice: Marsilio, 2003) a cura di Luciano De Giusti, Marsilio. Edizioni di Bianco & Nero, Venezia, 2003.
[23] “Fantasy” or “fantastical” mode might do a better job of translating fantasme, but the reader may choose to relate this use of the word to Deleuze’s other mentions of “phantasm” as it’s sometimes translated in his work or in the work of his influences (if there’s a connection to be made).
[24] As “little theater” has its own connotations in English (which may be misleading), the reader ought to read this as smaller-scale theatrical performances, chamber plays, one-act plays, commedia dell’arte troupes such as you find in The Golden Coach, etc.
[25] For the montage of attraction, see Cinema 1 pp. 180-1 and Cinema 2 pp. 79, 87.
For archival purposes, the augmented and new time stamped version of the transcription was completed in February 2021. The translation was completed in April 2022 with additional revisions added in February 2024.[NB: The transcript time stamp is in synch with the WedDeleuze recording link provided here.]