Notes on: Deleuze, G and Guattari, F. ( 2004) A Thousand Plateaus.London: Continuum. Chapter 15 Conclusion: Concrete Rule and Abstract Machines

Dave Harris


The authors this is a jolly clever book with lots of new terms and concepts to admire even if you haven't a clue what they mean. That's your fault anyway for being such an ill-educated plonker. It's probably quite funny and provocative too. God is a lobster! Hilarious!  Maybe just quoting them will help?


abstract machine

At last. This summarizes the themes one last time, maybe focusing on the structurating bits that form things out of the swirling possibilities (long overdue – but still only a philosophical forming not anything resembling an actual social forming). I have put the original summaries first, followed by calmer and less despairing details

Strata. There is a lot of stuff on these in the largely unreadable Ch. 3 . The chapter can be grasped initially as explaining in much more detail shifts from quantity to quality ( the two articulations)  If I have the energy I will add notes on this chapter -- one thing it explains is the insistence that inanimate matter 'expresses' itself:  Strata are formations providing form, content and substance (the lengthy discussion of various conceptions of these terms left me cold). I was amused to see much made of the term ‘double articulation’, which became very fashionable in British Cultural Studies, no doubt because the leading lights had somehow assimilated it from other discussions, as trendies do when they go to conferences. It is applied much more prosaically there, classically to mean the double influence of class and age in the formation of deviant youth subcultures. Here it means a double process of formation. In the only examples I could begin to grasp, D and G talk about the formation of strata of rock -- first molecules are laid down by water in a kind of loose pattern, then those molecules are squeezed ( ‘folded’) into sedimentary rocks. Or, in a biological example organic molecules are formed from elements then they are joined together eg in polymers: polymers then can form more complex and interesting emergent molecules in a qualitative shift from quantity. In the third main area, the social is a stratum – or  ‘the anthropomorphic’. Double articulation is described as a double pincer, like lobsters have -- hence ‘God is a lobster’ (45).

More detail:

The three major strata are the 'physico chemical, organic and anthropomorphic'(553).  Each contain forms and substances, or codes and milieu, in some diversity and variety.  There are both formal types of organization, and different modes of development and these produce a division into parastrata and epistrata, which can be considered as strata themselves.  In each case, there is a unity of composition in spite of the diversity of organization and development, relating to common formal traits or codes, and substantial elements or materials.  Strata are 'extremely mobile'and they can collide or act as substrata to others, as well as developing and organising.  There are also 'interstatic phenomena' transcodings, passages between milieu, running according to certain rhythms.  Stratification allows the world to emerge from chaos, 'a continual renewed creation', but they are then subject to the Judgement of God [as are all organisms, as we know from chapter six].  Classical artists can also create from chaos in this way.

Articulation constitutes a stratum, but it is always a double, articulating both content and expression, as in Hjemslev and his 'net' [the table offering particular combinations of content and expression, as I recall, so that content and expression each have  form and substance].  Content and expression do not correspond, nor does one cause the other, nor is it like the relation between signifier and signified.  Instead 'there is real distinction, reciprocal presupposition, and only isomorphy' (554).  However, the differences between content and expression vary for each stratum - for example, expression is 'linearized' on the organic and 'superlinearized' on the anthropomorphic.  This affects the combination of molar and molecular on each stratum.

In chapters three and four, we see that it is possible to move outside strata - for example matter is not just physiochemical: 'there exists a submolecular, unformed Matter'.  Similarly, not all life is organic: the organism limits life, and there is a powerful and lively anorganic form.  There are nonhuman becomings of human beings which break out of the anthropomorphic stratum, although it is a problem to reach this plane, or to draw a line leading to it, because outside of strata there are no stable forms and substances organisations or developments, or even contents and expressions.  We become disarticulated.  Non organic life is chaos, requiring extreme caution when we destratify, if we are not to end in chaos or a return to strata in an even more rigid sense ]as in the warnings about taking drugs in ch 10]

Assemblages. These construct territories from elements in milieux. They both assemble  like machines and enunciate (chs 4 and 5)  They are subject to deterritorialization, sometimes producing abstract machines.

More detail:

Assemblages are different from strata although they are produced there.  There are zones where milieu are decoded, and assemblages are able to extract a territory from them.  Assemblages always have to discover 'what territoriality they envelop'(555).  Territory is made of decoded fragments of all kinds, which exist first in the milieu, but are turned into 'properties' in assemblages, and even rhythms can take on the meaning of the refrain [chapter 11].  Territories transcend both the organism and the milieu, and produce assemblages that are responsible not only  for different behaviours, but more than that [the example seems to be the difference between territorial and milieu animals].  Assemblages still belong to the strata or at least 'pertain' to them, and this produces the differences between content and expression [apparently discussed in chapter four].  However, there is some independence from the strata, because we find expression becoming a definite regime of signs, while contents becomes 'the pragmatic system, actions and passions'.  There is of course a double articulation between them [between 'face - hand, gesture - word'].  In this sense, every assemblage is simultaneously machinic and enunciative: something is both said and done.  However, we also have expressions of 'incorporeal transformations'.  We do not find these in the strata, because expressions are not yet formed into sign systems, nor contents into pragmatic systems [baffling -no doubt something argued more fully in chapters 4 and 5]: these formalizations grant the incorporeal transformation a certain autonomy.  Regimes of signs proper are found only in the anthropomorphic strata, 'including territorialized animals', but they permeate all the other strata.  Regimes of signs and pragmatic systems of action can be seen as strata in their own right, although the relation between content and expression represents a new development.

Assemblages have another division along anothe axis as well, between territoriality and lines of deterritorialization, which are very diverse: some link with other assemblages, as when territorial refrains become courtship or group refrains [chapter 11], others open up the territory, to something more 'eccentric, immemorial, or yet to come' (556) [the example is the lied -- pass].  Some lines lead to abstract and cosmic machines.  These possibilities arise because assemblages have already decoded the milieu of the strata, to territorialize initially, but this also opens the possibility of deterritorialization.  Assemblages therefore include 'unformed matters, destratified forces, and functions'.  In order to understand the territoriality of the assemblage we must ask what is the regime of signs and pragmatic systems, but also to look at the 'cutting edges of deterritorialization' which are open the assemblage to abstract machines.  So we have four valencies: content and expression, territoriality and deterritorialization.  This explains a figure in Kafka, apparently.

Rhizomes. Unlike assemblages, the lines they develop go between points in smooth space (see—it does all link up) as in multiplicities. They destabilize trees. Their lines of escape can become lines of death as in fascism when creativity is compromised. More in ch. 1

More detail:

In a first kind of line, the line is subordinated to the point and 'the diagonal is subordinated to the horizontal and vertical'.  Here, lines are contours, and space striated.  There are metric multiplicities subordinated to the One [all discussed in chapter 14, although our heroes also specify chapters nine and one].  These are molar lines forming a segmentory e.g. arborescent system [or a circular and binary one].  Rhizomatic lines are different: the diagonal frees itself; Lyons Pass between things and points in a smooth space.  A consistent multiplicity is developed, as in masses or packs, multiplicity he's of becoming rather than of elements and ordered relations, fuzzy not exact aggregates and so on.  These multiplicity he's appear in psychosis especially schizophrenia, and in the practice of sorcerers.  Theoretically, their status correlates to that of spaces: smooth spaces of the desert type are not depopulated, but have a population of multiplicities.  Mathematics and music have helped develop the theory of multiplicity.

[As usual] the two types of multiplicity are related to each other immanently [referring to the 'political' chapter nine -- so this only happens in politics?].  It is better to speak of 'an arborification of multiplicities' (557), and this can developed when black holes resonate together or where stems get segmented and striate their surroundings [with special reference to of the Face, so chapter seven, as well as chapter 12].  So mass movements can be joined together at points of accumulation or stoppage and get segmented.  However there is always a potential where 'the stems of the rhizomes are always taking leave of the trees', establishing new connections, smoothing space,  providing 'deep movements' to disturb territories or language.

All three of the lines that compose us offer their own dangers, even the molecular lines with their micro black holes, and lines of flight which can turn into a line of death or destruction.

Plane of consistency: BWO

Only a detailed account here:

This plane is opposed to the planes of organization and development, which concern form and substance and their development.  The plane of consistency 'knows nothing of substance and form' (558) and it deals with haecceities, 'modes of individuation proceeding neither by form nor by the subject'.  The plane is both abstract and real.  It demonstrates relations of speed and slowness between unformed elements [longitude].  It composes intensive affects as well [latitude].  It also concretely ties heterogeneous and disparate elements, to consolidate fuzzy aggregates, or 'multiplicities' of the rhizome type'.  It necessarily acts in the middle, not by principle and not by some final origin or goal.  It offers consistent consolidations, 'never unifications, never totalizations', and its surveyors are 'Spinoza, Holderlin, Kleist, Nietzsche'.

Haecceities are 'inscribed' on this plane, and so are events,' incorporeal transformations...apprehended in themselves' ; nomadic essences, 'vague yet rigorous'; continuums of intensities, 'which go beyond constants and variables';  becomings,'which have neither culmination nor subject,but draw one another into zones of proximity'; smooth spaces' composed from within striated space'.  A BWO [also described as a plateau here] 'comes into play in individuation by haecceity' [their references point us to chapters six and 10], producing intensities, and operating through the medium of becoming or transformation and the smoothing of space.  There is also a powerful non organic life, 'a line of nomad art and itinerant metallurgy' (559).

Does the plane of consistency constitute the BWO or the other way about?  Are they the same thing?  They certainly both have the same power, just as lines do not have superior dimensions to points, nor surfaces to lines, nor volumes to surface, because all are fractal [desperately confused].  The plane 'sections multiplicities of variable dimensions.  The issue is really the connection between different parts of the plane.  Do BWOs interconnect?  How can continuums of intensity be extended [that is turned into something extensive?  That has long been my question - here are our heroes avoiding the issue by asking rhetorical questions of their own!].  Why does the middle actually effectuate  linkages?  [They can only offer us a sad metaphor: 'the plane is like a row of doors'].  There are concrete rules[political ones? pragmatic ones?]  for the construction of the plane, and these do exercise a selective role.  In drawing them up, we should reject empty and cancerous bodies, for example, the false version of smooth space, and lines of death and destruction.  What is both retained and preserved, and 'therefore created' [I love that 'therefore'] in consistency is 'only that which increases the number of connections at each level of division or composition'[which sounds like a rerun of the Spinozan ethic of joyful connections, unless it is Guattari on the liberating potentials of the transversal?].  This gives us a descending and an ascending order, because every time we divide or compose, we change the nature of the entity, or make it into a larger composition with a new criterion of comparison [Jesus!].

Deterritorialization. This involves flights of escape from territories but which can sometimes lead to reterritorialization. So there are 4 concrete types (allowing for combinations) if you want to get obsessive about this.

More detail:

Deterritorialization [which thankfully they abbreviate to D, because my voice recognition does not know how to spell the full term] involves leaving a territory by a line of flight.  Sometimes this can be obstructed and reterritorialized, and here D can be called negative.  Reterritorialization [which I am going to abbreviate to R, for the same reasons] can also occur, and it can be based on anything, an object or a book or a system.  The state apparatus actually performs D, but immediately overlays it with R, based on property, work and money.  Signifying regimes can attain a high level of D, but simultaneously a system of R based on signifieds and signifiers.  D can prevail over R but still remain only relative: the line of flight here gets segmented, and it can sink into black holes.  This happens with 'the regime of subjective signs' (560), where passion and consciousness brings about D, but only in this relative sense.  The negative and the positive forms are not just linked in an evolutionary way, but can interact -- sometimes, for example converging lines of flight can produce segments and an overall R based on one of them.  Mixed figures are common.

The territory itself produces vectors of D, because, perhaps, it is supple and marginal, itinerant, or because it interacts with other types of assemblage.  Even R can somehow imply a D: multiple or various forms, speeds and movements can produce a D element.  R does not imply a simple return to a territory, but rather to the same mixed relations internal to D, 'this multiplicity internal to the line of flight' (561).  This helps us explain 'the mystery of the "Natal"': the earth itself seems to be some 'focal point' outside territory, and we reach it only by D.  However, the earth and 'the glacial' [where did that come from?] is itself D par excellence,  belonging to the cosmos, representing cosmic forces which can be tapped by human beings.  In this sense, 'D can be called the creator of the earth - of a new land, a universe, not just an R'. 

This is what we mean by absolute D—nothing transcendent, not a quantity above all other relative quantities, but a' type of movement qualitatively different from relative movement'.  And absolute movement is not defined by a quantity or speed, but how it relates bodies ['"a" body'] to a smooth space, occupies it, as a vortex.  Relative movements however relate bodies  'considered as One' to striated space [they're making up as they go along, trying to make it all hang together].  When D operates in this relative or negative way, it permits Rs that obstruct the line of flight, or segment it.  Absolute D connects lines of flight, producing 'an abstract vital line' or 'a plane of consistency'.  However, absolute 'necessarily precedes by way of relative D', because it is not transcendent, and, conversely, relative or negative D requires a [nasty, overcoded, they call it 'limitative'] absolute to operate, something all encompassing or totalizing.  Conjunction [conjugation?This implies making all the lines belong to one set?]  unites and amplifies lines of flight, unlike mere connection [they admit they have not been consistent in this usage].  Conjoining lines of flight in this nasty limitative way is what leads to lines of destruction.

It is all a matter of whether we are thinking of negative or positive absolutes, whether we see the earth's as 'girded, encompassed, overcoded, conjugated as the object of a mortuary and suicidal organization' (562), or as something connected to the cosmos, 'following lines of creation'.  This gives us our four forms of D [presumably relative and absolute, by positive and negative absolutes].

Abstract machines (Diagram and Phylum) . Not abstract in the usual sense but connected to concrete assemblages which they open up . They ’constitute becomings’  in this way (562). They process unformed matter according to a number of functions or programs. They are always specific in their effects. They operate as plateaus on the plane of consistency. It is possible that there is some general abstract machine with all the qualities of the others.


There is no abstract machine in a platonic sense, something transcendental universal or eternal, because they are only operate within concrete assemblages.  They are defined by following decoding and D.  This makes the deterritorialized assemblage become something else, another type, the molecular or the cosmic, through becomings.  Abstract machines are singular and immanent, not dominated by forms and substances, and thus different from both strata and even assemblages.  They 'surpass any kind of mechanics. [and]  They are opposed to the abstract in the ordinary sense'.  They consist of 'unformed matters and non formal functions', aggregates of these as in phylum and diagram.  We can see this on the technological plane, which has definite substances like plastic or electrical wire, but also organizing forms: but something else is needed, 'a composite of unformed matters' which usually appear only by offering particular intensities, such as resistance or conductivity, or sometimes '"tensors"'[they say these are discussed in chapter five, which relates to linguistic forms, so we have the same problem - do we only find tensors in linguistic forms? Are we generalizing from the linguistic case?].

The abstract machines are 'effectuated in forms and substances' in actual assemblages, in different states [something like the phase states in DeLanda?].  But the abstract machine composes itself first, and its plane of consistency.  This means they are abstract singular and creative, but not concrete, actual but not fully effectuated.  They have been described by various heroes including Einstein and Galileo, Bach and Beethoven -- these names are used to describe singular machines.

Abstract machines do not work with form and substance, and also do not work with content and expression [since these are connected to forms and substances?].  The plane of consistency offers continuous variation, with each abstract machine as a plateau, putting variables of content and expression together.  Content and expression thus take form at the most abstract level [which they render as their 'highest level of relativity' (563), meaning D?] This is discussed in chapters four and five.  However, there are also traits of content, '(unformed matters or intensities)',and expression '(non formal functions or tensors)', and now these traits act as cutting edges of D, its active and passive elements respectively.  [Just in case this is too clear{!}], it can be the other way around—the key argument is that there are dissymetrical roles for content and expression even though they are 'elements of a single becoming'.  For that reason, we can't define continuous variation as something that includes both indiscernibly: instead, we have to specify traits and intensities of content and expression ['indefinite article, proper name, infinitive and date', we will recall from chapter 10], and each of these can lead when crossing the plane of consistency.

It would be wrong to consider unformed matter, the phylum, as just inert, since it is 'a matter- movement bearing singularities or haecceities, qualities, and even operations' (564).  Nor is the diagram simply composed with some 'inexpressive metalanguage': instead it features 'an expressivity - movement', something like 'the foreign tongue within each language', and also 'non linguistic categories within language'.  [For some bizarre reason this is described as 'nomad poetic lineages'].  Unformed matter is real [an equally odd phrase says that 'One writes... on the same level as the real of an unformed matter'], and matter can both traverse and extend 'non formal language', as in the examples of becoming - animal described in Kafka and Hoffmanstahl [see Ch 10].  The abstract machine can be revolutionary, 'all the more abstract for being real', operating outside signifiers or the subjective. 

Nevertheless, there might still be an overall transcendent abstract machine, where concrete assemblages 'are related to an abstract idea of the Machine' and can be seen as offering 'coefficients' of it, where these coefficients' relate to assemblage components like territory, D, R, earth, cosmos.  This can help us trace the 'entangled lines constituting the "matter" of an assemblage', and thus the different relations between the assemblage and the plane of consistency.  [An annoying animal example ensues, relating to the ways in which grass stems are used in various ways by birds, in the weird chapter 11 -- very unconvincing.  Naturally, our heroes wish to write this as a 'having different coefficients in assemblages of animal species that are nevertheless closely related']

However, there may be different types of abstract machines.  We can see this by examining the operations that constitute the strata and the assemblages, which do not just derive from the plane of consistency.  Thus strata organize themselves, as thickening or coagulation, and start to develop along another plane [a more concrete one?], which does feature substances, forms, contents and expressions.  This gives us consistency related to substantial elements and forms.  What is responsible is 'a properly stratic abstract machine' on the second plane [the second type].  There is a third type on the anthropomorphic strata [actually the 'alloplastic strata'], which is particularly useful for the formation of assemblages.  These manage a balance between D and R, between decodings and overcodings, a way of closing assemblages.  Thus there is 'an order - word machine' which over codes language, a 'faciality machine' that overcodes the body, and a 'machine of enslavement' that overcodes or axiomatizes the earth.  These are also real machinic effects. [They have to balance the creativepotentials with some account of oppression in real life? Oppressive, closing machines are equally real and immanent? ]

It is no longer a matter of placing assemblages on some scale to see how close they are to the plane of consistency, because there are different types of abstract machines which overlap and bring new qualities to the assemblages, machines of stratification as well as of consistency, and over coding abstract machines to close off assemblages.  Every abstract machine is linked to these other abstract machines: they are 'inseparably political, economic, scientific, artistic, ecological, cosmic—perceptive, affective, active, thinking, physical and semiotic'.  [So --open to poloitical struggle or beyond it?] The different types are intertwined and their operations converge in an entire 'mechanosphere'.

Now in bullshit:

  But if abstract machines know nothing of form and substance, what happens to the other determination of strata, or even of assemblages—content and expression? 1n a certain sense, it could be said that this distinction is also irrelevant to the abstract machine, [3] precisely because it no longer has the forms and substances the distinction requires. The plane of consistency is a plane of continuous variation; each abstract machine can be considered a "plateau" of  variation that places variables of content and expression in continuity. Content and expression thus attain their highest level of relativity, becoming "functives of one and the same function" or materials of a single matter [see 4, "November 20, 1923: Postulates of Linguistics” note 2 1—Trans.]. But in another sense, it could be said [4 ]that the distinction subsists, and is even recreated, on the level of and traits: there are traits of content (unformed matters or intensities) [and 5] traits of expression (nonformal functions or tensors). Here, the distinction has become entirely displaced, or even a different distinction, since it now concerns cutting edges of deterritorialization. Absolute deterritorialization implies a “deterritorializing element" and a "deterritorialized element” one of which in each case is allocated to expression, the other to content, or vice versa, but always in such a way as to convey a relative distinction between the two. Thus both content and expression are necessarily affected by continuous variation, but it still assigns them two dissymmetrical roles as elements of a single becoming, or as quanta of a single flow. That is why it is impossible to define a continuous variation that would not take in both the content and the expression, rendering them indiscernible, while simultaneously proceeding by one or the other, determining the two mobile and relative poles of that which has become indiscernible. For this reason, one must define both traits or intensities of content [1,2] and traits or tensors of expression (indefinite article, proper name, 4,10 infinitive, and date), which take turns leading one another across the plane of consistency. Unformed matter, the phylum, is not dead, brute homogeneous matter but a matter-movement bearing singularities or haecceities... ( 562—3) NB the numbers in the text which I have put in square brackets are set off in the left margin in the original – they refer to the chapters, apparently)

Clear as fucking mud.

back to menu page