Thursday, December 30, 2004

Reading :: A Thousand Plateaus (second attempt)

Originally posted: Thu, 30 Dec 2004 06:26:15

A Thousand Plateaus: Capitalism and Schizophrenia
By Gilles Deleuze, Felix Guattari

I first read this book -- or rather, a big chunk of it -- just over a year ago, and my review was not entirely positive. "I found the book to be excruciating," I confessed, "and abandoned it on p.178." But in that intervening year I found myself recommending it to a variety of people. The idea of the rhizome was too valuable and kept making too many connections with other things I was seeing. So the realization eventually sunk in that I would have to go back and read the whole thing. I read the Gulag Archipelago when I was 15, I told myself, so I should be able to make it through this thing.

And I did. The book is still maddening, but now I have a better idea why. And although I still stand by the general criticisms of my previous review, I do understand the authors' project much better. Sit back, because this is going to be a long review. With subheadings.


Jenny encouraged me more than once to finish this book. "Play it like a record," she told me more than once, moving her hand as if rotating a turntable, encouraging me to take seriously Deleuze and Guattari's claim that the book can be read in any order because it is a rhizome. If this book were a record, it would be something like Skinny Puppy's Brap or Severed Heads' Since the Accident: full of starts and stops, looping, tempo changes, motifs, unexpected juxtapositions, and refrains that appear to be external references but that enter into a new internal relationship full of tensions. These are two of my favorite albums -- but maybe that's because they don't take weeks to get through.

I listen to these records the way that I read books: straight through. So I "played it like a record."


Many of these refrains in this book, though, really are references to other work. For instance, take this passage from the allegorical chapter on "the Geology of Morals":

The professor cynically congratulated himself on taking his pleasure from behind, but the offspring always turned out to be runts and wens, bits and pieces, if not stupid vulgarizations. (p.42)

You see a lot of offhand (and unpleasant) statements like these in A Thousand Plateaus, statements that first appear to come from nowhere. They are so frequent that sometimes I felt that the book was authored with the French Poststructuralist version of Mad Libs. But it turns out that this is a direct reference to Deleuze's scholarship. In response to an anonymous critic, he once said:

But I suppose the main way I coped with [writing about historical philosophers] at the time was to see the history of philosophy as a sort of buggery or (it comes to the same thing) immaculate conception. I saw myself as taking an author from behind and giving him a child that would be his offspring, yet monstrous. It was really important for it to be his own child, because the author had to actually say all I had him saying. But the child was bound to be monstrous too, because it resulted from all sorts of shifting, slipping, dislocations, and hidden emissions that I really enjoyed. (Negotiations p.6)

What this tells me -- besides the fact that Deleuze thought of his scholarship in terms of sexually administered comeuppance, which I think is fairly creepy -- is that A Thousand Plateaus probably isn't the best Deleuze or Guattari book to start out with. It was the capstone of their individual and collaborative work, I think, and I'm quite positive that there were hundreds of references like this one. As I said in the previous review, you have to work your way into the system all at the same time, and that's because the authors deliberately choose not to frame the text or situate it. (They lie about its rhizomatic nature -- there is definite progression in the book, and the conclusion is genuinely a conclusion -- but that progression is obscured and each part of the book refers to concepts developed in the other parts and in previous scholarship.) And as we all knew, the authors have made no real attempt to make the book very readable. Here are two more quotes from Negotiations. The first is about Anti-Oedipus, their previous collaboration:

There are, you see, two ways of reading a book: you either see it as a box with something inside and start looking for what it signifies, and then if you're even more perverse and depraved you set off after signifiers. And you treat the next book like a box contained in the first or containing it. And you annotate and interpret and question, and write a book about the book, and so on and so on. Or there's the other way: you see the book as a little non-signifying machine, and the only question is "Does it workl, and how does it work?" How does it work for you? If it doesn't work, if nothing comes through, you try another book. This second way of reading's intensive: something comes through or it doesn't. There's nothing to explain, nothing to understand, nothing to interpret. (Negotiations p.8).

I guess I am implicated here, since I am after all writing copious annotations and a little review essay. And here Deleuze discusses A Thousand Plateaus, then in progress:

We'll do the sequel because we like working together. Only it won't be anything like a sequel. With a bit of help from outside, it will be something so different in its thinking and language that anyone "waiting" for us will have to say we've gone completely crazy, or we're frauds, or we couldn't take it any further. It's a real pleasure to confound people. Not that we just want to play at being mad, but we'll go mad in our own way and in our own time, we won't be pushed into it. ... We're going to stop compromising, because we don't need to anymore. And we'll always find the allies we want, or who want us. (p.9)

A Thousand Plateaus, then, was an experiment and they knew that it could be taken as mad. Well, mission accomplished. The book is tremendously self-indulgent in this respect, written for those in the loop and/or for the authors themselves, and written to confound the critics. Consequently, there's a lot of air here. I imagine that the Cliff's Notes version of this book would be about 60-80 pages. This review is considerably shorter, but I'll hit the highlights and try to make things clearer -- realizing the entire time that clarity is counter to the authors' intention. And I'm quite certain that I've missed a lot of things here, but that's what the comments section is for.


Let us summarize the principal characteristics of a rhizome: unlike trees or their roots, the rhizome connects any point to any other point, and its traits are not necessarily linked to traits of the same nature; it brings into play very different regimes of signs, and even nonsign states. The rhizome is reducible neither to the One nor the multiple. ... It is composed not of units but of dimensions, or rather directions in motion. It has neither beginning nor end, but always a middle (milieu) from which it grows and which it overspills. (p.21)

As I noted in the previous review, the rhizome is an antigenealogy (p.11, 21), an antimemory (p.21), "an acentered, nonhierarchical, nonsignifying system without a General and without an organizing memory or central automation, defined solely by a circulation of states" (p.21). But I missed the most important point, which is right there in capital letters: "RHIZOMATICS = SCHIZOANALYSIS = STRATOANALYSIS = PRAGMATICS = MICROPOLITICS" (p.22). Yes, rhizomatics is pragmatics. The veil lifted when I got to this point. Deleuze and Guattari's project is an exercise in pragmatics, a monist, materialist attempt to make sense of the world by examining interactions. It is, as Andrew Pickering claimed of his own pragmatic project, a "theory of everything" (TOE). Much of what follows in the book has to do with this essentially ecological understanding of the world and how it applies to materials as well as signs. For instance, in a later chapter the hand is described as a foot that has been "deterritorialized." Biological evolution is described using the same terms as writing and economics. In my previous review, then, I drastically underestimated the project: it's not a semiotics, it's a TOE.

And so it makes sense at the end of this chapter that the authors finally tell us that "the tree is filiation, but the rhizome is alliance, uniquely alliance" (p. 25). Yes, no wonder Latour is so intrigued by rhizomes, since alliance of humans and nonhumans is his obsession. Filiation/lineage/arborescence exists, but rhizomatic alliances among heterogeneous elements are what causes the whole mess to work.


This short chapter, a critique of Freudianism, introduces the notion of multiplicity. Freud's problem, the authors seem to say, was that he believed in essential coherence rather than relative coherences. He was always looking for the root of the problem, which always led to the same thing -- and away from the discovery of the rhizome, away from the discovery of multiplicities, which are "lines of flight or deterritorialization, becoming-wolf, becoming-inhuman, deterritorialized intensities" (p.32) -- that is to say, not simply different representations of the same thing, but different positionings of material assemblages that may have relative coherence. (For more, I commend Annemarie Mol's The Body Multiple and my review of it.) Here's the important part: Multiplicity

was created precisely in order to escape the abstract opposition between the multiple and the one, to escape dialectics, to succeed in conceiving the multiple in the pure state, to cease treating it as a numerical fragment of a lost Unity or Totality or as the organic element of a Unity or Totality yet to come, and instead distinguish between different types of multiplicity. (p.32)

Yes, multiplicity is a way to escape dialectics. We forget about dualisms, absolute coherence, and most of all dialectical syntheses. Alert readers will sense a parallel to Bakhtin's project, which I'll discuss in a moment.

Multiplicities can be arborescent (macro; extensive, divisible, molar, unifiable, totalizable, organizable, conscious or preconscious), or they can be rhizomatic (libidinal, unconscious, molecular, intensive, "composed of particles that do not divide without changing in nature" (i.e., molar) "and distances that do not vary without entering another multiplicity and that constantly construct and dismantle themselves in the course of their communications." Rhizomatic multiplicities are composed of particles, "their relations are distances; their movements are Brownian; their quantities are intensities, differences in intensity" (p.33).

Citing Elias Canetti, the authors provide another distinction: mass (crowd) multiplicities and pack multiplicities. "The leader of the pack or the band play move by move, must wager everything every hand, whereas the group or mass leader consolidates or capitalizes on past gains" (p.34). The first is schizo, the second is paranoid (p.35).

Nevertheless, the different types are not dualistically opposed: "There are only multiplicities of multiplicities forming a single assemblage, operating in the same assemblage" (p.34). And "there are no individual statements, there never are. Every statement is the product of a machinic assemblage, in other words, of collective agents of enunciation (take 'collective agents' to mean not peoples or societies but multiplicities)" (p.37).

And this is the crux of the authors' criticism of psychoanalysis. Finally, some light breaks through: "We are criticizing psychoanalysis for having used Oedipal enunciation to make patients believe they would produce individual, personal statements, and would finally speak in their own name. ... at the very moment the subject is persuaded that he or she will be uttering the most individual of statements, he or she is deprived of all basis for enunciation" (p.38). Psychoanalysis is looking for an absolute coherence, an absolute subject, an individual to diagnose. It fails to understand the social, dialogic relations that produce individual voices.

See Voloshinov's Freudianism for another stab at this basic point.


This is where the going really starts to get tough. The chapter is written allegorically, as a lecture delivered by the fictional Prof. Challenger: "The same Professor Challenger who made the Earth scream with his pain machine, as described by Arthur Conan Doyle, gave a lecture after mixing several textbooks on geology and biology in a fashion befitting his simian disposition" (p.40). Challenger's lecture uses geology as an extended metaphor for understanding how parts of the material world cohere.

He says: the Earth is "permeated by unformed, unstable matters" (p.40) which tend to flow in all directions. But it folds strata, "acts of capture" (p.40) which lock some of these matters into relatively coherent and stable aggregates. "They operate by coding and territorialization upon the Earth" (p.40), which is to say that the components of strata enter into mutual stable material-semiotic relationships. But at the same time the Earth continually becomes destratified and deterritorialized.

The Professor doesn't give a good example here, so I'll give it a try. Let's take Bateson's famous illustration of the blind man with the cane. (Bateson is one of Deleuze and Guattari's favorites, and his notion of the double bind figures prominently in this chapter.) In Steps to an Ecology of Mind (which I really need to reread), Bateson asks: is a blind man's cane part of him? When the blind man taps his stick along the sidewalk and feels a curb, where is the curb perceived? At the end of the stick, where the stick reaches the hand, or elsewhere? The answer is that the whole has to be understood as a system of blind man + cane + curb; it's a molar unit. This molar unit is an assemblage. The assemblage is not just a collection of atomic elements ("man," "stick," "curb"): each of these can be sliced or interconnected infinitely. A blind man's cane has its own components, and these interact with other assemblages; it's simultaneously part of a variety of them. It can be rearticulated, used for support (by a lame man) instead of sense (by a blind man). Assemblages make the unity of composition organized rather than random (p.71) -- they make sense of a heterogeneous jumble of infinitely recombinable parts, not just semiotically, but functionally (cf. multiplicity). Such assemblages are the surfaces of strata, which I take to be a sort of material-cultural space; it provides a unity of composition that "relates to formal traits common to all the forms or codes of a stratum, and to substantial elements, materials common to all the stratum's substances or milieus" (see the conclusion, p.502). Strata "consist of giving form to matters, of imprisoning intensities or locking singularities into systems of resonance or redundancy, of producing upon the body of the earth molecules large and small and organizing them into molar aggregates" (p.40) -- molar aggregates such as the blind man + cane + curb assemblage, I presume. The stratum might also include the city streets, the attitudes toward the blind, the significance of the cane's color (white), etc.

Strata come in pairs and are doubly articulated. "The first articulation chooses or deducts, from unstable particle-flows, metastable molecular or quasi-molecular units (substances) upon which it imposes a statistical order of connections or successions (forms). The second articulation establishes functional, compact, stable structures (forms), and constructs the molar compounds in which these structures are simultaneously actualized (substances)" (p.40). For instance, the authors say, think in terms of geological formation. In sedimentation, materials are cyclically laid down in a statistical order, separated from each other (layers). In folding, these layers are metamorphosed; it "sets up a stable functional structure and effects the passage from sediment to sedimentary rock" (p.41). Sedimentation forms a stratum, folding cements it.

A few things here about sedimentation and folding.

One, notice that here is the account of development and relative stability missing from rhizomes. Latour criticized rhizomes a while back for having no account of development or stability, but an account of sorts does seem to be here.

Two, the example of geology may appear to be a metaphor, but I take the authors at their word when they call it a "relatively simple case" (p.40, my italics). Remember, we're talking about a TOE here, an account that encompasses material, semiotic, and cultural aspects. It clearly encompasses evolution (we'll talk about that in a moment), so why not geology as well?

Three, their statement that "substances are nothing other than formed matters" recalls Bakhtin's statement that form is congealed content. I think the authors are often in sync with Bakhtin, although his project has a considerably narrower scope.

This double articulation is associated with Bateson's double bind. It connotes a dynamic tension between aggregation and cementing, or cycling and stabilizing. I would say that it's similar to the notion of contradiction, except that it's far more multidimensional and multivalent. Let's return to the blind man-white cane-curb assemblage. Suppose you take this assemblage into fiction, give the blind man heightened senses, and turn him into a superhero (Daredevil). The assemblage drifts a bit but still remains relatively stable. Daredevil is still a blind man; he still navigates city streets (on rooftops now, not curbs); and he taps criminals rather than curbs. He doesn't need a cane to navigate, but he still carries a white billy club -- the assemblage is still relatively stable, you see -- and it still functions as a tool. The club has been partially deterritorialized, partially dissolved from the assemblage and reconstituted in a different position, but it remains.

Let's take another example.

Not only is the hand a deterritorialized front paw; the hand thus freed is itself deterritorialized in relation to the grasping and locomotive hand of the monkey. The synergistic deterritorializations of other organs (for example, the foot) must be taken into account. So must correlative deterritorializations of the milieu: the steppe as an associated milieu more deterritorialized than the forest, exerting a selective pressure of deterritorialization upon the body (it was on the steppe, not in the forest, that the hand was able to appear as a free form, and fire as a technologically formable matter). Finally, complementary reterritorializations must be taken into account (the foot as a compensatory reterritorialization for the hand, also occurring on the steppe). Maps should be made of such things, organic, ecological, and technological maps one can lay out on the plane of consistency. (p.61)

What is deterritorialized on one plane is reterritorialized on another (p.54). The vision is one of ecological change, or a complex system -- though system implies too much in terms of preexisting forms. I think Deleuze and Guattari are getting at the notion that we live in a heterogeneous world with no a priori divisions or distinctions, but pragmatic distinctions can be made in multiple ways, both interpretive and functional. And like Latour, they're trying to use a common (symmetrical) language for both.

Also like Latour, they are interested in science as an abstract system that overcodes the other strata through language. Language has temporal linearity, or superlinearity, that distinguishes it from the spatiality of genetic linearity. The resulting "linear overcoding" allows translation, the ability of language "to represent all the other strata and thus achieve a scientific conception of the world. The scientific world ... is the translation of all the flows, particles, codes, and territorialities of the other strata into a sufficiently deterritorialized system of signs, in other words, into an overcoding specific to language" (p.62). One plane, one abstract system of representation. Science derives its power from this deterritorialization of planes into sign systems. (Remember Latour's argument that science is not performed on the world, but on representations of that world: tables of figures, diagrams, maps. And these representations are brought into being through acts of translation.)

The authors distinguish among three kinds of signs: indexes (territorial signs), symbols (deterritorialized signs), and icons (signs of reterritorialization) (p.65). These signs mark thresholds crossed in the movements of de - and reterritorialization (p.67). They occupy regimes of signs (p.65), relatively stable sign systems extracted from strata in which expression and content have been categorically separated (p.65). It's only in a regime of signs that a sign can be called a signifier (p.68). Only in a stratum that has codified the blind man-cane-curb assemblage, for instance, does Daredevil's white club signify blindness.

Content and expression cannot be reduced to sign-signifier. Regimes of signs "express organizations of power or assemblages and have nothing to do with ideology as the supposed expression of a concept (ideology is a most execrable concept obscuring all of the effectively operating social machines)" (p.68).

Writing this review, I realize that most of the major concepts of the book are discussed in this allegory. No wonder it was so inpenetrable. I haven't even gotten to the body-without-organs, which I'll tackle under Chapter 6 below.


"Language is made not to be believed, but to be obeyed, and to compel obedience" (p.76). This observation kicks off the chapter on linguistics, and you can see how it derives from Chapter 3's discussion of language as a deterritorialized, abstract system that represents other strata. Language makes sense of the world by fixing relationships, even though language itself is prone to the drifts and destabilizations that are present elsewhere.

If that sounds like warmed-over Bakhtin to you, you're partially correct. The authors approvingly quote Bakhtin here -- actually Voloshinov's Marxism and the Philosophy of Language, which was translated into French under Bakhtin's name. In fact, I was perplexed that they didn't quote from some of Bakhtin's other relevant work since the parallels were so obvious. But then I realized that it's likely Bakhtin's other works had not yet been translated into French at the time they were writing. I'll try to point out parallels with Bakhtinian thought as I go through the chapter.

(By the way, it's ironic that the authors, who are apparently fond of difficult prose, have chosen the most lucid book of the Bakhtin circle to cite. Perhaps that's why this chapter is one of the most lucid in the book.)

The vision of this chapter is what I would characterize as dialogic in the Bakhtinian sense:

Language in its entirety is indirect discourse. Indirect discourse in no way supposes direct discourse; rather, the latter is extracted from the former, to the extent that the operations of signifiance and proceedings of subjectification in an assemblage are distributed, attributed, and assigned, or that the variables of the assemblage enter into constant relations, however temporarily. Direct discourse is a detached fragment of a mass and is born of the dismemberment of the collective assemblage; but the collective assemblage is always like the murmur from which I take my proper name, the constellation of voices, concordant or not, from which I draw my voice. I always depend on a molecular assemblage of enunciation that is not given in my conscious mind, any more than it depends solely on my apparent social determinations, which combine many heterogeneous regimes of signs. (p.84)

The authors go on to say that self emerges from this assemblage. This all sounds suspiciously Bakhtinian; replace "regimes of signs" with "social languages" and "murmur" with "dialogue," and there you are. But the authors are interested in linguistics as part of a TOE, remember, so they go farther than Bakhtin in discussing language as material. "Representations are bodies too!" they exclaim (p.86). And: "An assemblage of enunciations does not speak 'of' things; it speaks on the same level as states of things and states of content" (p.87). Perhaps this is the answer to Pickering's complaint that Latour shuttles between the semiotic and the performative: the distinction is unproductive because signs can be both bodies and acts (p.87).

This brings us back to the distinction between content and expression that evidently preoccupies the authors. Expressions do not uncover or represent content; these forms "communicate through a conjunction of their quanta of relative deterritorialization, each intervening, operating in the other" (p.88). On one axis, an assemblage is segmented into content -- "a machinic assemblage of bodies, of actions and passions, an intermingling of bodies reacting to one another" -- and expression -- "a collective assemblage of enunciation, of acts and statements, of incorporeal transformations attributed to bodies" (p.88). On another axis, the assemblage is segmented into territorial or reterritorialized sides that stabilize it and cutting edges of deterritorialization that carry it away (p.88). The second axis sounds quite a bit like Bakhtin's distinction between centripetal and centrifugal forces, with the added point that centrifugal (deterritorializing) forces carry things away into other assemblages. And of course the first axis sounds a lot like a monist, materialist explanation of what Bakhtin calls dialogue.

The authors give a nice example using feudalism, once again mingling materials and signs in an undifferentiated (symmetrical) fashion. Here's a nice quote about how technologies fit into the picture:

Even technology makes the mistake of considering tools in isolation: tools exist only in relation to the interminglings they make possible or that make them possible. The stirrup entails a new man-horse symbiosis that at the same time entails new weapons and new instruments. Tools are inseparable from symbioses or amalgamations defining a Nature-Society machinic assemblage. They presuppose a social machine that selects them and takes them into its "phylum": a society is defined by its amalgamations, not by its tools. Similarly, the semiotic or collective aspect of an assemblage relates not to a productivity of language but to regimes of signs, to a machine of expression whose variables determine the usage of language elements. (p.90)

(Latour calls the Nature-Society machinic assemblage "NatureCulture.")

This brings us back to linguistics, or pragmatics in particular. The authors distinguish between external pragmatics (nonlinguistic) and internal pragmatics (linguistic) (p.91). Pragmatics, they point out, can't be conceived as "pretreated by a phonological or syntactical machine" -- pragmatics must be recognized as emergent from the "abstract machine," the totality of the assemblage being examined. Once we understand pragmatics in this way, as superlinear rather than placing elements in a fixed linear order, we achieve a rhizomatic understanding in which "the interpenetration of language and the social field and political problems lies at the deepest level of the abstract machine, not at the surface" (p.91). Again, rhizomatics = pragmatics and the equation necessitates a materialist, monist, symmetrical understanding of the world.


Now we look at regimes of signs in depth. Here, the authors define a regime of signs as constituting a semiotic system and as a "specific formalization of expression" (p.111). Signs are all about relationships; what is important about the sign is not itself but its associations (p.112). The authors reiterate the three types of signs they discussed in Chapter 3, indexes (territorial), icons (reterritorialization), and symbols (deterritorialization). They form a network or circle: "The question is not yet what a given sign signifies but to which other signs it refers, or which signs add themselves to it to form a network without beginning or end that projects its shadow onto an amorphous atmospheric continuum" (p.112). It's hard to build on this shifting sand, which means that paranoiacs in particular look for the master signifier that is the key to everything: "The sign that refers to other signs is struck with a strange impotence and uncertainty, but mighty is the signifier that constitutes the chain. The paranoiac shares the impotence of the deterritorialized sign assailing him from every direction, but that only gives him better access to the superpower of the signifier, through the royal feeling of wrath, as master of the network spreading through the atmosphere" (p.112). (This reminds me of Foucault's Pendulum and conspiracy theories.)

"But," the authors add, "what counts is less this circularity of signs than the multiplicity of the circles of chains. The sign refers not only to other signs in the same circle, but to signs in other circles of spirals as well" (p.113). And a regime of signs must assume this task of organizing signs into circles -- but it also "must provide the center with more signifier to overcome the entropy inherent in the system and to make new circles blossom or replenish the old" (p.114). This center of significance, the Signifier, is "nothing" -- a pure abstraction or principle. They call this redundancy of the signifier faciality and say that the face is what "releases and recaptures signifying signs" (p.115). That is to say, the face is the organizing principle of a regime of signs, that which collects and reterritorializes signs. When "the face is effaced" and "the faciality traits disappear," we have entered another regime (p.115).

Notice how this slots into the question of multiplicity. In a universe or "amorphous atmospheric continuum" of signs that float free, deterritorialized, we can make sense of things only by looking for a network of signs -- the network is the signifier, individual signs are not, and the network works because the deterritorialized signs become reterritorialized by their relationship to a coherent center. That center is "nothing" -- think of a tangle of string, whose center is nothing but more string, or a knot in a network, which is simply the network's material wrapped around itself. Or perhaps a dust bunny. We use such ephemeral centers to segment and organize the environment. But signs can belong to more than one network or signifier, as Annemarie Mol demonstrates so well in her discussion of atherosclerosis: in the hospital, the signifier "atherosclerosis" doesn't have a one-on-one correspondence with a particular object, it is constituted by an entire network of signs, materials, and enactments. And as Mol's study suggests, this is not merely a semiotic explanation. Signs are also material; they encompass everything, literally (p.117). (This is a TOE, after all.)

The authors compare a signifying regime to other types of regimes:

Presignifying. "There is no reduction to faciality as the sole substance of expression: there is no elimination of forms of content through abstraction of the signified" (p.117). Forms are plural and polyvocal. Segmentarity is plurilinear. Think of hunter nomads, the authors say.

Countersignifying. "This time, the semiotic proceeds less by segmentarity than by arithmetic and numeration" -- not arithmetic that is useful for dividing and unifying segmentary lineages, but arithmetic that refers to itself, "which itself determines functions and relations, which arrives at arrangements rather than totals, distributions rather than collections, which operates more by breaks, transitions, migration, and accumulation rather than by combining units" (p.118). Such a number is directed against the State apparatus. Think of warlike, animal-raising nomads, the authors say, giving Moses as an example.

Postsignifying. These regimes can all appear in each other and mix together (p.119). The postsignifying regime intersects with the signifying regime. It "is defined by a decisive external occurrence, by a relation with the outside that is expressed more as an emotion than as an idea, and more an effort or action than imagination" (p.120). "It operates by the linear and temporal succession of finite proceedings, rather than by the simultaneity of circles in unlimited expansion" (p.120). A signifying regime is paranoid and despotic; a postsignifying regime is passional, subjective, and authoritarian. The line of deterritorialization -- when a "packet of signs detaches from the irradiating circular network" -- is given a negative value in the signifying regime, but a positive one in the postsignifying regime (p.122). Again we return to Moses: "the paranoid Pharaoh and the passionate Hebrew." And the scapegoat: Israel became the scapegoat, following its line of flight, turning that line into a positive line (p.122). This postsignifying regime is related to the nomadic countersignifying regime: Israelites become warlike nomads who toppled States. But it is also related to the signifying regime: Israelites were nostalgic for Egypt and repeatedly tried to reestablish an imperial society (pp.122-123).

In the postsignifying regime, "faciality undergoes a profound transformation. The god averts his face, which must be seen by no one; and the subject, gripped by a veritable fear of the god, averts his or her face in return. ... It is this double turning away that draws the positive line of flight" (p.123). This regime is a regime of betrayal, in which man and God continually betray one another. Existence is existence in reprieve, indefinite postponement (p.123).

In this sort of passionate regime, the book becomes the sacred book, taking the place of the face of God (think of the ten commandments on the stone tablet). "It is now the book, the most deterritorialized of things, that fixes territories and genealogies" (p.127).

The chapter goes on like this, but let's skip a bit -- I'm not terribly interested in discussing some of the more detailed categorization. But I found it interesting that the authors equate absolute deterritorialization to "the earth iself" -- which I take to mean the entirety of existence, material in infinite semiotic connection, from which we must build regimes to "conjugate" and make sense (p.143). No universal logic exists; Abstract machines, diagrammatic functions, and machinic assemblages lead to regimes of signs, which in turn lead to languages. Thus, the authorws say, pragmatics is the fundamental element of logic, syntax, and semantics (p.148).

What I began to realize in this chapter is that the authors do indeed have a systematic way of discussing and analyzing the cases they do. Despite their frequently metaphorical language, they are trying to get at a TOE, and they see pragmatics (which, recall, is rhizomatics) as the key building block. Regimes of signs give rise to languages, but I think they are very close to what Bakhtin calls social languages.


Now this was a frustrating chapter to read -- the most frustrating in the book, even more so than Chapter 3. "So what is this BwO?" the authors ask hypothetically (p.150). Well, that's a good question. (One of my notes earlier in the book simply says: "BwO -- WTF?!?")

Let's take a crack at this question. Among the descriptions of psychological disorders and masochistic fantasies, the authors become relatively lucid here:

A BwO is made in such a way that it can be occupied, populated only by intensities. Only intensities pass and circulate. Still, the BwO is not a scene, a place, or even a support upon which something comes to pass. It has nothing to do with phantasy, there is nothing to interpret. The BwO causes intensities to pass; it produces and distributes them in a spatium that is itself intensive, lacking extension. It is not space, nor is it in space; it is matter that occupies space to a given degree -- to the degree corresponding to the intensities produced. It is nonstratified, unformed, intense matter, the matrix of intensity, intensity = 0; but there is nothing negative about that zero, there are no negative or opposite intensities. Matter equals energy. Production of the real as an intensive magnitude starting as zero. That is why we treat the BwO as the full egg before the extension of the organism and the organization of the organs, before the formation of the strata; as the intense egg defined by axes and vectors, gradients and thresholds, by dynamic tendencies involving group displacement, by migrations: all independent of accessory forms because the organs appear and function as pure intensities. The organ changes when it crosses a threshold, when it changes gradient. (p.153)

All clear? Recall that in Chapter 3, the Earth is described as a BwO (p.40). Referring to Bateson's term "plateau," the authors add that "every BwO is a plateau in communication with other plateaus on the plane of consistency" (p.158). The BwO has to do with the resistance to organization into strata. And I think the authors are, again, not simply speaking metaphorically when they say that the BwO is the Earth. "It is matter that occupies space to a given degree." The intensities and thresholds and gradients can be applied materially and culturally and economically.

The BwO is opposed to the organism: The organism is

a stratum on the BwO, in other words, a phenomenon of accumulation, coagulation, and sedimentation that, in order to extract useful labor from the BwO, imposes upon it forms functions, bonds, dominant and hierarchized organizations, organized transcendences. The strata are bonds, pincers. ... We are continually stratified. But who is this we that is not me, for the subject no less than the organism belongs to and depends on a stratum? Now we have the answer: the BwO is that glacial reality where the alluvions, sedimentations, coagulations, foldings, and recoilings that compose an organism -- and also a signification and subject -- occur. ... The BwO howls: "They've made me an organism! They've wrongly folded me! They've stolen my body!" The judgment of God uproots it from its immanence and makes it an organism, a signification, a subject. It is the BwO that is stratified. It swings between two poles, the surfaces of stratification into which it is recoiled, on which it submits to the judgment, and the plane of consistency in which it unfurls and opens to experimentation. If the BwO is a limit, if one is forever attaining it, it is because each stratum, encased in it, there is always another stratum. ... A perpetual and violent combat between the plane of consistency, which frees the BwO, cutting across and dismantling all the strata, and the surfaces of stratification that block it or make it recoil. (p.159)

But the BwO is not simply reality, and it becomes clear near the end of the chapter that "intensities" refer to intensities of desire:

The BwO is desire; it is that which one desires and by which one desires. And not only because it is the plane of consistency or the field of immanence of desire. ... Money, army, police, and State desire, fascist desire, even fascism is desire. There is desire whenever there is the constitution of a BwO under one relation or another.

Desire can be an organizing principle, I suppose, but not a totalizing one or a teleological one. Desire is postponed, immanent. The authors here use masochism as a metaphor for infinitely postponed desire in which routes and circulations are blocked. Again, we can turn to Negotiations to make a connection: Deleuze suggests there that in a control society, desire is infinitely postponed.

What an exhausting chapter. Frankly, the authors never quite break the surface here, but near the end they almost do. Again, I think that reading earlier works in the authors' oeurve would make this elusive chapter much more comprehensible.


Faciality was discussed in Chapter 5; it was the redundancy of signs that together form a signifier. Sometimes clumps of signs deterritorialize, break away from one regime, and reterritorialize on another face.

Why "face"? The term refers to the two axes of signifiance and subjectification:

two very different semiotic systems, or even two strata. Signifiance is never without a white wall upon which it inscribes its signs and redundancies. Subjectification is never without a black hole in which it lodges its consciousness, passions, and redundancies. (p.167)

Semiotics are mixed; strata always come in twos, at least; so the intersection of these two strata (white wall/black hole) is a special mechanism, a face. The person's face is what makes her words meaningful (p.167). It is an organ of meaning-making, and in this sense, the face is an absolute deterritorialization: "it is no longer relative because it removes the head from the stratum of the organism, human or animal, and connects it to other strata, such as signifiance and subjectification" (p.172). (This is in contrast with the hand, which is a relative deterritorialization of the paw.)

The authors give us four theorems:

First theorem. Deterritorialization never happens alone. There are always at least two terms, and each term reterritorializes on the other, serving as a new territoriality for the other. Reterritorialization implies a set of artifices that enables this reterritorialization (p.174).

Second theorem. Intensities and speed of deterritorialization do not necessarily correspond.

Third theorem. The least deterritorialized reterritorializes on the most deterritorialized.

Fourth theorem. The abstract machine is effectuated by the face, but also on other things that are "facialized" -- clothes, objects, etc. (pp. 174-175)

The face, the authors tell us, is Christ's. "He invented the facialization of the body and spread it everywhere" (p.176). Everything is facialized now, everything is a potential accumulator of clumps of signs.


Honestly, I got nothing important out of this short chapter. Sorry.


Segmentarity refers to the ways we divide phenomena to get a grip on them. The authors identify three kinds of segmentarity: binary (dualist -- men-women, adults-children), circular ("my affairs, my neighbor's affairs, my city's, the country's, the worlds"), and linear (proceedings) (pp. 208-209). These are all bound up with each other and overlap. "The notion of segmentarity was constructed by ethnologists to account for so-called primitive societies, which have no fixed, central State apparatus and no global power mechanisms or specialized political institutions," the authors say, but segmentarity applies just as well to our own lives. In primitive societies, "primitive segmentarity is characterized by a polyvocal code based on lineages and their varying situations and relations, and an itinerant territoriality based on local, overlapping divisions. Codes and territories, clan lineages tribal territorialities, form a fabric of relatively supple segmentarity" (p.209).

The term supple is important here, opposed to the rigid segmentarity that characterizes the centralization of State societies (p.210). They spend most of their time discussing rigid segmentarity, which characterizes modern societies. And here's where micropolitics comes in. Micropolitics works at the "molecular" level, as opposed to the "molar segments" of macropolitics. For instance, there are totalitarian states -- characterized at the macropolitical level as having "a rigid segmentarity and a particular mode of totalization and centralization" -- that are not fascist. Fascism operates at the micropolitical level, in the smallest political units, as cancerous cells (pp. 214-215) that are suicidal (p.230). Macropolitics are disrupted by contradictions, the authors say, referring explicitly to Marxism; but micropolitics are characterized by lines of flight (p.216).

In this context, the authors begin talking about flow, defined as belief or desire, "the basis of every society" (p.219). Such flows are not contradictions but escapes; yet again the authors use the example of the Crusades, which they call a connection of flows. By connection they mean "the way in which decoded and deterritorialized flows boost one another, accelerate their shared escape, and augment or stoke their quanta"; it is opposed to conjugation, in which those flows encounter "relative stoppage" that generally reterritorializes them and "brings the flows under the dominance of a single flow capable of overcoding them" (p.220). (This sounds a bit like Bakhtin's distinction between the centrifugal and centripetal to me.)

So let's get back to the three distinctions the authors have been sporadically making throughout the book. There are three kinds of lines in this map: primitive (tribe: territorial, lineal), rigid (epire: dualist, overcoded), and flight (war machine: decoding, deterritorialization) (p.222). "Power centers obviously involve rigid segments," the authors say, but the segments are also distinguished and united by a power center or "point of resonance" (p.224). This isn't a contradiction. Think in terms of Latour's work, in which power is defined as a consequence rather than a cause -- the longer and more entangled a network, the stronger it is, but part of extending the network involves coding (segmenting) the environment through representations. And that brings us to Callon, who agrees with Deleuze and Guattari that the most deterritorialized flow determines the dominant segment -- and money is the most deterritorialized of all flows (p.226). Power, the authors say, involves stopping lines of flight (p.229).


We get to affect now, "the effectuation of the power of the pack that throws the self into upheaval and makes it reel." Affect is "not a personal feeling" but a sort of affiliative feeling (p.240). Affect is a key concept in this chapter, which is about becoming through feeling affiliation. The authors' opening paragraphs discuss Willard, the movie in which a young man has creepy affection for his pet rat -- a sort of becoming-rat, the authors say. What goes into that sort of becoming?

Affect is a big part, of course. The authors continue in this vein by discussing becoming as a kind of splicing or grafting into another lineage, epidemic vs. filiation, contagion vs. heredity. They use the same metaphors that Haraway uses elsewhere: vampire, werewolf, cyborg, virus -- all of which reproduce by merging with other organisms rather than through lineage (p.241, 244, 245). If the authors lived in a Protestant country rather than a Catholic one, perhaps they would have included Christianity, which has been called a "mind virus" by detractors and an alternative to biological reproduction by its advocates (I Cor 4:14-15, I Cor 7).

Becomings involve entering into assemblages or complex aggregates (p.242). And this brings us back to multiplicity.

This is our hypothesis: a multiplicity is defined not by the elements that compose it in extension, not by the characteristics that compose it in comprehension, but by the lines and dimensions it encompasses in "intension." If you change dimensions, if you add or subtract one, you change multiplicity. Thus there is a borderline for each multiplicity; it is no way a center but rather the enveloping line or farthest dimension, as a function of which it is possible to count the others, all those lines and dimensions constitute the pack at a given moment (beyond the borderline, the multiplicity changes nature). (p.245)

Packs are multiplicity; "Each multiplicity is already composed of heterogeneous terms in symbiosis, and ... a multiplicity is continually transforming itself into a string of other multiplicities, according to its thresholds and doors" (p.249, their italics).

Later in the chapter, we get to the example of the tick discussed in chapter 3, this time in terms of affect. The tick's "three affects": it is attracted to the light, so it goes to the tip of a branch; it is attracted to the smell of mammals and abandons the branch when it smells one go by; it is attracted to skin and digs in (p.257). This list of attributes, by the way, is very Latourean.

Let's skip a bit. The authors later get into the question of assemblages, which all operate on the same plane in an antihierarchical fashion (p.268). (This reminds me of actor-network theory, again, but especially John Law's discussion of method assemblages in his book on research methods.) Each world is an assemblage effectuating the Cosmos (p.280). Becoming is an antimemory; becoming is rhizomatic (p.294).


The refrain is the authors' account of relative stability. Every milieu is coded, where a code is a periodic repetition; each milieu is vibratory. But each code is perpetually transcoding (p.313). The key of a milieu's existence is that it achieves a periodic repetition, but the effect of that repetition is to produce "a difference by which the milieu passes into another milieu" (p.314). (This sounds like Bakhtin's discussion of genre as both remembering its past and adapting to new conditions.)

Territories -- not the things on national maps, but more like animal territories -- are an act of marking that "territorializes" milieus and rhythms (p.314). "There is a territory precisely when milieu components cease to be directional, becoming dimensional instead, when they cease to be functional in order to become expressive. There is territory when the rhythm has expressiveness. What defines the territory is the emergence of matters of expression (qualities)" (p.315). Still talking about animals' territories, the authors give the examples of colors in birds and fish.

The assemblage, they say, is fundamentally territorial; "the territory is the first assemblage" (p.323). "In a general sense, we call a refrain any aggregate of matters that draws a territory and develops into territorial motifs and landscapes" (p.323). Ages are assemblages (p.346).


Remember nomads and the war machine? They were discussed in earlier chapters, particularly Ch. 5. Nomadology is opposed to the State and centralization. Its invention, the war machine, has been used by the state but at its heart is nomadic. The war machine has three rules:

1. Question hierarchies. The war machine is a band or pack, a rhizomatic structure rather than a hierarchical or lineal one. The leader is prevented from acquiring stable power because when s/he is opposed, it is by someone who might take a pack with them. The war machine is about alliances rather than hierarchies.

2. Blackmail by betrayal. And that means that the leader has to manage these alliances. Again, this all sounds very Latourean/Machiavellian, and we're reminded of Latour's understanding of power as a consequence rather than a cause. The leader is a leader by virtue of holding the alliance together, not by her or his personal strengths and virtues.

3. Volatile sense of honor. These imply a volatile sense of honor in which betrayal is valorized and the warrior has a "fundamental indiscipline" (p.358).

The war machine also has several axioms that I won't get into. I'll hit what I think are the highlights.

Using this frame of nomadology, the authors get to the question of science. The war machine perpetuates a "nomad" or "minor science" (p.361). They cite Michel Serres' work. These nomad sciences are not legalistic, but focus on consistency; becoming and heterogeneity; nonlinear (smooth) rather than linear (striated) relations; and problematic rather than theorematic modeling (p.361). Legalistic science consists of reproducing, making sure the same phenomena recur in the same striated space, "constantly reterritorializing around a point of view." Nomad science consists of following. In this "ambulant" model, "the process of deterritorialization constitutes and extends the territory itself" (p.372). In the legalistic model, "a 'method' is the striated space of the cogitatio universalis and draws a path that must be followed from one point to another." In the nomad model, one occupies "a smooth space that it must occupy without counting" (p.377). (There are obvious applications to Law's work on method assemblages.)

The authors also apply it to types of human organization. There are three types: lineal (primitive), based on clan lineages; territorial (State), based on metric power or the impulse to striate things; and numerical (nomadic), which uses numbers directionally rather than dimensionally or metrically (pp. 388-390). In the numerical (nomadic) organization, "the arithmetic base unit is therefore a unit of assemblage, for example, man-horse-bow, 1x1x1 ... and the formula becomes more complicated to the extent that certain 'weapons' assemble or articulate several men or animals, as is the case of the chariot with two horses and two men" (p.391). Nomadic organization uses numbers to describe assemblages, not to measure space.

There's lots more in this chapter, but let's leave it for now and get to the next one.


The last chapter talked a bit about the State. This one talks about how the State captures things: people, territory, the war machine. They see the State as animated by a rhythm between signs and tools; "the combination, sings-tools, constitutes the differential trait of political sovereignty, or the complementarity of the State" (p.425).

The State's first pole of capture, called "imperial or despotic," "corresponds to Marx's Asiatic formation" (p.427). It overcodes existing codes such as the linear codes of agricultural communities. This overcoding submits them to the despotic emperor. This overcoding is the Signifier -- something that makes sense in conjunction with Chapter 5. This system, they say, is a system of "machinic enslavement" (p.428). Great. What Marx didn't understand, the authors say, is that the State preceded rather than followed from developed forces of production. "It is not the country that progressively creates the town but the town that creates the country. It is not the State that presupposes a mode of production; quite the opposite, it is the State that makes production a 'mode'" (p.429). More startlingly: "Everything is not of the State precisely because there have been states always and everywhere. Not only does writing presuppose the State, but so do speech and language" (p.429). Indeed, language is used to bridge differences and develop State relations: "language was made ... for translation, not for communication" (p.430).

The State may be everywhere, but "the appearance of a central power is thus a function of a threshold or degree beyond which what is anticipated takes on consistency or fails to" (p.432). Thresholds of consistency coexist. One threshold is the Town's: "The town exists only as a function of circulation, and of circuits," that is, it exists because of what the roads take into and out of it. "It is a phenomenon of transconsistency, a network, because it is fundamentally in contact with other towns. It represents a threshold of deterritorialization, because whatever the material invoked, it must be deterritorialized enough to enter the network, to submit to the polarization, to follow the circuit of urban and road recoding" (p.432). The State's threshold is different: "it is a phenomenon of intraconsistency. It makes points resonate together ... it makes the town resonate with the countryside. It operates by stratification: in other words, it forms a vertical, hierarchized aggregate that spans the horizontal lines in a dimension of depth. In retaining given elements, it necessarily cuts off their relations with other elements, which become exterior, it inhibits, slows down, or controls these relations" (p.433). It takes the territory as an object to stratify and resonate. The Town breaks free when the State's overcoding produced decoded flows; capitalism, they say, is the fruit of the towns, arising "when an urban recoding tends to replace State overcoding" (p.434). But capitalism triumphed through the State (p.434).

Now we get to the apparatus of capture, which (again, following Marx) comes in three forms: rent, profit, and taxation. Of the three, taxation is what monetarizes the economy (p.443). The really interesting part is that capitalism deterritorializes by taking "materialized labor" (commodity) as its object; the State reterritorializes (p.454). In capitalism, the States change form and become models of realization for it (p.454). Indeed, capitalism is so successful partially because "it continually sets and then repels its own limits, but in doing so gives rise to innumerable flows in all directions that escape its axiomatic." It continually disrupts its own models (p.472).


We used the terms smooth and striated earlier. The difference is that of fabric and felt. Fabric is striated, constituted by at minimum two elements (usually vertical and horizontal), creating a delimited and metered space. Felt is an "anti-fabric": fibers entangled at the micro level. It is smooth (p.475), but smooth like patchwork: heterogeneous, amorphous, nonformal space (p.477). Smooth is nomos, striated is logos (p.478).

Striated space is the line between two points; think of latitude and longitude (p.480). Smooth space is the point between two lines (p.480); it emphasizes trajectories, vectors, haptic rather than optical sensation (p.479). Striated space is the space of the State, whereas smooth space is the space of the nomads -- as we saw in the chapter on regimes of signs.

One place where this discussion of two spaces is being hashed out, the authors say, is mathematics. They credit Riemann for developing the notion of multiplicity, which they say marked the end of dialectics (p. 483). There are two types of multiplicity: "one qualitative and fusional, continuous, the other numerical and homogeneous, discrete." Or: nonmetric and metric, acentered and centered; rhizomatic and arborescent; smooth and striated (p.484).

Striating involves making, not just measuring, differences (p.489).

There's a discussion of capitalism toward the end of the chapter, but let's save that for another time. I'm tired.


See, I told you they were liars. For all the discussion of the book as rhizomatic, the conclusion approaches a conventional conclusion, with summaries of the major ideas of the book.

I'm sure you can see why my copy of this book has so many post-it notes sticking out of it. The book is rich with ideas and frameworks, and reading it the second time has actually been a rewarding experience, though usually not an enjoyable one. As I said, I've left a lot out and I'm sure I've gotten some things wrong. So please do leave comments if you'd like.

Blogged with Flock


roy said...

Many thanks for the review and walk-through. I am impressed by the fleetness of mind, and the many impressive intellectual arabesques and pas de deux.

But – I have gone long since back to the simplest: i.e. Barthes “every use becomes a sign of itself” and (various semioticians, from de Saussure onwards) “semiotics is a system of difference” – which simply means that all uses are defined in (an adaptive, ecological) relation to all other uses, and all meaning of signs, therefore, are defined in relation to all other signs.

Consequently, in agreement with Derrida in particular, there is no ‘ultimate signifier’ – the chain of meaning is endless, and does not stop, or go to, any one point, or sign, or ‘being’. (Pretty much a Buddhist line of thought).

Having read through all this, summary, it makes sense, but so does Olympic ice-skating. Very impressive, but I don’t need to go there.

Mr. Blimpy (aka Aaron Thompson) said...

This is really great. Many thanks for this! Especially for the Latourian reading of this difficult yet intriguing book.

Clay Spinuzzi said...

Glad it was helpful!