January 18, 1972

Where things get tricky—I suppose—in schizo-analysis, is finding someone’s desiring machines. What’s the deal with your things, what are your machines? Then, if the person doesn’t answer in terms of machines, or if we can’t manage to find their machinic elements… obviously, that raises the question: what is our criterion? What allows us to say, ah, there, we’ve pinned down someone’s desiring machines? I think there are very solid criteria, and they specifically come down to this: desiring machines are molecular formations, never molar aggregates.

Seminar Introduction

The opening seminars  at Vincennes (1970-71 & 1971-72) not only provided a site for the ground work from which Anti-Oedipus emerged, but also the transitional discussions with which the long sequence of ‘plateaus’ would be developed. As Deleuze indicates in the opening seminar, he had already spent the 1970-71 academic year outlining the ongoing collaboration with Félix Guattari. Thus, with the book’s publication quite imminent (appearing in early 1972), these sessions allow Deleuze to refine the ongoing developments with Guattari, most notably concerning the tasks of schizoanalysis and the intersection of the fields of political economy and psychoanalysis, as well as the continuing dominant and oppressive impact of this intersection both on society and subjectivity. Note that Deleuze and Guattari had only published one essay derived from this ongoing collaboration, titled ‘La synthèse disjonctive’, in L’Arc 43 (on Klossowski), 1970, 54-62, but with the publication of Anti-Oedipus, numerous interviews will appear in 1972, notably in L’Arc 49 (on Deleuze) and in La Quinzaine littéraire 143 (16-30 June 1972).

 

English Translation

Edited

This extremely fragmentary session consists mainly of two responses by Deleuze to student questions and also of Deleuze’s reiteration of the three tasks of schizo-analysis. The first question concerns the axes of code and axiomatics proposed by Deleuze, then desire’s investment of the social field, and also the investment of the social field by desiring machines. While the first positive task of schizo-analysis is to reach the lines of flight of the unconscious, he recognizes that all investments are necessarily molar or social. Hence the second task, discovering how social investments function at the level of the unconscious, and he notes that this usually requires recognizing how desiring machines are repressed by molar apparatuses of the State or the family. Another question concerns the Schreber case, whether the social investment takes several generations for the paranoiacs to be established. Deleuze points to several approaches (anti-psychiatry; Lacan) toward this aspect, and he considers such chicken-or-egg questions difficult but necessary, yet provided one avoids the indefinite familial reduction. Deleuze manages to express a certain hope that rather than ultimately reterritorializing on artificial land, the movement of deterritorialization might create a new earth. The session ends with Deleuze disagreeing with a student’s Lacanian emphasis who insists that the generational question is relevant, to be pursued the following week.

 

Gilles Deleuze

Seminar on Anti-Oedipus I, 1971-1972

Lecture 04, 18 January 1972

Translation: Billy Dean Goehring; revised transcription, Charles J. Stivale

 

Richard Pinhas: One thing I’d like to bring up—not that I’m necessarily arguing against it—earlier you outlined the two main axes for our investigation, the first being “code and axiomatics” and the second being the way in which desire invests the social field, the question being how it does so. I’m curious why you didn’t take that line of inquiry further and ask why. What I’m getting at is something along the lines of what happens in fascism—to take a random example—there’s a relationship which shows us why the desire/social field relationship is sort of transformed; it’s the fact that the object [of desire] completely disappears. [Its] link has been severed, wrecked. This is illustrated rather well in certain surrealist films from 1940-40 [sic], where we see money depreciating, inflation, for example. And there’s an element that sort of makes the object of desire completely disappear; some essential aspect gives rise to, lends itself to fascism as we know it. I suppose that after a given period of transformation, we more or less find the same process—different, but similar.

Gilles Deleuze: You began with the necessity of asking “why.” I understand everything you said afterwards, but I don’t quite understand why this question is necessary, and how it fits in with the rest of what you said. I think there’s no need to ask “why” because this whole system of machines is due to functionalism. If you ask “why,” you’re back to working with categories of signifiers; it’s a risky question. I believe there’s a domain—the domain of machines we call machines of desire, or desiring machines—where what we get is functionalism, that is, our only question is “How does it work?”—“How” and not “Why.” And this is why certain ethnologists are still in touch with how desire manifests in the social field—when they ask themselves, “What exactly is the purpose of psychoanalysis? Does it really serve our purposes as ethnologists?” And they say yes and no, because for us, as ethnologists, at the end of the day what we’re interested in has nothing to do with what something means—however one poses that question—it’s actually: “How does it work in the social field?”

Then it’s clear where the argument goes from there: never does the function or the use of something explain its production. For example, the way in which an organism functions could never explain that organism’s mode of production, or how an institution functions could never explain how the institution itself is formed. I think this anti-functionalist argument is spot-on, but in what context? [It’s spot-on] at the level of large aggregates, of molar aggregates.[1] There, indeed, how something is used always comes second to how something is formed. But if we’re trying to approach the unconscious in terms of machines, in terms of factories, in terms of units of production, in terms of desiring machines, I don’t think these are big machines; they are micro-machines. The machinic unconscious, the unconscious of desiring machines, is essentially a micro-unconscious, a micro-logical, micro-physical unconscious. Or, if you prefer, it’s a molecular unconscious.

Yet, in other words, at the level of molecular formations, as opposed to molar aggregates, strictly speaking, there is no possible difference between form and function. If I say how something works, there’s nothing else I can say. For schizo-analysis, the question isn’t at all — I recognize that the question “What does it mean” is rather complex because at a certain point, “What does it mean” might simply refer to a signified, or as they say, the whole question is knowing, when talking about the signifier, whether the question is still “What does it mean,” or whether it’s something else. Personally, I think it’s the same question. The question is still “What does it mean,” only crossed out, but it still falls under “What does it mean,” while desiring machines, strictly speaking, don’t mean anything—neither in terms of signifieds, nor in terms of signifiers — for schizo-analysis, the problem is: “What are your machines?” And that’s not easy to figure out.

A guy comes in; what are his desiring machines? It’s not enough to point out that he likes to drive, that he has a fridge—that’s all related to his desiring machines, but that’s not what his desiring machines are. On the other hand, neither are his desiring machines fantasies; they aren’t imaginary objects that function as doubles—it isn’t the dreamed car as the real car’s double… Desiring machines are molecular formations that exist objectively in larger technical machines and in larger social machines. That’s why, I think, we have to critique everything that’s dream, fantasy—the unconscious doesn’t dream, doesn’t fantasize, no more than the unconscious is a theatre. Dreams and fantasies are by-products of one’s reflection. They are territorialities of Oedipus; dreams are always oedipal. If analysts always stuck to dreams, they’d necessarily come back to Oedipus.

There’s a wonderful part of Bergson where he writes that it’s no surprise that matter and intelligence dovetail together, that matter and intelligence are tailormade for each other, since they both come from differenciating the same one movement, and well, it’s the same with Oedipus and dreams. It’s not surprising that all dreams are inherently Oedipal, because the same movement constitutes both dreams and Oedipus.

But that has nothing to do with desiring machines, so where things get tricky—I suppose—in schizo-analysis, is finding someone’s desiring machines. What’s the deal with your things, what are your machines? Then, if the person doesn’t answer in terms of machines, or if we can’t manage to find their machinic elements… obviously, that raises the question: what is our criterion? What allows us to say, ah, there, we’ve pinned down someone’s desiring machines? I think there are very solid criteria, and they specifically come down to this: desiring machines are molecular formations, never molar aggregates—but I get the feeling I’m not addressing your comment—all I’m saying is that, at the molecular level and only at the molecular level, functionalism reigns supreme, and that’s because function, production, form, are all one, strictly speaking. A desiring machine is defined only by how it functions, which is to say how it’s formed, which is to say how it’s produced. As it stands, I’d say that the question “why” never comes up.

Pinhas: I think maybe I didn’t word my question the right way. I wasn’t trying to question the fact that it works, or to reintroduce linguistic-psychoanalytic categories, etc. What I was trying to ask is, after the question of how, there are processes—I tried to offer an example of something I think fits the description: fascism—there’s why it works, in that sense, and then there’s why it works in another way, if you will. It works this way at one moment, and meanwhile, in another country, for example, it works another way. I think that, in that context, over and above the question how, we might afterwards ask the question why.

Deleuze: Yes, yes, yes, yes, yes. That’s a very particular “why,” then. It’s a “why” concerning the nature of the social field invested by desire and desiring machines. Where the social field’s characteristics come from, at such and such time, under such and such circumstances. Here, again, it’s pretty complicated. We’d need to maintain, like, two practical guidelines or principles, alongside there being desiring machines, which run along molecular lines, what I’d like to call lines of flight.[2] It’s not even enough to reach lines of resistance in the unconscious. The essential thing about the unconscious is that it escapes; it goes along lines of flight. Yet Oedipus, fantasies, dream, and what not—far from being real productions or formations of the unconscious—are all tourniquets, plugging up the unconscious’s lines of flight or escape. That’s why we have to blow all that up to reveal its lines of flight, which then lead us to a molecular unconscious of desiring machines. These are molecular formations, micro-formations, whereby it becomes a non-figurative, non-symbolic unconscious. It’s neither figurative nor symbolic—it’s what Lacan calls the Real, only not just a possible real, but a functional real.

Then, we have to simultaneously maintain, basically, two poles: one pole would be one like—it gets tricky; fortunately, at one level we’d distinguish two poles: the first being investments of molar aggregates, investments of the social field and everything that comes with that, including family investments, the family being one molar aggregate among others, a molar sub-aggregate.

And then, at the other end, there’d be molecular lines of flight, just as you might distinguish between macro-physics and micro-physics. Really, then, schizo-analysis would operate in terms of units of production of the micro-unconscious of small, molecular formations—that would have to be our first principle, but at the same time—so, again, we’d need to distinguish between investments of molar aggregates and investments of molecular formations, of desiring machines, on the other side of social and technical machines. Which is how we would define one of the main activities involved in the practice of schizo-analysis: nothing can get off the ground until we get to the desiring machines of someone’s unconscious, i.e., their molecular functioning formations. If we can’t reach them, we can’t do anything; we’re still dealing with big aggregates: Oedipus, the family, etc. As I see it, the signifier is a molar sign, fundamentally, a sign structuring big, molar aggregates, thus completely unrelated to desiring machines.[3]

So, that would be the first task for schizo-analysis: reaching the lines of flight of the unconscious. Which means the unconscious doesn’t express itself, doesn’t attract; it escapes, and it causes its desiring machines to function based on its lines of flight. As Plato might put it, when it draws near its opposite, it either retreats or perishes[4]—in the same way, the unconscious either perishes under Oedipus or escapes through its lines of flight.

But, from another perspective—one which doesn’t cancel out the first—we’d have to make what seems like the opposite point: that all investments, no matter what kind, are necessarily molar or social. Any investment is necessarily the investment of big aggregates; all investments are investments of big aggregates by molecular formations one can identify as desiring machines. No matter what, desiring machines—their parts and their cogs—invest big molar aggregates.

The second task of schizo-analysis would be to discover the nature of these social investments in a person, at the level of the unconscious. And I’m saying the two aren’t contradictory: in one, we say there are two positions: big aggregates defined by social and technical machines, and on the other side, molecular lines of flight and desiring machines. With the other, it’s that all investments are molar and social, just that social investments have two poles: a paranoid pole, which we could just as easily call fascist, reactionary investment, which consists in subordinating desiring machines to big, repressive apparatuses, to the big apparatuses of the state or of the family. The vulgarized psychoanalytical framework has to be flipped: paranoia is what comes first, what comes second is oedipal neurosis, and then third is Narcissus. The Oedipus complex is first of all a paranoid idea; it’s only secondarily a neurotic feeling, i.e., the neurotic is someone who’s been fooled, who’s been fooled by the paranoiac. It’s a father’s notion, and it’s not a father’s notion regarding his son—he doesn’t care about his son—the paranoiac is someone who starts hallucinating their social field by sadistically subordinating all their desiring machines to the repressive apparatuses of their social field.

At any rate, desiring machines are within real machines; they don’t resemble them, as in that absurd way of understanding dreams, and so on…it’s not about saying, ah, yes, uh, uh—it’s not about making psychoanalysis some kind of gizmo, uh, for market research, where we go, hmm, what dream car lies behind the real car? That’s not it. Desiring machines exist objectively. They exist in the cogs and parts of social machines and technical machines; they just can’t be extracted at the macroscopic level. That’s why it takes a microscopic analysis to draw out someone’s desiring machines. Someone who’s paranoid isn’t worried about their son. Schreber’s father is the paranoiac. You can see how his paranoia works, and in this regard, you can see the extent to which psychiatric understandings of paranoia are reactionary. If they didn’t link paranoia to family events, they’d be forced to acknowledge that paranoia, precisely because it has nothing to do with family relationships, is an inherently fascist, reactionary investment of one’s social field. The paranoid man’s delusions aren’t about his wife or about his son; his delusions are about race, education, culture. The schizophrenic too, but in a different way. At its root, the paranoiac’s delusion is “let’s create a pure race”: a sense that everything is in decline, that we’re no longer pure Aryans. Although he’s looking at his little kid, that’s only secondary. And there he applies—Oedipus is always an act of reduction [rabattement], application—he presents himself as the great pedagogue, the great restorer to the races. Schreber’s father started with a whole other dimension: the deviation of his social field. Our world’s been ruined; let’s recreate the pure race—and with what? With machines.

First, the paranoiac is delusional about the social field. I argue that paranoia ought to be defined in absolutely non-oedipal terms, as a particular way of investing the social field. A kind of investment of the social field, an investment that completely subordinates the system of desiring machines, or molecular formations, to greater molar aggregates. That in mind, in application to his family, as a privileged molar sub-aggregate, he applies his reform, his pure race, and produces an oedipalized little guy. Which is thus a product of earlier paranoid investment. It’s in this sense that the Oedipus complex is something experienced by the neurotic son inheriting his paranoid father’s big idea, but the latter’s idea isn’t oedipal; it consists in investing the entire social field.

The other idea is that everything is social investment, only it ranges between two poles: from paranoia, which broadly subordinates desiring machines to the apparatuses of large aggregates, to schizophrenia, schizophrenic escape, molecular lines of flight, desiring machines—it’s the same thing. And it’s no less deeply connected to the social field than paranoia’s sweeping integrations. It isn’t more delirious; it’s just a different sort of delirium. Because delirium has like two ends of the spectrum (constantly oscillating), and the schizophrenic side of delirium carries out the inverse sort of subordination, subordinating big molar aggregates to molecular formations. Which is why lines of flight don’t just consist in doing something; they also entail driving something away.

As a result, schizo-analysis basically has three tasks.

–  A destructive task: taking a big curette to the unconscious, i.e., has to blow up: Oedipus, dream, fantasy, myth, tragedy, no more theater—at the same time (not later on).

— Its first positive goal: reaching someone’s desiring machines. And you can never get at them directly; all we have are machinic indices, which might as well be practically hidden—that’s fine, that’s inevitable, there are only indices, not at all in the sense of “why” that I denounced earlier, but because they belong to a different regime, they’re micro-formations and only along lines of flight, or what amounts to the same, lines of decoding, of deterritorialization. The unconscious constantly deterritorializes itself; there are endless lines, lines of flight, abstract lines—again, neither symbolic, nor figurative, nor imaginary, nothing of the sort—lines of flight marked out with machinic indices. The only thing noteworthy about dreams is that they’re full of machinic indices. In dreams and fantasy, we get nothing but indices of desiring machines, and we get them in the form of molar machines; as often as daddies-mommies appear in dreams, so too do machinic indices. It isn’t a desiring machine because desiring machines aren’t dreamt, but they’re machinic indices. All the time, in fantasies, little machines appear; you have to follow their escape routes [voies de fuite] to get out of dream analysis; you have to delve into these kinds of machinic lines in order to go further.

— Alongside the work of scraping everything out, the sort of violent countermeasure to the unconscious pseudo-formations that have to be ruthlessly destroyed, stamped out, and the main positive task of uncovering desiring machines, the second positive task—there are criteria if you follow the lines of flight as far as you can—the second task is uncovering the nature of the social field’s unconscious investments, granted that the social field’s preconscious investments don’t work the same way. You could have rather revolutionary preconscious investments and still have paranoid, reactionary unconscious investments. What analysis needs is some equivalent to the revolution of abstract painting—that is, it needs to reach parts of the unconscious that are no longer figurative or symbolic: a molecular, abstract, machinic unconscious.

A student [comment about Schreber’s father]: It doesn’t start with the father’s generation; we have to go up a level (the grandmother), which is also quite something (dreaming of having a “real” man). You get the sense that it’s the paranoiac’s father who should be committed, and not his delusional son (delirium being the imaginary resolution of his fantasies). The one driving things is perfectly socially integrated, and it’s in that sense that he has delusions about the whole social field and then afterwards applies it to his son—but he’s firmly planted in his social field, which is why he avoids getting sent to the asylum. Whereas those who end up in the asylum, like paranoiacs, are those who aren’t connected enough to be spurred on by familial investments; there’s a barrier that absolutely prevents them from having any delusion about the social field. It takes three generations to make someone paranoid; as for schizos, who’s to say?

Deleuze: I take issue with your wording, but that’s beside the point. If there were only paranoiacs, there never would have been asylums. The paranoiac gets along fine. The paranoiacs who do get hospitalized are either really poor, so they think they’re little leaders, or else it’s because of schizoid elements that always get mixed in with paranoia. If they’re hospitalized, it’s certainly not because of their paranoia; it’s owing to some latent schizoid element, lying beneath the paranoia.

[Maud] Mannoni’s argument about the decision on Schreber’s case as the first act of anti-psychiatry misses the mark, since the reason Schreber was released was that his schizophrenic elements did not prevail.[5] I’m reminded of another case, perhaps the opposite of Schreber’s, of someone with quite pronounced paranoid aspects and even more pronounced schizophrenic tendencies: Nijinsky, the dancer. No way the court would let Nijinsky go free—he went around shouting, “I am the clown of God, death to the Stock Market, the Stock Market is death, money is death!”[6]  You could argue that saying that is unreasonable since it was just said here.

With rich paranoiacs, it’s great; they have a fundamental social function. The paranoiac is well integrated, and no one who’s paranoid will get committed to an asylum unless they fall into one of those two categories.

My second point addresses matters regarding the three family members. Looking at the three generations, we find with the most ardent anti-psychiatrists—Gisela Pankow, [R.D.] Laing, [David] Cooper—what do we end up with? When it comes to psychosis, the Oedipus complex obviously falls apart, so the question is how to salvage Oedipus. As [Guy] Rosalato naively puts it: “How do we bring psychosis back into an Oedipal framework?”[7] Posed in that way, the problem can play out in one of several ways. We can take a structuralist route—which isn’t at all what Lacan is doing—we can use structuralism to form a structural Oedipus complex and figure out where psychosis fits into said structure. Or—these aren’t mutually exclusive—we can extend Oedipus. If the Oedipus complex doesn’t work, we widen it a bit, i.e., we bring in the grandmother, the grandfather; we argue that Oedipus no longer involves two generations but three. We’ll have to account for the grandfather. If that doesn’t work, we’ll bring in a fourth generation.

If I’m privileging the role of the father or the role of the grandfather, it’s certainly not for the sake a new take on Oedipus. At a certain point, I could ask: what comes first, the father or the child? That is, the chicken-or-egg questions doesn’t make any sense—but at the same time, it must be addressed. The father or mother comes before the child. If I claim that the father has primacy over the child, that can be interpreted in one of two ways. [First,] regressively, as an infinite regress—every child has a father—we can keep climbing further and further back up to the presupposition of a first father, the Father of the primal horde, for example. But there’s another, completely different way of interpreting it, in a way where we don’t get caught up in an indefinite familial reduction. It could mean—and this is where the question becomes nonsense—what comes first is actually the social field covering both father and child, who are simultaneously immersed in said social-historical field. In truth, the idea that the father has primacy over the child means that social investments come before familial investments.

That means understanding the unconscious as a cycle, as per Marx’s well-known response to Aristotle regarding the “cause of man [sic]”—yes, but there’s a cycle, the cycle whereby man produces man. The unconscious of desiring machines is a cyclical unconscious. Accounts of schizophrenogenic families fail to explain anything about how schizophrenia is produced; the most ordinary family mechanisms get portrayed as schizophrenogenic mechanisms. The father has primacy over the child, but not as a father—it’s that the social field and social investments have primacy over father and son.

Why do we only find indices of the little desiring machines that invest the entire social field? These machinic lines of flight are necessarily lines of deterritorialization as such, because deterritorialization is like the inverse or opposite of the movements of reterritorialization. Even Beckett’s most extreme protagonists can’t completely deterritorialize themselves; they integrate small patches of land[8], Malone’s room, a trash can. The movement of deterritorialization can only be grasped through the gene and the nature of the reterritorializations carried out by an individual.

It’s always by way of a reterritorialization’s countermovement that we can determine the degree of deterritorialization. For example, the neurotic has already found a foothold to cling onto: Oedipus. The analyst’s couch is the second thing that won’t budge, the bit of land one must cling to—otherwise everything gets rocky.

We have to demonstrate how capitalism never stops deterritorializing and, with its axiomatics, reterritorializing. For example, fascism was a process of reterritorializing the masses, however terrible. You can only read someone’s deterritorialization—and how far it’s gone—that is, someone’s schizophrenic terror, through their counter-terrors, the reterritorializations they carry out. That’s why we shouldn’t think of perverts in terms of their impulses. We should think of them in terms of their territory: perverts are folks who want neither the territoriality of Oedipus nor the territoriality of the couch. That won’t do—they invent artificial territory, artificial groups. They have their own way of reterritorializing themselves, and if all else fails, as a last resort, you reterritorialize yourself as a body without organs, i.e., lying catatonic in the hospital, the poorest territory of all. You’ve recreated your little foothold.

What matters is that the movement of deterritorialization isn’t simply susceptible to getting taken up in a perverse reterritorialization, be it psychoanalytic or perverse in the strict sense, but that the movement of deterritorialization is strong enough to—cleaving to its revolutionary lines of flight—create for itself a new type of land. Perhaps that’s what Nietzsche means when he says that one day “the earth shall become a place of healing.”[9] Perhaps, rather than being reterritorialized on artificial land, under certain conditions, the movement of deterritorialization might come to be the creator of a new earth. It would be nice, at any rate.

Eric: Three generations are relevant. It doesn’t go any further than that. There’s been a psychotic episode—there must be a mechanism there linking it to something, the “name of the father,” triggering the patient’s hospital psychosis. The same way you distinguish between schizophrenia and hospital schizophrenia, there’s also a difference between psychosis and hospital psychosis. Hence Lacan’s work, which demonstrates the role of exclusion, which is linked to the father in a certain way—not the real father, but the father as an investment of the social field—in this idea of hospital paranoia.

Deleuze: We’ll have to pick back up on this next week. I disagree.

 

Notes

[1] Per Hurley, Seem, and Lane’s translation, ensemble here has been translated as “aggregate.” See, for example, Gilles Deleuze and Félix Guattari, Anti-Oedipus: Capitalism and Schizophrenia, trans. Robert Hurley, Mark Seem, and Helen R. Lane (Minneapolis: UMP, 1983), p. 181.

[2] Ligne de fuite is a ubiquitous term in Deleuze and Guattari’s work (and in the scholarship surrounding their work). In addition to translating fuite as “flight,” it also appears here as “escape,” or “getting out.” For readers with a background in electrical engineering, ligne de fuite is a term for “leakage distance” or “creepage distance.”

[3] The tasks of schizo-analysis are defined in Anti-Oedipus, chapter IV.

[4] Phaedo 102d.

[5] See Anti-Oedipus, p. 364. Deleuze and Guattari cite Chapter 7 of Maud Mannoni, Le Psychiatre, son fou et la psychanalyse.

[6] Deleuze and Guattari quote Vaslav Nijinsky’s Journal, or Diary (Paris: Gallimard, 1953), in Anti-Oedipus, p. 77.

[7] A possible reference is to Guy Rosalato’s Essais sur le symbolique (Paris : Gallimard, 1969), cited in Anti-Oedipus, p. 210.

[8] In this translation, terre appears as a patch of land, a foothold, a territory—for the sake of more comfortable English, the translation may obscure the word’s connection to Deleuze and Guattari’s paired terms, “deterritorialization” and “reterritorialization.”

[9] Thus Spake Zarathustra.

Notes

The transcript was developed by WebDeleuze and modified in March 2023, and the translation was completed in March 2023, with additional revisions and descriptions completed in September 2023.