Notes on: Guattari, F.  (2011) The Machinic Unconscious.  Essays in Schizoanalysis, translated by Taylor Adkins.  Los Angeles: Semiotext(e) Foreign Agents.

Dave Harris

Thanks to Chandler again

[Formidably difficult material, as usual.  As if the frequent references to offers and discussions that I have not read are not difficult enough, the style is also close to impenetrable, with eternal sentences, and this quality of 'Schizo flow' that others have mentioned.  I should be lucky to pick even the most modest bones out of this mushy soup.  One thing of interest is that quite a lot of this appears to have been reproduced in Thousand Plateaus,(TP)  sometimes almost word for word, like the bit that surrounds the famous statement 'there is no language', page 7 in TP.  I suppose the kindest thing that could be said is that this account is slightly more accessible than the pretentious rubbish in TP]

Introduction: Logos or abstract machine? [NB see also the discission on machine vs structure in Psychoanlaysis and Transversality]
Is the unconscious still a useful concept?  Can it still be understood and translated?  Modern conceptions see it is a structural matter, with very little left of Freud or Jung, structured like a language, even a mathematical language as in Lacan.  For Guattari, the unconscious affects all kinds of perceptions and actions, affecting the possible itself and all forms of communication, not just linguistic ones.  He uses the term machinic unconscious to stress that it is full of 'machinisms that lead it to produce and reproduce these images and words' (10).

There is a need to reject both classic notions of causality, and structural abstractions.  Instead, we need to consider interactions with objects, space and time, without worrying whether they are material, semiotic, or transcendental.  We should be working with 'abstract deterritorialized interactions' produced by abstract machines, especially as they traverse different aspects of reality and 'demolish stratifications' (11).  These interactions operate on a plane of consistency which crosses place and time.  They should be grasped as the 'quanta of possibles'.

Assemblages fix and unfix 'coordinates of existence', constantly deterritorializing and singularizing, and establishing new replacement territories—'machinic territorialities'.  These processes are universal, and are only slowed down or thickened on the '"normal" human scale'.  The same goes for normal understandings of causality and temporal sequence.  [Thom is cited on the apparent smoothing or reversability of space and time.  Machines are not just understood in terms of their current manifestations and they produce a plane of consistency enabling all sorts of other intersections. Thom is on to this, but too likely to understand it in terms of mathematics, thus remaining at the abstract level, unable to talk about the process of singularization, which Guattari here refers to as 'extracts' from the cosmos and history].  Referring to abstract machines reminds us that we are not just talking about normal processes of abstraction to get to universals, and the need to think of mechanisms as well as assemblages.  There are no universal general assemblages—universality is a function of power.

The search for some systematic formal order, over expressions, for example, as studied in the social sciences is impossible, because it really depends on 'political and micro political power struggles' to attempt to stabilize an essentially drifting language.  Attempts to reduce behaviour and language to binaries or digits could be extended to all social phenomena, but we will not have grasped the essence of the phenomena, unless we make some assumptions that everything aims towards stabilized equilibrium.  Structuralism attempts to deal with contingency and singularity by probabalizing them along synchronic and diachronic axes, but this is only a tidying up, and it misses altogether those assemblages that produce 'rupture and innovation'(13). 

The tendency to aim at general axioms in science, and at pure concepts in philosophy, has meant the dominance of epistemology.  The efforts to connect certain privileged denunciations with a transcendent order should be understood instead as a form of power, justifying 'the social status and the imaginary security of its pundits and scribes in the fields of ideology and science' (14).

It's possible to develop a formalism operating with 'transcendent universal forms cut off from from history', or to see 'social formations and material assemblages' as embodying them so that they can be inferred.  Some encodings will come to appear to be natural, and accidental connections with 'sign machines' will appear as general laws.  These were sometimes misunderstood as comprising a metalanguage—but language, enunciation, states of affairs and subjective states are really all on the same level ['flat ontology'] .  There are no independent subjects or objects— their seeming independence is really an effect of deterritorialization.  Abstract machines connect with deterritorialization and this is what produces apparent universal causess, laws, and pre-established orders.

It is common to argue that we can never possible to develop abstract conceptions free from 'invasions' from social assemblages or mass media.  However, it is 'assemblages of flows and codes' that differentiate form, structure, objects and subjects in general in the first place  so these invasions are just specific observable cases (15).  Abstract machines do not code existing social stratifications from the outside.  What they do is to offer transformations inside a general deterritorialization, by constructing 'an "optional subject"', a kind of focusing of possibilities, 'crystals of the possible' (16).  They assemble components in order to affect realizations, never logically or in a law-like way, but contingently, never simply passing, for example from the complex to the simple, and never establishing a systemic hierarchy.  For example, the most elementary elements can bring forth new potentialities and invade more complex assemblages.  We need to lose terms like 'the elementary' when describing what goes on and refer instead to a 'molecular level', and this can never be simplified or reduced.  Molecules can provide keys or seeds for more complex and differentiated developments.  The molecular is more deterritorialized, and this is essential for more complex assemblages to arise.

So abstract machines can not be understood in terms of subject and object, nor in standard logical terms [like the subject of a sentence and its predicate?], And nor can they be understood in terms of denotation/representation/signification.  Instead, they offer something more general, 'the order of subjectivity and representation, but not in the traditional form of individual subjects and statements detached from their context' (17).  One consequence is to move away from anthropocentric conceptions, to move from, for example, logical propositions to 'machinic propositions' and to examine 'non - semiologically formed matter' [eg birdsong].  This will make [conventional, human] coding or signifying look more contingent.  There will also be an emphasis on singularity as something that does not just contain 'a limited number of universal capacities'.  Assemblages can be undone to reveal other possibilities, overcoming the alliance between universal thought and 'respect to an established order'.

Linguistics and semiology claim a privileged place by claiming to be able to solve problems in other disciplines.  But this is really a matter of high status and an appearance of scientificity.  Saussure, for example, has simply borrowed by lots of psychoanalysts, following some sort of tacit agreement over boundaries of domains.  It is this 'shared problematic' that will be investigated in particular (18).  Instead, it is important to look at issues which will help us revive the notion of the unconscious, and understand pragmatics in a new way.  We also want to look at two particular issues: 'faciality traits and refrains'.  Then we want to develop 'a schizoanalytical pragmatics' to address political and micro political problems, and then to go on to develop new semiotic entities based on pragmatics in the form of 'a "machinic genealogy"'.  Then we will talk about faciality traits and refrains in Proust [see my summary --sic-- here . See also Deleuze's commentary on Proust]

A glossary might help [ha!]

In order to replace formal analysis with analytical pragmatics and schizoanalysis, we need to replace a tree with a rhizome or lattice.  Instead of dichotomous choices, as in Chomsky, rhizomes can connect any point to any other point.  Each semiotic chain can refer to a number of 'encoding modes: biological, political, economic chains, etc.' (19), meaning we can discuss sign regimes and also forms of non signs.  We will look at the relations between segments at different levels inside each semiotic stratum, in order to demonstrate 'lines of flight of deterritorialization'.  We will emphasize pragmatics rather than underlying structure, machinic unconscious rather than the psychoanalytic unconscious, something that resembles a map rather than 'a representational unconscious crystallized in codified complexes and repartitioned on a genetic axis': this map will be 'detachable, connectable, reversible, and modifiable'. Tree structures can appear within rhizomes, and conversely the branches of the tree can send out buds in a rhizomatic form [but we need a 'pure' model?]

We can consider pragmatics as having different components.
  1. Interpretative components imply more importance for semiology and signification.  There are two general types of transformations involved— analogical, as in iconic signs, and signifying, requiring more linguistic semiology.  A type becomes dominant only within particular modes of subjectification of power—(re) territorialized assemblages of enunciation for analogical transformations, and individual assemblages for signifying transformations. 
  2. Non interpretative components are more general, with interpretative ones as 'a particular or borderline case'.  They also have two types of transformation: symbolic transformation, operating at the level of perception or gesture and also involving other verbal and scriptural 'levels that escape from analogical redundancies'(20); diagrammatic transformation, which involves a non semiotic deterritorialization produced by 'mutant abstract machines…  working simultaneously within the register of material and semiotic realities' [I assume this meant transforming from one specific shape to another governed by the overall diagram.Mutant possibilities arise with deviant groups --madmen or radicals].

At the semiotic level, there are also two types of redundancy.  One involves 'redundancies of resonance'[and this is something to do with faciality and refrains] (21).  A second type is machinic redundancies' 'or redundancies of interaction' [something to do with diagrammatic components, presumably the ways in which one gives place to another within the overall diagram?  As in evolution through the machinic phylum? OR, the condensed versions of language in 'restricted codes' or ideologies?].

At the existential level, there are three levels of consistency.  The first one is molar consistency, relating to 'strata, significations, and realities', and this seems to have to do with phenomena as [collectively?]  perceived, including whether or not they are seen as complete objects, subjects or individuals.  The second one is molecular consistency, relating to assemblages and how machines are embodied in them.  The third one is abstract consistency that apparently 'specifies the "theoretical" degree of possibility of the two preceding consistencies'.  by combining semiotic and existential types, we can get 'six types of fields of resonance and fields of interaction'[up to now I had blamed Deleuze for this obsessive classification and tabulation].

The first note refers to differences in terminology, especially with Chomsky, and the second one distinguishes semiology from semiotics.  Semiology is 'the translinguistic discipline that examines sign systems in connection with the laws of language', with Barthes as the example, while semiotics proposes to study sign systems 'according to a [pragmatic?] method which does not depend on linguistics', and the example is Peirce.

Chapter two.  Escaping from Language

[Lots of stuff here about different linguistic theories.  The main points seems to be to argue that pragmatics is far more important than any other aspect of language, and that general theories, including structuralism, are really describing particular clusters of pragmatic utterances, some of which come to seem important for political reasons]

Functionalist accounts of language operated with phonological chains organized around binaries.  Language was seen in terms of information theory, with messages and redundancy and so on.  Social and political context were ignored, so that linguistics can appear to be scientific and 'serious'.  Generative linguistics, as in Chomsky, saw functionalist models as describing only the surface activity produced by underlying syntactic structures.  Some followers wanted to describe the syntax in terms of mathematics or types of logic.

Even a more recent emphasis on enunciation [Foucault? Someone else I think] has failed to grasp the social and political context, and it is still common to study enunciation in general, or something abstract, 'an alienated enunciation' (24).  In all these approaches, pragmatics were seen as something to be dumped in a waste basket, either ignored, or grasped in 'a restrictive mode'.  Syntax and phonology dominated.  Enunciations were supposed to be located at particular 'structural junctions', but were never seen as contingent or singular.  Language was just assumed to be able to represent a social system.  It was also assumed that semantic and pragmatic fields could be digitalized—relying on contents and context as something that can be formalized and seen as the result of a 'system of universals'(25).  Sometimes this means that semantic creativity depends entirely on syntactic structures [Guattari wants to suggest instead Hjelmslev's notions that, apparently creativity finds an origin in 'the concatenation of figures of expression and figures of contents' -see below].  Generally, creativity is seen as something marginal or deviant, but this does not explain how originally marginal words become mainstream.

Linguistics claims to be able to explain all domains that use language, and this makes it 'imperialist'.  There are claims to be able to explain pragmatics in a neutral way, as something secondary.  Guattari wants to explain pragmatic contents in terms of 'collective assemblages of enunciation'(26).  This means that 'there is no language in itself', that what we have is specific language of all kinds being produced via 'an abstract machinic phylum'.  There is no direct connection with social structures, and no general structure that produces statements separately.

We can start with the classic distinction between language and speech [and deny it in favour of a pragmatics of enunciation].  The pragmatics of the unconscious or schizoanalysis will be particularly hostile to Saussurian structuralism, since it shows that language does not have 'a domain of its own'(27), that it is always an open system connected to 'all the other modes of semiotization'.  It is only political or micro political processes that close it up and call it a national language or dialect, or delirium.  Language is suffused with 'borrowings, amalgamations agglutinations, misunderstandings'.  The same goes for anthropological structures, such as the prohibition on incest, which, when examined in detail, consist of 'rules that can be bent in all kinds of ways'. It is power formations and power centres that unify languages and establish boundaries [leading to the bit that reappears in Thousand Plateaus. The whole discussion on language is very similar in fact -- maybe a bit clearer].

Guattari the recycler:

Machinic Unconscious: 28

Language is stabilized around a parish, fixed around a bishopric, and installed around a political capital.  It involves by flowing along the river valleys, along the railway lines; it moves through oil spots (example of the Castilian dialect). [it is just assumed we all know about the Castilian dialect!]
Thousand Plateaus: 7

Language stabilizes around a parish, a bishopric, a capital.  It forms a bulb.  It evolves by subterranean stems and flows, along river valleys or train tracks; it spreads like a patch of oil

Note 4, p.333 says:  'Although I wrote them alone, these essays are inseparable from the work that Gilles Deleuze and I have carried out together for many years.  This is why, when I am brought to speak in the first person, it will be indifferently with that of the singular or plural.  Let one not see there especially the business of paternity relating to the ideas which are advanced here.  There as well as here it is all a question of "collective assemblages".  Cf.  Our book in collaboration: A Thousand Plateaus...'

Individuals are always moving from one language to another, using suitable language for a father, for a lover, for a dream and so on.  These utterances display a ' whole ensemble of semantic, syntactic, phonological and prosodic dimensions…  [As well as]…  poetic, stylistic, rhetorical, and micro political dimensions'.  Linguistic mutations appear from these languages, as a matter of frequency of use [unstructured by social orders, that is, unlike the examples above?].

The eternal distinction between synchrony and diachrony is also suspect.  As with other formalizations, linguists claim an authority by claiming to see abstract linguistic competence behind actual performances.  But there is no general structural competence.  Languages reflect heterogeneous social and political assemblages, as 'carriers of ... indecomposable historical singularities' (29).  The same can be said about structural analysis in chemistry, biology, economics or psychoanalysis.  There are no universals in transcendental form, but only 'abstract machines that differentiate themselves, on the basis of the plane of consistency of all possibles' at particular points of the machinic phylum.

What are these abstract machinisms?  We can get some ideas from early Chomsky, as long as we abandon the later idea of universal syntax, and notions of 
depth and surface.  We might begin by analyzing minor languages, including those associated with children or slum dwellers.  We begin to get an idea of the 'linguistics of desire', but this will not be grasped by formalization [and Chomsky is singled out in particular for developing a topology as a tree structure, as his system developed and was codified].  If we return to the earlier model, we can reconsider grammaticality as 'one of the modalities of the abstract power set into play by the most decoded capitalistic flows' (30).  We also have to discuss what grammaticality actually means: is it some fundamental axiom producing a generative structure, or is it a 'marker of power and [only] secondarily a syntactic marker' (31)?  There is a normative element in it, in that the only normal individuals form grammatically correct sentences, and those who don't are seen as deviant to be 'interpreted, translated, adapted' and sometimes enclosed in institutions.

Linguistic competence is seen as neutral and universal, outside of contingency, but it is difficult to think of defining it separately from performance, and thus the concept is always used to judge performance, from a position of power.  Chomsky's model does point to the excess of competence and linguistic capacity compared with actual speech or performance, but it is wrong to see this excess [the structure or machine producing excess] as somehow producing performance.  It is the other way around, and 'the machine itself is produced by its production'.  There is no 'innate faculty'.  Competence and performance interact.  Thinking of competence as 'the machinic virtuality of expression' can help us deterritorialize rigid statements like stereotypes, or syntaxes, seeing them as tied to a specific social territory, a particular local kind of competence.  Alternative competences can sometimes attempt to seize power, as when 'a patois becomes aristocratic, a technical language contaminates vernacular languages, a minor literature takes on a universal importance' (32).  These transformations affect 'all the resources of language'.

Linguistic competence is not universal, nor are speech acts.  They are all associated with networks 'of various semiotic links (perceptive, mimetic, gestural, imagistic thought etc.)'.  Statements are produced by 'a mute dance of intensities', operating on social and individual bodies.  There might be some relatively universal features [one is ' the morpho-phonological organization known as double articulation'].  It is obvious that power formations constitute particular fields of representation, however, often in an overcoded manner, displaying 'contingent relations between heterogeneous layers' (33).

Speech acts, especially Habermas's version, have been studied by modern linguistics in opposition to Chomsky and systematic competence: the latter is seen as arising contingently out of speech acts, which are regulated by individual, social, linguistic, and psychological factors [apparently associated with somebody called Brekle].  This is a step towards analyzing real speech acts concretely, which would mean an abandonment of dichotomous or mathematical structures, but we still have an assumption of universality—as in Habermas and universal pragmatics .  It would be quite some project to identify these in every speech act and in every possible one!  Habermas goes on to say that some are institutionalized in cultures or societies, and this has led some people to identify particular characteristics as more likely to be universal [a list on page 34].  This must be arbitrary, however, and again focuses on normal adult enunciation rather than idiosyncratic performances.  We should abandon any attempt to find universals in favour of 'the full acceptance of problematics pertaining to a micro politics of desire and all sorts of macro politics' (35).

What sort of power do we find in linguistic fields?  Power is not just found in ideological superstructures, and nor is it something that only regulates 'well defined social ensembles'.  Power engages 'the whole complex of "extra human" semiotic machines', seen in the power of the ego and superego which make us afraid or neurotic.  Combinations of these forms of power can produce stable layers of competence in various activities, including individual semiotic activity, that relating to social machines, machinic forms themselves, and systems connecting domains including ' deterritorializing lines of flight, components of passage, etc.'(36).

Making formal distinctions and pursuing abstract foundations of language means we miss the effects of 'collective assemblages of enunciation…  The true creative groups concerning languages', and operate instead with either individuated or universal subjectivity.  These collective assemblages can indeed display formal or individual qualities, but it would be wrong to turn these into abstract categories.  Instead, we need to examine 'the same types of processes of universalization that every power formation has utilized in order to be given the appearance of a legitimacy of divine right', and this particularly applies to expansionist capitalism.  We clearly can structuralize and binarize, but it is a mistake to think that we are describing eternal structures, things that have actually produced specific social and cultural forms.  Instead it is 'processes of power and machinic mutations' which have produced particular regions of creative potential, and these are the fundamental processes: we can make them more complex, but we cannot decompose them any further.  These machines are either connected to assemblages and transform them, or remain virtual.

In order to become stable, particular fields have to show themselves to be suitably collective in managing diverse performances, even marginal ones.  They have to engage in coding a whole range of 'overflowing' semiotic assemblages, based on 'the dominant mode of semiotization that they set to work' (37), particularly being able to mobilize abstract machines in finance or science, for example. 

These processes involve 'semiological subjection within fields of resonance' and 'semiotic enslavement within interactive fields of machinic redundancies', or some coordination of the two.  We can expect to find dominant grammaticality being imposed, ideological assemblages dealing with content, and 'diagrammatic assemblages of enslavement' operating as referents [the examples are 'flows' of abstract labour as the essence of exchange values, flows of monetary signs as the substance of the expression of capital…  Linguistic signs adapted to standardized interhuman communications' (38)].  The intention is to make each individual into a suitable speaker and listener, adopting particular linguistic behaviour that is compatible with capitalist modes of competence.  Capitalism has indeed produced a pervasive and direct language that seems natural, operating even at the unconscious level through 'presuppositions, its threats, its methods of intimidation, seduction, and submission'.  This makes the search for a new conception of the unconscious particularly urgent.

Again we have to deny that capitalist language somehow expresses the apparently fundamental needs or conditions of human beings, and see these instead as 'semiological transformations dependent upon a given context within a power system' (39), with an increasing intolerance towards alternatives.  Such language overcodes the signifying machines of the state and its institutions.  It replaces 'ancient sedimentary structures' of societies and communities with 'molecular chains of expression'[no doubt in the Hjelmslev sense, to include non linguistic expressions].  Contents appear as necessities, representing dominant opinion and 'the persecuting refrains of the ubiquitous Superego'.  Classic forms of desire are detached and transformed into the 'polarity of subject and object', so they can then be attached to social needs.  They become inseparable from the significations of mass media, and they become individualized in a fixed, non-nomadic way.

We can detect these tendencies in societies before capitalism, and anyway, such societies already showed the existence of less dominant capitalistic flows.  However, modern societies, beginning with the middle ages have led to a loss of control over these decoded flows [ones that break with the old symbolic codes].  One result has been 'generalized Baroquism'(40) 'leading to capitalist societies strictly speaking' [so some sort of linguistic determinism?  Also detectable in Anti Oedipus].  Capitalism features the 'semiotics and machinic enslavement of the flows of desire', as a response to reterritorialized codes.  This is 'correlative' to the emergence of new forms of social division between sexes and ages, divisions of labour, and other forms of 'social segmentarity'.  All forms of sense are located in a social hierarchy.  There is a constant effort to rethink and modify codes to incorporate 'in detail every significant relation'.  Children have to undergo an apprenticeship in the use of such language, including when they become sexualized and socialised.  They will encounter, for example a 'regime of pronominality and genders that will axiomatize the subjective positions of feminine alienation'.  Again there is no universal or common language between these categories.  Any languages claiming to be national, such as 'those spoken in the French Academy or on television' are really metalanguages, preserving social distance with other languages', and forcefully imposing over coding.

We need to return to Hjelmslev, but not the bit where he wants to axiomatize language. [Here is a  good discussion of this, reducing, for example, narrative form to a series of what looks like Boolean logical connectors]. And try this nice condensed summary in the irreplaceable site on semiotics by Chandler:

  • Plane of content: For Hjelmslev and Barthes, the signifieds on the plane of content were: substance of content (which included 'human content', textual world, subject matter and genre) and form of content (which included semantic structure and thematic structure - including narrative). See also: Plane of expression
  • Plane of expression: For Hjelmslev and Barthes, the signifiers on the plane of expression were: substance of expression (which included physical materials of the medium - e.g. images and sounds) and form of expression (which included formal syntactic structure, technique and style).
We can use these categories, remembering Hjelmslev's own reservations about these terms.  He seems to suggest that notions like planes of expression and planes of content arise from everyday use and are arbitrary.  In practice they are interdependent and understood only in opposition to each other.  It is unfortunate that this looks like the same as Saussure on signifier and signified respectively, which means that linguistic structures seem to dominate the whole practice of semiotics again. 

However, we can rescue the general notion of semiotics using these terms, and denying the abstract universality of signifier and signified [indeed, in Thousand Plateaus, all sorts of inanimate objects including rock strata are can be seen as offering expressions — I think this is very misleading in practice].  In particular, we should see forms only as expressed by or in substances, including non linguistic ones.  [I think the argument here is that nonlinguistic expressions, involving rocks or whatever, are seen as the original form from which linguistic forms have emerged].  We need to consider the nonlinguistic in order to grasp better  'the abstract machinism of language'(42).  [The argument might be that we need this abstraction to explain the appearance of semiotics substances of all kinds].  Hjelmslev goes on to argue that we should not see the syntagmatic dimension as a mere product of the system—again, processes are not based on universal codes, but arise from assemblages which support those codes.  We also need some 'basic materials' which can be used in the process of expression.

The issue is to decide what makes some semiotic components creative, and what might institutionalize them.  Again it is pointless to look at the qualities of language itself, since it can produce proliferation, mutations and standardization.  Nonlinguistic components are often crucial to break conformity or to 'catalyse mutations' (43).  There is no creative potential in the formal units of content.  The pragmatic level of enunciation, especially assemblages of enunciation, and molecular matters of expression are where we should look: these bring into play abstract machines.  This helps us to break with overcoded languages by going back to assemblages, where standard significations are on offer, supported by hidden power formations.  We need to preserve 'this systematic politics of "good semiotic choice"', by going back to assemblages and components that produce 'signs, symbols, indexes, and icons'

Chapter three Assemblages of Enunciation, Pragmatic Fields and Transformations

[This is very technical and I do not know enough about linguistics to fully follow it.  Allow me to present a plain man's gloss. An initial starting point might be the 'semiotic triangle', which connects signs, thoughts and objects in various formulations.  Thank goodness for Chandler who provides this:

Variants of Peirce's triad are often presented as 'the semiotic triangle' (as if there were only one version). Here is a version which is quite often encountered and which changes only the unfamiliar Peircean terms (Nöth 1990, 89):

semioitc triangle

  • Sign vehicle: the form of the sign;
  • Sense: the sense made of the sign;
  • Referent: what the sign 'stands for'.

One fairly well-known semiotic triangle is that of Ogden and Richards, in which the terms used are (a) 'symbol', (b) 'thought or reference' and (c) 'referent' (Ogden & Richards 1923, 14). The broken line at the base of the triangle is intended to indicate that there is not necessarily any observable or direct relationship between the sign vehicle and the referent. Unlike Saussure's abstract signified (which is analogous to term B rather than to C) the referent is an 'object'. This need not exclude the reference of signs to abstract concepts and fictional entities as well as to physical things, but Peirce's model allocates a place for an objective reality which Saussure's model did not directly feature (though Peirce was not a naive realist, and argued that all experience is mediated by signs). Note, however, that Peirce emphasized that 'the dependence of the mode of existence of the thing represented upon the mode of this or that representation of it... is contrary to the nature of reality' (Peirce 1931-58, 5.323). The inclusion of a referent in Peirce's model does not automatically make it a better model of the sign than that of Saussure. Indeed, as John Lyons notes:

    There is considerable disagreement about the details of the triadic analysis even among those who accept that all three components, A, B and C, must be taken into account. Should A be defined as a physical or a mental entity? What is the psychological or ontological status of B? Is C something that is referred to on a particular occasion? Or is it the totality of things that might be referred to by uttering the sign...? Or, yet a third possibility, is it some typical or ideal representative of this class? (Lyons 1977, 99)

Typically, these components are connected in the form of collective enunciations or assemblages which not only specify the content of the links, so to speak, but also privilege the legitimate ones.  Guattari is able to develop this model and add dimensions to it, and then to offer a table of the possible combinations.  I'd take this to be the 'diagram' of possible forms of enunciation.  The diagram offers machinic variations, as various connections are established between different options, including the mysterious 'resonances'.  However, there are mundane variations as well, since each of the points of the triangle can demonstrate change—objects can change as in technological progress, signs can change as languages develop, including developing flexibly to exploring metaphors and analogies and that, and thoughts can change as culture is developed.

It is also possible to see how linguistic assemblages are connected to cultural and social assemblages, and this clearly develops on from all the stuff about over coding in Anti Oedipus.  Here, Guattari sets up some ideal types societies, including the one taken to be the most primitive form, which looks pretty much like mechanical solidarity in Durkheim.  Signs and thoughts are tightly confined by a collective agreement to enunciate in a particular way and to rigorously police any deviations.  By contrast, capitalist societies are heavily individuated, with apparent autonomy given to signs and thoughts, although in practice, these are overcoded, that is subjected to dominant meanings expressed over and over in a redundant way.  Now read on…]

[Not yet...While I am here, the phrase 'redundancy' occurs a lot ,in a linguistic sense, and it normally means repetition so as to convey meaning as in 'overcoding'. However, according to a primer on grammar I found on the web, it can also refer to material that is surplus to identifications of a particular grammatical unit and 'In generative grammar, any language feature that can be predicted on the basis of other language features'.

There is also much reliance on 'resonance' as a linking mechanism between components and systems. Too much reliance in my view where it helps Guattari duck out of being more precise about what leads to what. I think of it in terms of DeLanda's dicusssion about how signals, when brought into close association, can line up their wave lengths as they dampen the variations in each other -- this is my take on the normal notion of physical harmonic resonance as in guitar strings etc. However, it must surely be used as a scientific-looking metaphor here, just to explain how all the forces converge in semiology and social and psychological systems to produce the reified capitalist subject. He might as well have said 'pixie dust'.

Content and expression are attached together in assemblages of enunciation.  Originally, before formed up human language, we have only 'components of semiotization, subjectification, conscientialization [see below] , diagrammatism and abstract mechanisms' (45).  These are joined by systems of correspondence and translation between language and culture.  Some of these appear as common sense, but there are many other possibilities.  Assemblages are partially autonomous from 'the plane of content' and their 'angle of signifiance' depends on the conditions of the semiological triangle, a way of '"holding" a given subset of the world'. Note that word 'signifiance' which I am translating in my layperson's way as the capacity to signify or to offer significations.  [Massumi's notes on translation and acknowledgements suggests  more technical definitions.  Signifiance refers to the syntagmatic processes of language, and interpretance to the paradigmatic ones in a '"signifying regime of signs"' (xix). Beneviste, who was the originator of the terms, has his work translated into English thus: '"signifying capacity" and "interpretative capacity"']

The subject itself is not just a result of the play of the signifier, but the product of heterogeneous components, some of which produce meaningful dominant realities.  Individuation is itself the result of certain social organizations and the operations of the unconscious, the way it manages 'the libidinal topics of the social field' (46).  The notion of subjects, approved contents, and the law are determined by relations of power.  There is no universal content, universal world.  For example, there are no universal mechanisms of male domination based around the operation of the phallus either—the apparent ubiquity [redundancy] of phallic forms simply shows the effects of particular authoritative institutions and ways of representing persons [which is what I think is meant by 'faciality', but we will come to that].

There is no intention to separate out and isolate assemblages of enunciation and desire.  These assemblages can deterritorialize.  However, deterritorialized desire can lead to capitalist forms of subjectification, because of the effects of 'semiological subjection and semiotic enslavement'.  It will not be easy to pursue deterritorialization in the name of molecular revolution [a note on page 336 explains why—each side of the semiological triangle contains a number of diverse apparatuses and possibilities, and any molecular revolution would have to deal with all of them—including scientific and economic assemblages, ideological apparatuses, and various established 'modes of perceptive encoding', mass media and so on].  The more obviously contingent nature of power relations conceals other attachments to these assemblages.

Each assemblage has a 'machinic nucleus', comprised of abstract machines, operating as 'crystallization of a possible between states of affairs and states of signs' (47)—like those mysterious virtual particles in contemporary physics, it is useful just to presuppose their existence.  Abstract machines produce 'redundancies of resonance (signification) or redundancies of interaction ("real" existence)', depending on whether they are fully embodied 'in a semiological substance' or exist only [virtually] on a machinic phylum.  Abstract machines like this display three types of consistency:
  1.  molar consistencies, with strong crystallization and stratification, operating through 'weak resonance', as in formal translation, and 'weak interaction' between stratified codes.  The difference between molar and molecular reflects alternative consistencies.  Thus molar forms include 'the world of stratified, identified, or hierarchized objects and subjects', which constrain the operation of abstract machines to a reproductive function: schizoanalysis at this level operates only with interruptions of consistency, including challenging the resonance between the levels by referring to 'weak  diagrammatic interaction' (48) [that is pointing out other possible combinations on the basis of the overall diagram?]. 
  2. molecular consistency, with less stratification, with consequent strong resonance, for example between the subfields inside the overall semantic one and relations between them, as in 'poetic or mystical effects', and suitably strong interaction, including particular kinds of passage [and again faciality and refrains are going to explain this].  Interactions at the molecular level mean it is not really worth separating out components assemblages or fields.  Abstract mechanisms are actualized.  At this level, constraint has not set in, including that resulting from power struggles, which gives schizoanalysis a particular 'fundamental micro political stake' in this level (49).
  3. abstract consistency, where we can see machinic elements in a more pure state, without redundancies imposed by social constraint.  Nevertheless, there are capitalistic abstractions, where a whole series of resonances and semantic fields are coordinated.  There are also [more possible] consistent ways in which sign particles behave.  As a result, there is always 'a "potential possible"' which is never fully dissolved in existing fields and components, and which is found in various uncaptured 'matters of expression' [the example turns on the ways in which computers have been captured by instruments of social control].  Schizoanalysis aims to transform this situation by displacing consistency by emphasizing potential passages and new machines to challenge existing semiotic fields.  [an obscure diagram on page 50 summarizes the argument, not very helpfully]

Combining the molar and molecular and the abstract types of consistency with the notions of redundancies of resonance and interaction, provides the six cell table (51) (see below).  These types of fields are interrelated and dynamic.  Some can 'swell' until they are totally redundant or powerless [cells in the second column?], asignifying components can develop from signifying ones, and pure abstraction is never possible [bottom right cell,column 4], even with computers which always retain 'some semiological terminals that are human'.

We can see what happens in these interacting assemblages by looking first at those that express capitalistic power.  Capitalism develops 'a world of simulacra' from semiological redundancies.  It is the passage between referents, expressions, and representations that produce these simulacra.  This process even 'simulates diagrammatic relations' (52), but without any critical potential, since it claims to represent the actual world with no room for creative dissent, only its occasional version of 'Life, the Spirit, and Change'.  Any abstractions will look synthetic and arbitrary [one example is '"extremist" religious assemblages'], and any other abstractions are likely to be reterritorialized.  In particular, subjectivity and objects of desire are framed, by connecting them, through what looks like reasonable resonances, to control mechanisms, such as superegos, exchange values, or 'an imaginary museum'.  There is no attempt to connect them to any absolute [as in mechanical solidarity] , but rather to 'coordinated systems of power', which manage lines of flight and lines of dissidence as an example of  'oppression, distancing, autonomization, and alienation from an increasingly domesticated plane of the signifier' (53).  The process of homogenization also involves splitting individuated subjects from assemblages of enunciation, the reproduction of conservative signifiers, sometimes drawing upon limited versions of the diagram, and overcoding of speech and writing through the control of syntagms and paradigms.  However, there is always a possibility of new kinds of deterritorialization.

Capitalistic abstraction therefore operates to neutralize and recuperate lines of flight and machinic possibilities ['machinic indexes'], and is linked to particular institutions.  Religion is one, but now there are more fragmented and diversified mechanisms including 'public infrastructures and facilities' and the mass media.  Together, these produce 'a gigantic net, composed of points of potential signification', and it is impossible to escape, because all assemblages of enunciation are woven in.  There is a constant alteration and recalculation so that the 'thresholds of deterritorialization' are always trimmed for the established order and made coherent.  Lines of flight will be confined within this horizon, and machinic assemblages will have to deal with contents that appear to be universal.  [Sounds like an elaborate way to describe hegemony].

If we turn to Hjelmslev, we avoid the simplifications of signifiers and signified, and we can extend the notion of content and expression 'to all the various assemblages of enunciation' (54).  There is more to 'semantic reality' than just signification.  Semantic fields refer to particular operations, particular transformations—'analogical interpretations' these involve redundancies of resonance and interaction.  Particular assemblages can be derived from signifiation processes or 'diagrammatism', and all actual assemblages affecting humans are mixed.  Possibilities are closed off by power and an emphasis on the subject [the 'personological pole'], but there is always a potential for new possibilities, since abstract machines always exceed any apparently 'fixed and universal coordinates' (55).

We can build on the pragmatic component suggested in the Introduction to produce 'four general types of mixed assemblages of enunciation'[sigh].  Everything depends on how content is provided:
  1. In 'analogical generative transformations', semantic contents envelop referents, and are generated by 'fields of interpretance'.  They are territorialized and collective assemblages [as in mechanical solidarity or childhood assemblages before the full development of language].
  2. 'Generative linguistics semiological transformations' develop more signifiance on the plane of content.  The referent is not seen as the same as the signifier.  The syntagmatic dimension is better developed, and individuated assemblages emerge from the collective ones. 
  3. 'intensive symbolic and asubjective transformations', where there is a more reflexive grasp of the links between enunciation and referents [entirely my own words here], helping us seen more diagrammatic or machinic dimensions.  This can involve a transformation of existing semiotic components, as in 'mystical or aesthetic desubjectification'.  Assemblages of enunciation are still collective even if they appear to be unique to one individual, since the individual still really operates as 'a non - totalizable intensive multiplicity'
  4. 'asubjective diagrammatic transformations', featuring 'asignifying contents' these can deterritorialize the other assemblages of enunciation and 'machines of expression and semantic formalisms'.  They are connected to referents on the 'machinic plane of consistency' (56), and appear as 'machinic assemblages of enunciation'. [NB I think the 'asignifying' bits mean two things -- non-linguistic things like musical or scientific instruments that still express themselves in the work of Hjelmslev, and linguistic bits that have no external meanings  but make sense only as internally consistent performances -- hence D&G denying that their books signify anything]

These different assemblages are not that discrete.  They can deterritorialized, as a result of 'a signifying component (example: groups of educated children)'[wha? Some sneaky way of reintroducing the creative subject?].  They can inject asignifying machinic elements into symbolic components, as with drug induced meditation or 'diagrammatic writing in repetitive American music'.  They can alternate, as when they slide 'from subject groups toward subjugated groups' [I think this is an example of what he calls the 'individuated economy of enunciation'—the impact of economic factors in other words].  These alternations and differences arise because they represent specific states of the diagram, specified in an attempt to maximize their efficiency, as a result of the operation of a micropolitics [the example given is poetic enunciation, which operates with particular symbols and modes of subjectification, associated with particular regimes of signs].  The table that results also helps us fit in other linguists and their categories [which I have skipped].

Transformations are possible from one type to another, for example from (3,in the list above) symbolic intensive into (1) analogical [the example is when a fascist group colonizes and monopolizes particular symbolic components].  Diagrammatic (4) can transform into semiological (2) , which strengthens ordinary semiological understandings, and the machine is 'subjectivized' [the example is when a writing machine is operationalized by an author -- I think].  Signifying discourses can become asignifying if the semiological (2) transforms into the symbolic (3), avoids existing formalisms and explores new possibilities.  Semiological (2) can transform into diagrammatic (4) and this is where common understandings join up with diagrams to escape the limits of analogy and overcoding.  [This is a kind of creativity that explains scientific or musical innovation -- and revolutionary politics?].

Particular combinations of components and assemblages have produced three 'particularly important limiting fields' (59).  We have to diagnose more elementary components to consider more complex possibilities.  These assemblages have certain lines of passages, which have become typified [reified] by social scientists.  Another diagram summarizes the three 'pragmatic' fields, page 60. [and this is going to link to the attempts to understand social forms as codes as in AO?]

We can understand territorialized symbolic fields by considering elementary cases, including those of childhood madness and 'archaic societies'.  These fields occupy stratified territories and are unified by articulated codes, and by endless elaborations 'of a substance of universal expression'.  So primitive societies can develop myths which are not just the result of articulated codes of gestures or perception, but are related together by how the group has been territorialized.  The group becomes the signifying substance and they act out different semiotic combinations as 'a sort of pragmatic rhizome'(61), but one which limits any escape route [and the example is given of an 'index' which prescribes particular responses to events, including what to do if the initial responses fail.  This is also described as 'a "signifying synthesis"'].  There is no hierarchy established between these semiotic activities.  Mechanisms which deterritorialize can be tolerated, but are treated as the same as those which territorialize.  All possible disturbances are excluded, even elements of diagrammatism.  Symbolism is not open to analogical or semiological processes of translatability and transformation.  Any higher levels of signification are monopolised by existing chiefs or 'mechanisms of semiotic enslavement' such as the restriction of access to technical or writing machines.  Deterritorializatiion is subject to the desires of the group.  Conventional ethnography has not grasped how this can lead to a resistance to external factors like new religions, and have tried to explain such resistance in terms of social grids, or notions like 'invariant significations and stable relations of exchange' (63).  But the contents of the symbolic fields are connected to each other differently.  Seeing this connection between symbolic contents and social roles is an 'accomplishment of the hegemony of capitalism in the 19th century', a clear example of a type of 'dictatorship of the signifier'.  These days we know that non western semiologies can reverse colonial understandings, and the same goes where 'dream semiologies' assert themselves over more conventional kinds of semiotics.

It is conventional in linguistics to assert that iconic components depend on linguistic components, that symbolism is always linguistic.  Instead, the subordination of images to linguistic expressions is the result of particular conditions: one such set of particular conditions separates language from acts of speech, but Saussure has universalized this.  There is no necessary type of transformation [as above]: everything depends on the way in which enunciation and intersubjective communication are individuated [ at the level of individual groups here] .  Semantic fields are only moments of transformational fields, and they rely on 'a certain type of asignifying machine of expression'.

Signifying, individuated, and abstractified fields have been described already in terms of capitalistic assemblages.  The old territorialities are transformed by signifying systems.  Indexes are abstracted and connected to each other [systematized] .  Abstract codes produce particular potentials for icons and indexes.  All social life is open to such abstraction.  Analogies represent only the first stage of translating semiotic links into each other.  Even signifiance is managed as another modality of reterritorialization and subjectification of contents: the latter organizes systems of 'double articulation', in paradigmatic and sytagmatic coordinates (64), and these are more deliberately articulated with assemblages of enunciation at this stage [more so than in capitalism].  At least analogies preserve some autonomy for strata of expression, and there is no imposition of a final signified, embodied, say, in a dictionary, or controlled by rigorous rules of linguistic combination.  Asignifying sign machines remain arbitrary, to some extent, at the level of expression.

As a result of this degree of looseness, also found in analogies that are not particularly well regulated, certain elements of diagrammatism remain, in a mixed system.  It is possible to find 'empty signs without semantic content, for example the phonic or graphic images of the word "table" are seen as a table' (65).  Diagrammatic possibilities are limited because they are seen as 'quasi objects'.  However, there is no symbolic authoritarianism.  Syntax and logic can operate on significations and propositions, and every day reality can be extended by pragmatic implications. 

All this is soon to be captured  by capitalistic economy, however.  The formal subjectivity accompanies spectacular social rituals, and an abstract matrix based on differences emerges.  After a state machine develops, signifying power becomes more autonomous, regulated by the equivalent of a 'stock market'(66).  State machines operate at the molar level, to stratify and restratify, and at the molecular level where it 'articulates and controls all the cogs of the economy and society'.  Molar operations are regulated by 'a collective semiological substance' that offers differential relations and degrees and types of power.  Molecular ones are regulated by sign particles that interact in order to recuperate territorialities: for example different sorts of time are converted into the quantifiable time of work values.

Abstraction is the key to managing intensities and releasing them in particular linear and flat 'old territorialized rhizomes'.  Linear models and preferred sequences are imposed, so that, for example older forms of communication in primitive speech, combining song, gestures and mimicry, are decomposed, broken into elements and then rearranged in a syntactic order.  Deterritorializing potential is measured and compared.  Layers are organized in hierarchies.  Non linguistic components are neutralized, and any that escape are controlled by particular kinds of 'diagrammatic machines'.  Whole networks of semiotic operators [social and cultural organizations] control any kinds of escape, and only permit new machinisms or sign particles if they are compatible with dominant systems [including 'systems of abstraction and formal syntax'(67)]

[The example is the way that new ways of writing music introducing polyphony and harmony were originally constrained by appealing to some vague notion of 'suitable temperament', together with an effort to fix orthodox forms of musical syntax through lessons and academies.  This compromise eventually broke up with additional processes of deterritorialization, in the form of abstraction that we have already discussed, ways of stripping components out of contexts and reorganizing them 'according to a machinic rhizome' (68).  An alternative seems to be to impose a tree structure 'of implication'.  Apparently, these are the twin dangers affecting any attempts to get machinic in signifying fields—an uncontrollable proliferation on the one side, or a 'sclerotic syntaxization' on the other.

Diagrammatic, collective, machinic and asignifying fields are when sign machines operate without the 'significantive processes of subjectification', and these examples become more common with 'each passing day'.  We now realize that more normal semiotic machines and material or social machines can both be traced back to 'the same type of abstract machine', and this is to inform 'political pragmatics'. 

The old distinctions, like the ones between matter and form become irrelevant.  Polyvocalism takes on a new form, no longer connected to various territorialized assemblages, like the personal or the technical.  Instead, we know operate with machinic populations and non human machines.  Despotic overcoding is no longer possible.  Everything develops on the plane of consistency of abstract machines and their potentials.  Stratification increasingly has to respond towards deterritorialized elements.  Fully machinic rhizomes develop, showing vectors related to 'multipolar, multisubstantial, multidietic coordinates' (69).  There is a machinic process of destratification challenging hierarchies, since machinic components cannot be connected to stratification.  They develop their own phylum.  We can study this by considering the virtual, the theoretical and experimental dimensions of analysis.  Matters of expression themselves become diagrammatic, as well as semiotic contents, and it is no longer possible to separate out apparently material intensities from human and emotional ones: rhizomes connect assemblages ['intensities' (70)]
in all of these areas, without privileging any of them.  We should not see material assemblages as more territorialized than semiotic ones.  However, some assemblages are more intensive than others, so that material assemblages can constrain, but only temporarily, mathematical machines.  However, generally, 'abstract machines fully exploit their capital of possible'[sic].  Codings, and molar stratifications no longer apply, and even subjective feelings are affected, as when a scientist falls in love and this affects his research.  [We also find for the first time the mention of 'black holes' of subjectivity, which I understand here as meaning dead ends, where all signifiance ceases: erotic lines of flight can overcome these].  Passions are no longer separated from public life, and the special category of intimacy ceases to be effective.  Psychoanalysis started to investigate these interactions, but 'quickly stopped along the way'.

Pragmatics is not just about human communication.  Communication presupposes a relation between 'things and signs' (71).  Semiological assemblages are special only because 'they produce fields of redundancies of resonance on the basis of sign machines' [that is they can name and discuss an awful lot of things].  General pragmatic fields have four principal categories, which nevertheless interrelate [more obsessive classifying].  There are two generative fields, 'dominated by semiological components', 'fields of interpretance' relating to semantics, and 'fields of signifiance'[where social and political organizations underpin particular meanings?].  There are also two transformational fields with no interpretative components—'symbolic fields and diagrammatic fields'.  And another table, extending the earlier one.

What happens is noninterpretive transformations intrude into interpretive and generative ones, especially analogical and signifying fields.  We cannot reduce these pragmatic fields of enunciation by insisting on some universal subject positions as in Lacan, however.  We distinguish these fields only as a result of a 'methodological necessity', and to insist there is no priority.  We want to produce 'a "rhizomatic" analysis', and could have done this from other resources including 'black holes, faciality and refrains, power formations' (72).  [I think I will reproduce the table on 72]: (NB 'ex' means 'example')

Pragmatic fields interact with some of the earlier fields we have discussed, to add a certain flexibility.  For example, territorialized fields feature analogical transformation, it was argued, but there is more than one type, and it's possible to use elements from 'symbolic, diagrammatic and signifying transformations' [the example is not at all helpful, and considers the speech of primitive societies and their ability to refuse systematic significations, relying indeed on symbolic techniques to resist reductionism, but at the same time opening the possibility of 'a signifying economy'].  The second example turns on individuation, which takes specific forms according to the dominant type of transformations, but still retains 'deterritorialized symbolic transformations (similar to the Gestaltist Figure/Ground)'.  It is also possible for the diagrammatic generation of additional symbolic formalisms, which can restore some notion of content to the 'black hole of subjectification'[none of the transformations here make any sense, but they include faciality and refrain again, so perhaps clarity awaits.  Other transformations seem to refer to couples or paranoia.  There is also a reference again to this work that I cannot find defined anywhere - 'consciential'.  {I have seen it connected to the term 'conscient', which means just 'conscious', sometimes in the sense of being self- aware. Maybe 'consciential' just means referring to the development of consciousness in this sense?} Such subjective deterritorialization is often stabilized by collective appreciations].  The third example suggests that diagrammatic pragmatic fields are still 'haunted by subjects of enunciation' even though they have been already exhausted by individuation—for example the continuing idea of a 'listener - speaker' (73), or the notion that speech is a matter of the mouths of individuals.  Machines are the real source of the enunciation.  Nevertheless, even diagrammatic fields feature 'artificial reterritorializations', usually appearing as mixed assemblages.

[After all that appalling stuff, I don't think we've got much further than the schema in Anti Oedipus, which combines social formations with linguistics operations, probably by reducing social life to linguistic operations, as long as we allow for the fact that it's not just humans who communicate.  There seem to be three possible combinations of social and political organization with these ways of making sense: in mechanical solidarity [what they insist on calling "primitive" societies] there is a despotic system of coding, where the collectivity polices systems of enunciation, represented by the chief or leader; then there is a colonial phase, where new forms of coding intrude, and there is a thoroughgoing attempt to understand everything in terms of these new codes—this is over coding; then finally there is a capitalistic phase, when codes take on a certain autonomy of their own and are not coordinated, except through markets.]

Chapter four Signifying Faciality, Diagrammatic Faciality

[Delirium or schizo flow at full volume here, as our hero attaches all sorts of weak significant links to the notion of a human face.  It seems to go like this: first we can pick up on existing critiques of the individual, maybe even borrowing Althusser.  Then we render the  core of individualism as the face, by insisting that facial characteristics are really important in establishing one's position in social life.  Here, Guattari is presumably speaking on behalf of the French middle classes, acutely aware of the appearance of their faces?  There is a brief flirtation with the notion of the face as a landscape.  Then we can conceive of the face as a dual system of signification—a white screen, with black holes for eyes. {Eyes are centres of subjectivity as in 'black holes', the screen refers to the cultural and social contexts where subjectivity expresses itself?}. It works well enough if we ignore ears, noses and mouths -- Guattari must have been impressed by those over-exposed close- ups of women's faces where most of the detail is washed out by the light  The notion of a screen helps us pick up some paranoid stuff about the media.  The black hole is a glamorous way of referring to some notion of personal semiotic collapse into madness.  Luckily, the signification involves a binary system, so we can pile in against binaries and how they confine rhizomes.  Right at the end we find faces are good at delivering binaries because they operate with them -- seems to be some sort of echo of the stuff in sociology about how individuals build up dyads of selves and others or it might be a borrowing from Husserl (as Deleuze does) talking about how perceptions develop from a reciprocity of persepctives. It is all very much held together with sticky tape -- Blue Peter philosophy!

Finally, we can pick up on all sorts of Foucauldian paranoia about controlling gazes.  Voila!  A system is complete.  It all stays stuck together, floating along, no doubt on a plane of consistency {the subjective definition of, that is}   When things look ridiculously speculative, homely examples are provided, like the ways in which women have to use makeup in order to achieve a presentable face, or how black skin on faces help us identify -- black people.  There are the offhand references to anthropological study of 'primitives', and throwaways to various pictures or novels.  This is also Guattari at his most conservative, with evil capitalism controlling and dominating all attempts to break out from the system of faciality.  Of course, there are always abstract possibilities of lines of flight, and a face with a long hair 'shook the world' in the 1960's but, by and large, capitalism seems to have it covered. It even seems to survey everything with an empty controlling eye: capitalism as Mordor!]

Faces can be important socially, and we use them to judge people as trustworthy.  This is a form of territorialization.  There are other connected signifying substances such as voice or accent—and refrains.  'A voice is always related to a real, imaginary, or composite face' (75).  In particular, a consciousness is often based on the triangle that connects eyes, noses and foreheads.  This has apparently been recognized by gestalt psychologists, especially in working with infants.  None of this is natural or universal, of course.

'A certain module of faciality, with its tolerated deviant types' (76) dominates all forms of expression, and determines 'the strategies of the subjection of desire'.  This is the triangle mentioned above.  Specific forms of expression are based on this more fundamental form.  We can consider it as a system based on a circular white screen, on which resonate 'the semiological triangle, the ego, and the object'.  Thus 'a facialized consciousness' develops, circulating to pursue new notions of the possible by connecting to singularity traits around certain black holes.  Thus the capitalist notion of the individual, '"person"', is constructed 'in a fundamentally Manichean way', and there is no alternative except nonsense or the end of social life.  This 'abstract faciality speaks at the heart of speech', and manages the black holes of semiotic collapse while constructing 'personological structures of power'.  This is a binary system and it supports 'the universal translatability and responsibility of statements', as a 'subjective totalization', reducing everything to 'my consciousness'.  The system has the effect of apparently naturalizing meaning and things, and thus helping to reproduce 'the reigning socio - semiotic order' (77).  Bodies are shaped to strategies of power and politics.  Faciality employs redundancies and resonances [rapidly becoming the all purpose term for a relationship which avoids any of the inconveniences of having to say whether one causes or determines the other], although it 'occupies the determining [sic] place within pragmatic fields' to do with sex and social bodies.  As usual, it relates both to these global binary and phallic systems on the one hand, and more singular traits. 

There might be particular 'faciality- types', including the normal everyday one, which sometimes gets interested in other lives, say in the past, which seemed equally normal.  Even here 'the normality of yesterday supports that of today' (78).  This is because currently normalized faciality always superimposes itself on landscapes and territories.  It does this as it 'wards off the black hole of senselessness', and this is why alternative significations have to be domesticated.  The threat of semiotic collapse is ever present, with any individuated enunciation.  As a result, 'modern intersubjectivity is now primarily founded on a vacuous faciality, a blind face-to-face between two absent gazes' (79).  Facialization is therefore political at the molar level, operating to produce a consensus on signification, and at the molecular level, trying to avoid 'signifying traps'.  Both forms are produced to control deviant individuals with incompatible faciality traits.

Capitalistic faciality unifies modes of subjectification, places them in systems of coordinates and hierarchies, developing 'an empty semiotic screen', a screen of reflexivity, on which can be displayed individual consciousness.  Any threatening components, including asignifying ones are blocked.  It is this that homogenizes significations, regulating the plane of consistency and the machinic phylum of matters of expression.  The face is as important a regulating mechanism as the phallus, but it operates differently.  Face, phallus and [capitalist] self consciousness all reterritorialize abstract flows and produce an illusion of power in the form of an imaginary appropriation.  However, they have different origins, and only seem to be the same because they have been standardized 'by the action of power formations with a hegemonic inclination' (80).  In particular, facialities are connected more generally to interactions in social fields, and 'the face is always tied to a landscape as its foundation': you can travel from one faciality to another as different social circumstances require different expressions.  All these are collective, however, shared even with animals, and affected [ie showing affect?] institutions, such as the '"a priori" faciality of the doctor', or the robotic gaze of the policeman (81).  Phallic functions are grounded on the asignifying system of the socially divided sexes, and these can be personalized by faciality traits like facial expressions, grimaces, winks and so on.  It is not a matter of thinking in terms of theatrical masks: actual persons and intersubjectivity are themselves semiotic components 'assembled from machines of capitalistic abstraction'. 

Again, there is no universal subjectification, only individuation at the intersection of modes of production of the semiotic and social.  There are the redundancies of faciality, the binaries of the phallic, refrains, and nothing else 'except a hopelessly empty consciousness (the Sartrean "being-for-itself")' (82).  Consciousness struggles against black holes, but does not intervene in modes of production.  It is caught between resonance with these modes of productions, producing 'pseudo - territorialities', and permanent deterritorialization, between tranquilised faciality, and 'anxiety which aims at the reality of nothingness'.  The media play a major part in producing the former.  The whole system tries to mask semiotic subjection, through normalizing faciality, self consciousness, and feelings of belonging 'to a "mother language"', and controlling the range of assemblages and components. A constant adjustment goes on there to control turbulent desire, ranging from domesticating the 'animalistic facialities' of childhood, through binary options for adults, to the actions of the mass media in producing 'substitute ritual and totemic facialities' (83).  These replace earlier territorialities like clans or ethnicities.  However, there is no reification or alienation, because there are no remaining 'instinctual representations'.  Instead, capitalist subjectification develops from a diagram of its own, aiming at specific functions for specific power formations.  The possibilities of the rhizome are neutralised, including its 'animal, vegetal, and cosmic eyes'.  The media purify representations, arrange them according to 'dominant coordinates', and construct 'a vanishing point' for everything else.  These representations resonate with 'all local significations'[major paranoia well installed by now].

[The bullshit is thick and deep in this next section, smothering all the self supporting assertions].  Faciality is not iconic, but signifying. Speech and language is never just about transmitting messages.  Faciality and facial substance links elements of discourse to dominant significations, or dismisses them as nonsense: in this way it stabilizes signification.  The way it does this is by sorting out options for the possible, one relating to the past and one to the future, a binary.  This regulates 'paradigmatic [in the linguistic sense, a set of associated meanings] equivalences and axioms' (84).  To do this, faciality is detached from other components, and then neutralized so as not to interfere with those components, the better to regulate them into 'standard deviations and passages' (85).  In other words, faciality is constrained itself, except with schizophrenics or autistic children.

Before modern '"mechanization" [the bad type] and informatization', various 'mnemotechnic montages based upon the domestic environment' organized the 'mental space of reference', but these have been deterritorialized, leaving a modern system of 'dichotic deduction'.  We can see this, apparently in some current university exam techniques, aimed at producing not necessarily correct answers, but 'normal' or acceptable ones, representing 'the aptitude of an "elite" for cutting a "brave figure" in all circumstances', a particular example of how facial syntax is compatible with political and social semiotics syntaxes [just asserted].  Apparently, the integration of language and face can take place according to what is proper, defined by different institutions and agencies, including institutions.  These can then feed back into the 'collective Imaginary' (86) [Foucault on stilts].  Faciality acts as a superego, regulating enunciations, by making them resonate with dominant discourses.  Deviants have to be managed, even if they arise from 'a resistance to social learning…  a rejection…  of power'.  Facialized consciousness plays a key role here: it produces a 'landscape - face', which is apparently 'focalized in a [subjective] black hole', the management of which results in 'the illusion of a homogeneous world of signification'(87).  It also produces a facial syntax, providing an 'illusion'that the abstract machines rely on 'centralising structures, a monosubjectivism and a monotheism'.  [There must be such regulation in an evil system like capitalism, and therefore there must be a mechanism: it is this that is 'discovered' in all this blurb about faciality].

'Natural sociality' [whatever that is] and excessive machinic effects have to be neutralized.  The substance of the signifier becomes the whole of the possible.  No rhizomatic connections are permitted, and nor are other 'eyes'—a central, empty eye dominates capitalist faciality [welcome to Mordor!], and it 'knows of only one thing' [an eye knows? Eye, nose and throat?] that everything must conform to the dominant notions.  There are no mysteries to this 'inquisitorial gaze of the signifier'.  It's just like Big Brother or other dictators.

Landscapes are facialized in this dominating way, so that even new facialities, as in abstract art, are themselves 'interpellated' (88) by the dominant form in the present [apparently, Proust attempted to return to an earlier form of human faciality].  Only minor forms of resistance are permitted, though—Proust's experience is like a drug experience, beginning with a noise, speech, or gesture, and then releasing 'an efflorescence of intensities of desire'.  These examples show that the regulation of enunciation is never total, but depends on 'particular contexts and micro politics'(89).  The system does its best to focalize in such a way as to make all forms of subjectification subject to 'the regime of a general culpabilization'[in the sense of responsible for, or guilty about?].  In this way, all the specific elements can support each other and add together to produce 'inhibition and powerlessness'.  The development of a strong ego shows one process; 'fundamental images', and 'symbolic structures' offer other examples.  In particular, Anglo Saxon psychology domesticates the subjectivity of the child, imposing dominant faciality to control relations with others and her own interior [through the '"four - eyed machine", of eye to eye contact]

Territorialized pragmatic fields also divide an inside from outside, of reassuring kinds to establish harmony [the outside 'invests' the inside, and the inside is installed outside].  All the threatening subjective forces are 'inscribed' on to the regulated territory of 'social, religious, sexual and ludic activity' (90) [Marcuse on stilts now].  Sometimes, the landscape can be represented by a particular 'prototypical face'—Jesus, or the President.  Polyvocality is managed in this way.  All earlier territorializations are replaced by functional substitutes such as the nuclear family.  The 'third dull I' also colonises ethnic territories.  Discrimination is sometimes obvious, but also produced by constellations of facialities.  This is another job of the media, to adjust facialities, make them fit with institutional ones, without producing dysfunctional conflicts or complacency.

It is a matter of accepting binary options, or nothing, sense or nonsense.  This is how any new semiotic sign particles that might threaten the system are woven back into it, even though capitalistic faciality can produce threats of its own [as an abstract possibility, because it continually needs to adjust and doesn't always get it right?].  The weaving is not always complete, and it has to pass certain thresholds.  At the concrete level, power formations might show certain conflicts among themselves.  The options are located only in fuzzy sets, and need to be worked on at an elementary level before they can be incorporated into 'propositional dichotic operations' (92). 

The problem is that 'capitalistic messages' must make sense to individuals, but also to the general system.  Obviously, all archaic or animalistic facialities are deterritorialized, and new institutional versions, appearing in a landscape of universal reference, replace them.  However, this can also cause a crisis for subjectivity for some individuals, 'black hole effects—semiotic lacunae'.  There is a general problem with any fully 'unifying and reductive fusion'[however, it always seems to work in the end?].  Human signifiance also needs to be maintained as something universal [Habermas on stilts!].  Faciality does this, and so do refrains.  In this way gazes are doubly articulated, prevented from becoming lines of flight, but also prevented from becoming over integrated. 

This is an 'empty faciality', which acts as an attractor to the more specific kinds, but it is this tension that provides 'the supposedly essential anxiety of the human condition' (93).  Specific facialities are threatened with collapse into a black hole, or are 'subjected to the tyranny of facial redundancies of symbolic, personological, and Oedipal identification'.  All these options are simulacra of persons with proper names.  Capital and the libido join together here to produce stable kinds of individuated enunciation [but we have to understand libido as an effect of semiotization, and that faciality also can draw from unconscious components --everything links to everything else to avoid any challenging prioritization, which makes you wonder why the fark we spent so much time distinguishing the terms].

In this way, 'an entire syntax crowns hierarchizes, and adjusts to the various normalizing power formations'.
There is historical variation. Facial formulas become 'prototypes of men, women, and children', for example, and are seen as normal at particular times in particular places.  This is the only kind of subjectivation compatible with capitalism, and the formulas manage the possible and its heterogeneity, without simply overcoding them.  Ultimately, the system is one of 'centralization, an arborifiation, a gridding, and a finalization of all the means of expression over the signifying substance' (94).  Threats are projected on to deviant facialities.  A woman has to adopt particular features of faciality to become sexually available in an acceptable way, 'thus expressing her submission to phallocratic powers'. 

However, 'In 1968, a faciality with long hair shook the world' [so what happened to break this apparently omnipotent system?].  Some cultural breakthrough was achieved, challenging conventional notions and advancing new 'amazing propositions', relating to the environment, childhood, and homosexuality.  Clothing and ways of speaking also expressed this new view, but these became relevant only when ' reframed on the sensitive deterritorialized slab which has become the face'[always?].  [Guattari leaves 1968 inexplicable here—it is not not even one of those exceptions that prove the rule as below]

In capitalism, however 'everything must be foreseen and made calculable'(95), so that syntax is developed systematically so that it can arrange faciality.  Facialities, corporeality and 'landscapity' are no longer tied to local territories but are manipulated in a 'pragmatic syntax of bipolar standards and deviant types'.  Institutions developed to manage deviance, and they depend on the 'laws and jurisprudence governing the dominant faciality'.  The redundancies involved are focused back and embodied in 'a "supreme faciality"', and those that do not fit are ignored by 'the empty eye of power'.  New metalanguages in psychology and law are needed to manage this deviance.  Generally 'The universe of dominant significations does not tolerate any escape over which it lacks control', and all expressions have to support it or risk the consequences: 'it's all or nothing!' (96). Signification becomes univocal, and it gains social truth only if it is validated by the central system.  Soft forms of control are mobilized if harsher ones do not succeed or cannot be repeated, through 'Ambiguity, persuasion, half lies, and half truths'.

Focusing on specific facial movements and what they tell us about 'facialized conscience', smiles can be interpreted as 'an insane grimace or insolent mockery', submission can look underhanded, 'a pout beyond the standard will become a mark of contempt; too old a face or too wrinkled will inspire fear', and 'tan skin, beyond a certain threshold, will spark mistrust' especially if accompanied with a deviant accent.  Sexual choice is posted on the face so that no threats to phallocratic power may arise [that might need revising!]. Capitalistic control even extends to 'the texture of matters of expression' (97), using the features of certain linguistic systems to offer a complex exercise of power and subjection.  If all was to work or well, every expression would end up as a term on a series of 'automated binary choices likely to be computerized'.

Faciality explains how deterritorialization proceeds, helping to stratify different territories, relating inside and outside, for example, or events to 'syntagmatic articulations' as particular activities of the signifier.  'No new semiotic conjunction is conceivable, no creative nomadism, no "amazing" encounter, no flash of desire' (98) [really gloomy stuff then—activists will be cross].  All is controlled by the ultimate 'threat of annihilation'.  Supreme icons are developed to manage and reduce the dimensions of multiplicities, or anything created.  This is outside capitalistic signification and redundancy, and thus 'order will be guaranteed forever, everywhere'.  Even really insignificant events are colonized by faciality components, and this is the reverse of what is normally considered to be the transmission of information: all concrete options within a body of messages are the result of divisions from outside.  All statements can be reduced to binary forms, but it should not be thought that the binary is natural—statements have to be carefully prepared first, by managing originality, and reducing content to expression.  The model of the 'facial "treatment" of speech' (99) is one where particular traits of expression are mapped 1 to 1 with a definite set of semiotic objects, with all other possibilities provisionally abandoned: a binary treatment can then ensue [which part of the set maps the expression—first half or second half?]. 

Faciality 'creates the conditions of reduction for the entirety of reference'—it is something taken for granted that abstracts the value of things, manages intensive multiplicities and their heterogeneity, and coordinates them through binaries [still a lot of fatuous speculation in my view, where this 'must' follow from what has been said earlier, and so on. It all seems to depend on people agreeing that faces are simple and easy to read and so everything else must be].  The effect is to make dominant semiologies look simply logical, and conceal their origins in relations of power.  The system works to produce discrete objects through a 'dichotic method of opposition'(100) [just a repetition really, 'dichotic' = 'binary'].  These are not produced either by theoretical data processing nor by 'diagrammatic microbreaks'[which awaits below, apparently]. 

So far we have been discussing facial sets at the molar level, which produce gestalts rather than logical or diagrammatic groups. They work like decision trees, with 'a series of yes or no questions'.  This works because of 'synchronic waves of redundancies of resonance'[or fairy dust], and faciality is important because it is itself binary—I see another and therefore learn about my own faciality, so that 'Every effect of meaning is instituted between two deterritorialized winks' (101) [isn't this the old reciprocity of perspectives assumed in interactionist sociology?].  This is how we develop classifications of our perceptions, as a kind of 'facial syntaxization of content', and obviously much content is overlooked [my term— see, I can do it!].  The system of proper names helps to offer a suitable categories of clusters of qualities, and this is very useful for capitalistic semiotization: in archaic societies, every substantive acted like a proper name, but this has been broken by deterritorialization and individuation-- 'nothing appears proper to any group or any person'.

Diagrammatic faciality traits work differently, through 'machinic redundancies and abstract machines irreducible to data processes'.  The redundancies subject desire, and the abstract machines release it in acceptable ways.  Here we move from trees, simulacra, and black holes to consider rhizomes.  Signifying layers are no longer arranged in concentric zones or controlled by empty eyes.  Instead, there are lines of flight following diagrammatic possibilities, and 'a becoming - imperceptible' which avoids the effects of the gaze [ the old postmodern desire to disappear, I thought, in my notes on TP] .  Signifiers are no longer tied to referents, and we can develop new types of 'semiotic break', based on contact with matters of expression.  Material flows are in direct, unmediated relation with systems of coding.  As a result, machinic indexes are openly articulated, not neutralised.  As a result, they can be used to open new possibilities, breaking with systems of stratification, linking with each other, and 'launching semiotic bridges between matters heterogeneous until now' (102).  They permit the emission of sign particles and the growth of networks.

These machinic rhizomes are not simply at another stage in the evolution of capitalistic organizations of meaning.  They already existed in archaic societies.  There are constantly subjected to the development of new arborescent systems, even in societies that rest upon them [is this just the familiar argument about revolutionary societies being open to subsequent bureaucratization?  Weber on stilts?].  However rhizomes in archaic societies were still territorialized, unlike modern ones.  Powerful groups now have to constantly monitor the operation of such rhizomes, especially focusing on 'the adaptation and recuperation of asignifying machinisms' (103) [computers?], so there is no fixed opposition between machinic rhizomes and earlier semiotic systems, no inflexibility for the first, and no eternal resistance for the second.  Capitalistic systems were already beginning to develop 'semiotic quantification', for example, and this is now inseparable from 'machinic quantification'.

As a result, it would be wrong to see the 'data processing revolution' as inherently radical.  Nor can it be simply opposed by 'the humanistic conception of science', based on some eternal division between human beings and machines.  'Today, machinic conscientalization has become inseparable from human conscientalization'[very debatable], computers have become more and more integrated into enunciation, and 'it will become almost impossible to make a distinction between human creativity and machinic invention'.  The only barrier at the moment is an insufficiently quantified semiotization, but computers are developing [a note references a French theorist, writing in 1979, predicting that computers will soon be able to manage undecidable propositions, relatively short and discrete theorems, and 'propositions uniquely demonstrable through gigantic calculations'(343)].  Machinic conceptions are already affecting human semiotization [much more likely] .  Computers deterritorialize particular sign machines [another note refers to fuzzy computers dealing with fuzzy sets, 343].  All kinds of social assemblages can be dealt with like this, leading either to mystification or 'a different politics of the capture of the liberating collective consciousness' (104).

Data processing still tends to work best with binary systems which are reductive.  There is still resistance to the idea that the future can be calculated based on the tendencies of the past, or that an economic base will determine events.  Enthusiasts, however, do think they have managed the possible, by considering it in binary terms.  Nevertheless, at the moment, capitalistic modes of thought still dominate such enthusiasm.  A revolutionary machinism would want to deny the split between material and semiotic processes, bring deterritorialization of time and space back under the control of 'a new type of assemblage of enunciation', new types of faciality traits, new relations to bodies, 'sex, the cosmos' (105).  It might even consider 'reverse causalities and inversions of time, as astrophysicists who study the interactions inside black holes currently do' [gullible prat!].  Certainly, our conception of time and space will change, and become more particular, in terms of assemblages and context.  We might pursue more '"Internal" deterritorializations', exploring our own notions of our perceptions of the outside world, or our sexual behaviour, and these will be matched by external forms relating to the environment and society.  The new 'desired face will utilize registers as diverse as those of singular identifying components, power formations with standardised facialities, creative diagrammatic faciality traits...'.

It is impossible to classify concrete machines without looking at the 'framework of particular arrangements specific to each type of assemblage' (106).  These are not systematically organized.  The components do not always have the same importance, and can vary from one situation to another.  Some components are organized as constellations, sometimes with a cyclical mode ['sleeping, waking, meals'—this idiot has done no sociology, and so he can only use banal examples].  Some are organized in central hierarchies, some have dominant facialities or refrains with their accompanying everyday meanings.  Some can interrupt resonances.  Some can 'catalyze a rhizome', by undoing global redundancies.  Thus considering machinic consciousness as a whole, 'nothing is determined in advance', and no relations, such as those between molar and molecular can be predicted.  In particular, eyes can destabilise, or dominate right at the centre of semiotization [with a reference to a weird avant-garde French writer, who seems to like putting some words in capitals—where have I seen that before?, 344].

[Thank God I've got through this chapter.  I know my man, though, and equal horrors await. The opnly hope is that it is running down and there will be some repetition]

Chapter five The Time of Refrains

[Refrains structure time, as when we 'beat time'.  Refrains also structure group activity and play a major part in communication in general, even though music can be considered to be 'asignifying'.  I'll bet there's a capitalist version which is nasty, though]

Individuals and groups share the same a 'basic range of incantatory refrains' (107).  Some languages are atonal.  Some divisions of labour are underpinned by semiotic assemblages, and they seem to exhaust the range, with no 'pure' specialisms.  Capitalism also has an obsession with order and regulation of the heterogeneous.  This simplifies and rationalizes, even automates, writing, song and dance, or more creative elements are left to deteriorate.  In music, or complex rhythm disappears in favour of basic rhythms, and melody and timbre have been simplified. 

Refrains are 'the basic rhythms of temporalization'.  Capitalism imposes the same rhythms, ['it is always the same song'] and this helps organize assemblages serially.  At the same time, specialist options within the overall rhythms are allowed even encouraged to develop as a kind of [bad] diagram.  As a result, refrains can be seen as micro political, operating at the most intimate level of personal time, and our connections to the living world.  They are always associated with a face, as we saw, when faces and voices were examples of 'significative redundancy' (109).  Western music claims to be a universal model, sometimes with permitted flourishes, such as connections with fulk music or other genres.  This followed deterritorialization.  Musical tastes and styles have been imposed.  Some people mistake deterritorialization for a process of abstraction and purity, increased creativity, liberated from the limits of everyday refrains, available to all [but this means 'mass mediated']. As usual, it is deterritorialized material that is individuated, with an obvious role for the mass media: televised programmes take the place of local nursery rhymes and lullabies, general advice is given to the lovelorn in the form of 'antipsychotic catchphrases' (110).  This makes us feel we all belong.

Some people have advocated solitary precapitalist songs and speaking, to escape the network, but this is no longer easy in the 'general regime of inter-subjective mush' (111), where everything is mixed, including 'cosmic flows and investments of desire' in the most banal daily newspapers.  The solitary life may no longer be possible.  This links to the triumph of the signifier in structuralist psychoanalytics.

This could be seen as an aspect of 'The "modernist illusion"', that sees us as having dominated all the old traditional relations with life or thought, through an intermediary world between matter and image.  Kafka describes the situation, particularly in terms of 'capitalistic sound in our relation to time'[it sounds a bit like Huxley's morning for traditional songs in Brave New World] Kafka describes the results as 'an unbearable hissing', and this reminds Guattari of binary music, 'a hissing black hole'.  Western music could be seen as a fugue based on this original empty note.  [Then there is a reference to Schumann which I do not understand]

Modern attempts to transcribe primitive music into western terms simply failed to measure the 'singularity traits', imposing conventional notions of rhythm, for example.  Primitive societies work to different, longer, less consistent rhythms.  We can see some of this 'by returning to the rhythms of our childhood with the incessant breaks' (112).  Institutions like the school and the military, or corporations mean these refrains get 'purified, ascepticized'—a deliberate link between the modernist illusion and public hygiene [mentioned in Foucault I recall]. 

The schizoanalytic perspective does not mean regression, but rather building on those bits of childhood 'that associate refrain redundancies with faciality redundancies' to inform pragmatic fields.  The original territories have been swept away, together with their 'components of conscientialization' (113).  [All sorts of additional links with courtly love or Nazism seem to swarm unbidden into his head at this point -- schizo flow no doubt]. Extreme possibilities are then on offer:
  1. A 'hyper territorialized subjectification'focused on the domestic, involving 'movements of the material couple and their children'[closely regulated domestic activity, a bit like the Kabylia for Bourdieu, or time management systems?]. 
  2. New diagrammatic possibilities leading to new technology to produce 'chronographic enslavement of human functions' [time and motion?]. 
  3. Rhizomatic mutations which open new possibilities to break out of capitalistic forms.

More generally, we can compare the redundancies of refrains with those of faciality as before [with all the gloomy micro politics?], But whereas it is easy [?] to see the central role for faciality , refrains seen more passive, and time less manipulable than space.  However, refrains show the same tensions between radical creativity ['diagrammatic hyperconsciousness' for Guattari] and conformity ['opaque consciousness of resonance'].  We're going to do this in particular when we analyse Proust, but the main discussion here concerns 'the domain of animal ethology'[!]

[This is going to be fun.  I think the way this is going is to say that animals must be able to communicate.  If not, he would have to go back and revise what he said about an animal component of human semiotic activity earlier.  I cannot follow by biology, and wonder if anyone can either because it is so French and obscure.  I rely on Delanda, as ever, to put this into more accessible terms, denying that there is a blueprint either chemical or behavioural, that guides animal behaviour, but more a system of flows of hormones and zones of chemical intensity that get concretized and produce solutions to problems.  I think Guattari is a bit flustered here, judging by the spades of bullshit, and
he provides us with the wasp-orchid cliche in the end.  There also some marvelous asides, including a superb note 13 (345) on the courting behaviour of peacocks.  Apparently, female peacocks are subjectified by the black eye holes on the peacock's tail, that is turned into a female ready to mate.  Some naive behaviourists have denied that mating behaviour leads to orgasm which raises doubt about the communication involved: Guattari will have none of it and argues that peacocks enjoy 'A "courteous" orgasm [which] attaches itself to the partner with a lure and probably releases the hormonal components necessary for the following events'].

It is a 'preoccupation' to deny explanations of  animal behaviour in terms of a hierarchy of instinctive components rooted in nervous centres, as somebody called Tinbergen attempts.  Instead, we can talk of the emergence of certain 'spatio temporal coordinates, social coordinates, etc.' emerging from a typical refrain, 'between stratified systems and diagrammatic processes'(115).  Animal behaviour is not to be explained by combinations of inhibition and release mechanisms, although there are lots of biologists have tried to use these to develop hierarchies of behaviour.  There is even a tendency to deduce 'the existence of a "soul"' in these hierarchies, illustrating the tendency for purely functional analyses to have to rely on 'transcendent authorities' for their ultimate authority.  [Then a superb bit of bullshit lasting half a page, which says, I think that the complexity of animal behaviour cannot be explained just by specific components attached to particular points of a singularity].  Assemblages change as a result of changes in 'stratified substance-form relations' (116), and this is 'rhizomatic creativity', which cannot be grasped by empiricism or structuralism.  It depends on 'a rhizomatic conception of interassemblage pragmatic fields', and these explain innovations in the development of an animal capacities and species.  Here, the term refrain helps us to understand better than the usual oppositions 'between the acquired and the innate', between biological determinism and complete freedom [educationalist have missed a trick here—they could get all Guattarian on the old debate between nature and nurture].  The refrain can be seen as acting like a catalyst or an enzyme which facilitate interactions at the molecular level, in 'spatial and rhythmic' relations: this greater flexibility permits diagrammatic strategies and semiotic bridges, with local links rather than a final end ['no blueprints' is how Delanda puts it].  We can now explore limit states, and breaks of mechanism, diagrammatic potentialities and creative lines of flight 'through which evolution selects its adaptive paths' (117)

He now needs to rethink the point made earlier about black holes as the end of creativity.  Now, black holes can produce redundancies which help deterritorialize.  They can contain 'certain innovative processes', at least in a preliminary way.  These emerge long-term, perhaps after a million years of controlling innovation, or as a result of specific catastrophes.  Animals sometimes encounter 'a behaviour-crossroads' where different options appear to preserve the species, and we see there 'interrogative pauses', or special occasions where refrains, bodily components and facialities interact [and this is where the delirious diversion into peacock mating behaviour occurs through a note].  These are examples of 'the living rhizome' (118). These display 'complex semiotic metabolisms' which lead to certain assemblages, the disaggregation of others, 'sorts of phase transitions'.  None of this happens a priori, except for some underlying desire and life force [almost].  All concrete strategies are found within the overall rhizome, and none can be preferred.

For example, what is the difference between animal and human desire?  It is not just that the latter depend on speech and law, while the former 'remain fixed within systems of ritual fascination and ostentatious expenditure tied down to a passive Imaginary'.  The codifications of animal desire are as rich as those of human desire.  We find the same sort of interruptions by micropolitical black holes, as in courting behaviour, which gets disrupted by episodes of aggression or other sequences.  These are the same as blockages in humans which can also produce compulsive repetitive behaviour,  'behavioural stereotypes' which are widely found in human affairs as in phobias or obsessions. There may be intermediate behaviours between animal desire and human desire.  The most difficult difference to resolve is the way in which those 'interrogative pauses'are dealt with—human beings often simply accept social repression in a resigned way, unlike 'the vitality of animal desire' (119).

Human beings have lost some of these capacities, and have developed excessive individuated enunciation, and this is undoubtedly helped deterritorialization of refrain and faciality, although it's difficult to say which happened first.  Deterritorialization then spreads like 'gangrene'.  We tend to see evolution as creative because it has led to deterritorialization, but we have to resist seeing this as progress or dialectic: it is rare that the contradictions are completely overcome, more common to find them circumvented through displacement of the problems, sometimes even reviving the old territories.  Nor can deterritorialized assemblages be seen as somehow richer or more inventive.  For example, although the 'contemporary technico-scientific phylum' has been very creative, so has the 'living phylum' that preceded it in human affairs.

There are no essential differences between humans and animals, but they do have different assemblages of semiotization, and different components.  For example, human beings have had to invent ways to restore social order, such as faciality and refrains, and this can also produce 'new machinic territorialities'.  But animals also preserve order in a number of diverse ways—scent trails, territorial songs and so on.  It is a mistake to see them as innate, however—mottled Australian sparrows [sic] use their colours as a code to regulate personal space, but it is not that this machinery came first, more that selective pressure resulted in its introduction: again we can't see this as progress.

What about wasps and orchids?  [A favourite example, scattered throughout their work].  Orchids develop colours that look like female wasps, so that male wasps try to make with them and thus transfer pollen.  This seems to be a closed, natural system, but we can see this as originating in accidental encounters and remaining permanent as a result of selective pressure.  There is a role to be played by 'abstract machinisms that have been stratified to some extent in the machinic nuclei of the genome', but we cannot reduce this behaviour to the genetic.  Instead there is a 'pathway' between the genetic, the behavioural, and the ecological, producing 'collective semiotizations' (121).  New pathways are still in play.  The interaction between wasp and orchids therefore shows 'a "surplus value of code"' (122), something that cannot be reduced to interlocking purposes, but which 'functions like a mutant a wasp-orchid species', as a new 'evolutionary line of flight'.  In effect, actualized fields are hardened and stratified to facilitate certain consistencies and encounters, as an example of innovative deterritorialization.  This happens with refrains as well, which get closer to the machinic nuclei the more they are hardened or enclosed—but this makes them even more useful for machinic power.

[Forking hell!  Now we're going to have a chat about baboons!].  We want to look at how the apparently opposed opposites including acquired and innate, individual and social, and economic and cultural need to be examined to separate out fixed elements and 'transversalized operators' (123).  We can see the latter at work in a more evolved animals, such as baboons and vervets.  These animals display 'three social types of assemblages': 
  1. an internal hierarchy, expressed in aggression and combat, often simulated, and favouring dominant males.  This is different from external territorial disputes, and is not tied to a specific territory any way. 
  2. collective defence of the territory, against neighbouring troops.  
  3. individual flight from predators.

So the semiotization here is connected with intrinsic sexual components, but also a social field ['a field of inter-assemblage facility- corporeality' (124)]: baboons are facial because they look each other in the eye when displaying aggression.  For other species, social behaviour comes before sexual display, so it is wrong to assume that sex and the social are always connected in the same way.  In particular, the 'image of sex organs' becomes a form of intimidation only in particular spatially determined areas.  These are not 'partial objects', but only semiotic operators, one of the points that articulate assemblages and action.  [The term 'silhouette' makes several appearances in this discussion, presumably as some stripped down version of faciality? This is actually not as innocent as it seems, because when we come on to discuss birds we have a problem -- birds do not have faces! To talk about silhouettes then would look a bit tactical and opportunist -- better toget it in early].  These connections are mistakenly disconnected in laboratory procedures, isolating particular genetic and cultural components, sometimes inspired by Freudian or structuralist [functionalist?] notions.  Instead, we should be referring to various kinds of choices produced by biological imprint, individual actions, or social relations, so as not to 'crush the rhizome of socio-biological assemblages' (125).  Within this rhizome there are individual components, with biological factors; group components with collective rituals and socialization procedures; species components with mutations and genetic spirals, and even 'techniques of delimitation' [a note refers to the habits of birds in enclosing species], and 'symbolic attachments'[surely the most controversial of all attempts to blur the differences between humans and animals?].

This abstract machinism [rhizome] governs evolution for all animals, operating in all dimensions, including chemical and chromosomal ones.  Apparently, some 'primatologists' at least are thinking along these lines, for example to account for altruistic behaviour [sacrifice of an individual to preserve the chances of a relative—the note here refers to Proust!].  This is an example of how the molecular phylum traverses 'individuals species and milieu'(127) and produces '"molar" casualties'.

[ I have highlighted below the bits that reject all the usual libertarian notions of freedom and becoming as a matter of releasing our subjectivity]

This makes us question the notion of freedom as somehow meaning independence from material things: it is an option within a rhizome, and even nervous or digestive systems display 'a certain kind of freedom and even grace', as we realize if things go wrong.  Seeing the semiotic system involving 'genetic or automated regulation via a harmonious apprenticeship' has advantages: seeing semiotics as entirely a matter of consciousness and freedom leads to problems with explaining 'anguished interrogations or spontaneous blockages'.  'Freedom is not created with subjectivity!'  Machinic freedom does not mean blind automatism, but an ability to focus on 'capacities for life and expression…  what moves, what creates, what changes the world and humanity, in other words, ...individual or collective choices of desire'.  This is better than thinking of an eternal opposition between subjectivity and biological or economic destiny, or, between freedom and the innate.  All these dichotomies arise from power formations and are used to divide assemblages. 

Freedom is a matter of 'the give and take of quanta of deterritorialization emitted by refrains, facialities, etc.  and carried by the ensemble of the components of an assemblage, whether these quantum be material or libidinal, individual or group - oriented, private or public' (128).  It is a matter of negotiating degrees of freedom, especially in 'micro political confrontations', which involve all components, both molar and molecular and those of 'abstract consistency'.  Nothing is determined in advance.  General laws never work at the local regional level, so there is room for 'An intelligence and even a sort of cosmic consciousness' [when it comes to discussing the evolution of a species]. Human societies often do the opposite and set up hierarchies, based on fear, and in this, they resemble 'societies of hymenoptera (production for production's sake, systematic segregation, generalized gulags…)'[a note reminds us that dreams often offer more possibilities of freedom].

Positivists are often obsessively prudent, however, wanting to preserve rigid categories.  This is 'a completely essential, political issue', as when Tinbergen anthropomorphizes and thus projects his notions of hierarchy, and sees animal behaviour as somehow 'good'.  This is an ideology, and so are the opposite views involved in 'transcendental meditation, breathing methods etc.' (129).  At least the ethologists have helped understand some aspects of human semiotic activity [and the note refers to the use of techniques to help us drive cars or play instruments].  The point is to ask whether highly differentiated structures must be associated with constraining and oppressive hierarchies and the loss of personal freedom.

[Back to Australian mottled sparrows!].  Males display and bring gifts of blades of grass, then pretend to be a young sparrow eating the grass.  We can understand this as 'a semiotic index' referring to a chain of behavioural sequences, and a silhouette 'which functions here as an equivalent of faciality'[thought so].  This could be used to understand flirting and welcoming in human societies, which works at the level of very fast mimicry, 'and whose encoding is most probably hereditary' [the note refers to raised eyebrows, or dilated pupils and so on].  Both cases show how semiotic indices emerge, articulated on the human face, or the birds' use of objects as signs.

Birds have no faces, because their heads have not been deterritorialized, but expressive machines do not depend on faces.  They all operate with centralizing specific semiotics at the non verbal level, followed by '"mental" territorialization ' (130) producing particular icons.  What this illustrates is the limits of Freudian interpretations again when it comes to humans, and the need to consider 'uncommon abstract machinisms' (131).  This is one way in which we can illustrate 'an animal "becoming-human" by showing us that signs and tools do not merely belong to our own societies after all', and a corresponding 'human "becoming-animal"', where we realise that many apparently human behaviours really 'relate to an ethological montage'[so this is what becoming amounts to?  Finding analogies which make us suspect common mechanisms, as a kind of transcendental deduction?].

The usual version is that the activities of the sparrow reflect some archaic residue based on nesting behaviour.  Guattari prefers to think about 'concrete machines (machinic indexes or diagrammatic operators) working within machinic assemblages', but not at the level of human intentions or enunciations.  The machines use materials locally and contingently, with no general formula.  This is a universal deterritorialization which does not assume any rational progress.  These local concrete deterritorializations can also produce change that larger levels, again without seeing this as progress.  It is not that all residues have been surpassed.  They retain their semiotic power, and emerge by accident.  This can be seen with the increase in sociability in finches [!].  There can, however, be 'machinic progress'(132), as various potential and undecidable cases emerge.  However such progress is 'political and not normative'[not related to any transcendental progress].  This depends on the creative lines of flight in a rhizome: the solutions can be variously 'elegant' and possess 'grace and beauty', which are not just detectable to humans.

The grass blade is a pretty obvious sign, but there are other components in the rhizome in the form of various nonrepresentative and asignifying 'investments'(133), and these produce a lifestyle.  We are now going to look at birds especially finches.  These will illustrate how the notion of the territory gets semiotized, and how specific refrains affect detailed behaviour—in other words, how the refrain relates to both the social and the individuated.

As the discussion of faciality and refrains shows, 'Matters of expression' are not just passive transmitters of information, but 'actively participate' in  overall process of semiotization, and are equally affected by stratifications or lines of flight.  They are components of abstract machines, and they can be relatively deterritorialized and used in new connections.  They are sometimes held in reserve.  It is a mistake to operate just with empirical connections and correspondences.  We need instead to consider processes like deterritorialization, changes in 'thresholds of viscosity' (138), the effects of black holes, the operation of rhythms and redundancies.  One consequence is that we have to abandon the distinction between form and matter—both are open to these transformations.  Quantification in the normal sense needs to be accompanied with a sense of the intensities of work.  Faciality and refrains work at both levels, within 'form, substance and matter'.  They are located in and 'effectuate particular spaces and times'.

The example of the Australian sparrows indicates that there are different possible relations between material objects and processes of deterritorialization.  In this case, deterritorialization produced a number of heterogeneous components to provide a large vocabulary used by different kinds of birds.  With songbirds, songs are highly territorialized, however, and this has even produced effects on evolution of species, since it has helped to isolate specific populations.  Songs include alarm calls.  Finches are able to mimic the calls of the other birds when alarmed, as a kind of 'territorial frequency jamming'(139), showing how a particular song or call can make up an 'asignifying, behavioural language' [surely a gross exaggeration].  Sparrows can inflect their songs [here rendered as their 'refrains', no doubt profiting from the ambiguity of the term].  All this apparently indicates that songs can become dominant inside the rhizome.

Nevertheless, although this indicates highly ritualized behaviour, variations are still possible [the examples given are peculiar noises made by storks or starlings, or the apparently risky behaviour of the nightingale with its loud song].  Even these anomalies are explicable though, as part of the rhizome.  The mottled sparrows [again],  use grass blades as signs, and also display a simulated return to childhood.  Two other components are also important, highly colored plumage and stereotyped refrain [in the sense of song.  Is Guattari simply deliberately using this word ambiguously?  Does the whole thing depend on linguistic slippage?].  Apparently, the sparrows have to learn this song through apprenticeship and if raised with different birds have different songs—indicating both imprinting and active semiotization.  Further, there is a hormonal component, because only males sing, and females will if you inject them with male hormones!  So there is still a combination of deterministic components.  There is possibly an even closer link with these with deterritorialized activity like singing.  This also explains how deterritorialized human language has produced a tremendous power over behaviour and environment and excellence survival behaviour faced with external threats [I really do not know what to make of this promiscuous analogizing.  It reads like the worst kind of Desmond Morris type popular television].  Overall, a marvelous diagram on page 143 explains the tremendous complexity affecting the canary's reproductive cycle—physical, biological, perceptive components, and closely interconnected activities such as mating and nest building, together with responses to external stimuli and multiple effects.

It is artificial to distinguish between semiotization and the constituent components which are encoded.  Elements may be arranged in various ways, sometimes stratified, sometimes automated, sometimes capable of generating new assemblages.  Sometimes there are 'zones of semiotic collapse, of black holes, which in turn can become generators of over-deterritorialized lines of flight' (144).  The convention of establishing different orders of explanation, the physical and the semiotic, for example ignores transversals, and ignores material effects of expression.  Does social faciality simply reflect 'innate faciality traits' of the kind found in animals?  [Does the Pope shit in the woods?] .  What about the connections between free memories and involuntary memories?  If some aspects of memory, such as short term memory, seems to have a biological base, how does this explain free memories?  We have to say that some of the most deterritorialized aspects like faciality, ideas, or abstract machines are obviously material [in contact with reality].  If not, we are led to empiricist and positivist explanations with unexplained escalations between the levels of chemistry and life, or matter and spirit.  Some components of faciality and refrains can be located within brains, but it is a mistake to assume that these local elements explain molar levels.

'Biological assemblages depend on psychological and social assemblages as much as the latter depend on the former' (145).  It is equally possible to argue that components affecting imagination faciality or music can modify the social field, but also bodies and biological systems.  Ethology is about to adopt this more complex perspective, and 'abandon its childhood illnesses (taxonomism, reflexologism, behaviourism, neo-vitalism, etc.)' (145-6).

So the great question remains—what holds heterogeneous components together in an assemblage?  Not some transcendent hierarchy, not an underlying physical or chemical structure,  but contingent components taking on the role of 'transcoding and deterritorialization (what we have called "components of passage" or "diagrammatic components")' (146).  The discussion of refrains does allude to some notion of how biological rhythms are synchronized.  This has led to mistaken metaphysical arguments, based on foundational 'vital rhythms', sometimes even derived from some basic molecular rhythm.  This is the same sort of reductionist argument that we found with memory and its neurological locations. 

However, it did illuminate the problem about the articulation between the various rhythms and cadences found in life.  There are indeed  'infra-biological molecular rhythms'(147) as an example of machinic types of interaction.  However, it is more difficult with complex compositions.  Components of passages here are not simply effects of transition like this but 'the bearers of diagrammatic keys' hidden beneath empirical consistencies.  They offer connections between 'possible worlds and real worlds'.

Finches illustrate this [not bloody finches again!].  The finches' song operates with 'two types of rhythmic and melodic levels' which together articulate the song into definite stanzas in a particular order.  It is not just a matter of social programming, nor the mysterious vital rhythms—isolated finches can still sing like this and seem to be able to select melodies from a range of those used by other birds.  However, the articulation of the song relies on apprenticeship.  There is a third part 'left to improvisation and competition', developed through competitive singing.  Overall, the diagram indicates 'a constant entanglement involving heredity, apprenticeship, experimentation, and improvisation', and these differentiated forms are the components of passage, and produce deterritorialization and reterritorialization, and connections between territories, individuals, and species.

[Very dubious stuff here I reckon.Good fun to tease the romantics with examples of animal communication, as the behaviourists did, but all seems to depend on a decision to use the terms 'language' and 'communication' to cover these activities. Anthropomorphism seems endemic. Empirical data is somehow now conclusive {finches really do have songs divided into stanzas} even if explanations have to be reformulated. Cases are cherry picked. The most dubious usage is the slippage in the term 'refrain'.

This general problem is also found in DeLanda and his insistence that we ignore the role of human language in constructing social order because human language is somehow 'additional', just with relatively unimportant extra bits like the ability to argue, reflect, criticize and speculate!]

Chapter six Reference Points for a Schizoanalysis

[This is the most sustained  antihumanist argument so far.  There is nothing special about human subjectivity.  As with all kinds of consciousness, it is an emergent property of assemblages, themselves produced by machines.  An infinite number of components are assembled, and they face towards the possible — which seems to be in this case the actualized—and the virtual.  The implication is that other assemblages have subjectivity too, not just human ones.  The unconscious in particular is machinic like this, and it is impossible to regulate it through laws or probabilities.  We need to think of all the factors that produce the impulse to act, for example, {as a note argues}]

The unconscious consists of machinic propositions not semiological or scientific ones, and any attempt to use these risks reductionism.  Existing terms all depend on some basic notion of a pure subject confronting a pure matter, and 'concepts must be folded on to realities, not the other way around'(149).  We have to be careful not to turn distinctions into binaries—this can even affects the work done earlier on faciality.  There are faciality traits that signify, in human affairs, and a mechanical faciality that regulate perception and desire in ethology, at the molecular level.

For that matter, the molar and the molecular must not be seen as a conventional binary opposition between, say, 'large-small, passive-active'.  For example, there can be faciality at the passive molar level [the Freudian imago] but also an active molar form [the example is in Proust, I think it refers to constructing the image of a woman or a conventional partner]. Similarly, at the micro level, and referring to artistic style, there can be important refrains or 'modular stylistic cells' (150) that are imperceptible, even though they control entire works.  At the molar level an entire work can be regularized  [made consistent]  as well.  What is really happening is that the conventional oppositions should be understood as options within 'the same musical or poetic phylum'.

The molar and molecular constantly interact and links can be '"piloted"'either beginning with visible assemblages, or with '"invisible powers"' from matters of expression.  At the same time, all the normal combinations 'can be confronted with antagonistic machinic options'.  It is not therefore a matter of concrete assemblages passively responding to external interactions.  They are produced by abstract machines and the plane of consistency on the one hand, and the 'concrete machinic phylum' on the other, and this cannot be fitted in to macro formalism or micro probabilsm structuring an initial randomness.

This self-management in 'life, thought, and the socius' (151) has produced problems for scientific theorizing.  One approach has been systems analysis, sometimes utilizing mathematical formulae [and the example is Von Bertalanffy].  However, 'the framework of these theories reveals itself to be incapable of preserving the concrete richness of its object, in particular the attachments to micro-social assemblages'.  The analysis only works if we assume that subsystems remain in the same category as the main system: we can then establish hierarchical relations between system and subsystem, based on a notion of complexity.  However, other subsets are always possible [ in this case 'possible' means potential -- I use it in this sense throughout, and use the term 'actualized' to mean that which is clearly possible because it exists], and these can 'remain in waiting and in reserve, and become "functional" only under certain circumstances' [the example is chromosomal systems which can be triggered in particular circumstances].  It is also possible that a particular subset acts as the key to the whole system.

The main point however, is to argue for the preservation of multiplicity and heterogeneity, with possible new emergences at 'new points of metabolic crystallization'.  This will help us operate between notions of forms and notions of chance.  We have to see all populations, of objects subjects or statements, as displaying 'the same machinic optional matter' (152).  This produces both actualized and possible mutant realities.  Assemblages can be dominated by elements that begin with a molecular choice or 'an insignificant line of flight', and these can be completely heterogeneous compared with the structure.

At the same time, we should not see operations taking place on some original 'cosmic pulp'.  Instead, we have to insist on a general rhizome traversing assemblages and strata, while at the same time explaining the emergence of 'active machinic nuclei' producing singularities.  The notions of faciality and refrain hold the two possibilities together, and this indicates that material actual components are always threatened by 'quanta of deterritorialization' which must either be redomesticated, or allowed to lead to new assemblages [ and some of those are functional].

We can put the problem in terms of the relation between 'actualized flows and codes' and 'abstract machinic propositions' (153).  In the second case, metabolism is predominant.  Machinic choice at the nucleus of assemblages introduces an element of creativity.  It is possible to exist without such a nuclei and to be stabilized instead around 'redundancies [of stratification] or black hole effects'.  These provide for 'a molar existential politics' as opposed to a 'molecular existential politics' where there are machinic nuclei in, or close enough to affect, an assemblage.  The first case operates with visible repetitions of fixed coordinates, while the latter operates with 'the machinic plane of consistency of possibles'.  This is the basis of the real difference between the molar and molecular levels of articulation—the metabolism is associated with machinic nuclei, and it can reach a 'threshold of consistency'.  If intensive forces cross the threshold, machinic deterritorialization spreads through the network, and radical possibilities are opened [in the dramatic prose of Guattari, there is 'another universe [which emerges] which traverses all visible universes in time and space', obviously trying to pick up on the glamour of quantum theory].  The new coordinates can be challenged by new possibilities.  However all this can still be considered as material, which is not confined to the actualized alone.  Similarly, general laws and singular cases never stop interacting, and intercrossing between assemblages goes on all the time, between biological, semiotic and machinic components and processes.  There can be no single transcendent law or system of laws to cover or regulate all these possibilities.

All this helps us discuss the problem of how different options emerge from pragmatic transformations.  We can consider this in terms of the molar/molecular relation, and this helps us move the discussion away from human enunciation and human components alone to 'the point of view of the things themselves, so to speak' (154).  This implies that there is some 'subjectivity or a protosubjectivity in living and material assemblages': their components do not just interact passively.  Approaches like Freud that tried to explain unconscious processes in terms of thermodynamic analogies and underlying material energies are mistaken.  Freudian terms like consciousness and inhibition are just figurations inside  'machinic unconscious', and they can play different roles. For example, inhibitions can produce disorder. This process can also operate at the molar level, so that 'molar centring' can be installed on the molecular level.  Similarly, rhizomatic assemblages at the molecular level can produce '"molar plates"'(155), although they might not always be stabilised. 

Consistencies at both levels also relate to each other, and differ only in terms of their ends.  Molar politics puts an end to molecular politics.  But if we begin with the molecular, there is a risk of invading the scientific field itself with '"micro politics" and the "subjective"', or projecting spirit on to matter, this time at the molecular level.  However, human subjectivity works at all stages if we see it in machinic terms.  This is not the normal subjectivity, not the speech that produces the world, nothing transcendental or symbolic, not an archetype, structure or system, but something 'radically atheistic'. 

What does machinic freedom look like?  It is a matter of questions of degree, thresholds and their crossings, some notion of 'discursivity, deliberation and choice' that does not involve the conventional subject, codings and semiotics that do not depend on human language.  We can see examples of such genetic knowledge and machinic consciousness in the 'enslavement of a driver to his car'(156) [gets rather ANT-ish here] .  Signification and desire have nonsemiotic sources.  'A thousand machinic propositions constantly work upon each individual, under and over their speaking heads'.  Faciality and refrains reduce the complexity by short-circuiting connections in rhizomes or recentralizing them around black holes.  But capitalistic subjectivity must preserve a certain element of freedom, even if this is only 'a certain abstract perception of time and space'.  Generally though, the domestication of desire permits the dominance of norms and redundancies.  These processes make it 'absurd' (157) to see unconscious subjectivity as merely a matter of speech and symbols.

All human notions of subjectivity and consciousness 'coexist within biological, economic, and machinic processes'.  There is no eternal subjectivity or super-subject uniting these different operations, nor can we reduce it to a series of 'standard micro subjects, localized in the brain.  There is no pure and universal signifying substance or content.  The dominant 'white, conscious, male adult subject' has emerged from 'the disciplining of intensive multiplicities'.  Only when we have abandoned the notion of the controlling Cogito expressing itself in enunciation, can we see the unconscious as populated by 'molecular packs'.  There is an infinite number of creative assemblages, components, lines of deterritorialization, abstract propositional machinisms.  We need to analyze these in order to understand the unconscious.

Sometimes complex combinations appear to be simple or primary, as with the libido.  Other components generate much more elaborate and conscious assemblages.  Schizoanalysis should deal with all of them, seeing them not in terms of some subjectivity or free will, but as arising from 'objective constraints' (158).  These can be contingent such as a particular hormonal flow affecting a component in a refrain, or other unnatural and astonishing combination.  'Everything is possible on condition that the enacted connections are compatible with a set of machinic propositions'.  These in turn depend on the choices within a machinic phylum, and whether they can cross the threshold to become actualized.

We can reconsider the notion of freedom and the subject to focus on any singularity point.  Subjectivity is no longer homogeneous, but produced by 'infinitely diversified and complexified' nuclei.  Subjects cannot be separated from their machines.  Every material assemblage displays 'a degree of subjectivity' (159) and there is also an element of 'machinic enslavement' in every subjective assemblage.  We have no need to assume some underlying subjective process or life force.  There is nothing exceptional about human subjectivity, and other types of machines, can '"attach" their essence' to flows and codes.  Human worlds participate in the same phyla and planes of consistency.  Molecular subjectivity, not forms emerging from molar levels is what we need to study—'the living, free, creative part of machinic nuclei, and the economy of the possible and its point of abundant growth on the real'.  This is what we mean by the unconscious.

Developmental stage notions of the unconscious are not the same as machinic production—sequences here cannot be determined and are not automatic.  Particular children are parts of an 'individuated organic totality' (160)—as when biology dominates adolescents and releases new 'machinic indexes'and eventually a new abstract machine, which externalizes itself in different areas.  But even here, other components can affect this stage.  We cannot separate 'the interactions of the social and the biological',  with humans any more than with animals [so there is a collective or species dimension for us too?].  The sociological and the biological similarly interact.  We see this with the ways in which the biological stages like infancy and adolescence are domesticated subsequently, although they can also lead to 'collective desiring machines' like youth cultures 'or May '68 etc.'(161), after an initial 'quick, powerless deterritorialization'.

['Ive highlighted some of the remarks on education]

Other examples can be found in the early attempts at writing in a child, and how the 'school machine' reworks the original disconnected machinic indices [to domesticate creativity and polyvocality].  This can be seen as a moment 'tyrannised by an anxiety to conform with the dominant norms'.  What gives the school semiotic assemblage such power to manage the intensities of a child's desire?  It is not a matter of direct repression.  What is required first is an abstract machine that conveys repression, shapes subjects, normalizes particular competences: it is not just a matter of psychological maturity or the management of oedipus.  Rather we need to 'study concrete social constellations and their particular technologies of semiotic subjection' (162), including schools and families [with hints of Foucault and the similarities between schools and barracks].

We can't study abstract machines using the conventional psychological categories or processes.  Particular children develop particular coping strategies, which may involve regression or archaic territorialization.  One option is to develop 'the body without organs', an option for the 'enuretic child' [one unable to control their flow of urine]. [So nothing mysterious or Spinozan about the BWO here—it is a defensive phantasy?  There is a reference to little Hans a bit lower down, and his peepee machine].  Other ways of withdrawing the self involve becoming overdependent on the social circle, including those provided by the educational and therapeutic machines with 'sado-masochistic' (163).  regimes of repression or behaviour shaping.  But none of this can be grasped by conventional psychoanalysis of discourses.  Indeed, 'readaptive procedures'like the ones above can had seemed to offer a better understanding to the child, unless she 'plays the game of repression'.  Even here, abstract machines will still be able to produce new singularities to mediate psychoanalytic procedures.  There is no overall domination of the child's personality, even with secondary symptoms [bed wetters might be unable to do multiple division, but perfectly capable of other logical operations].  What secondary symptoms reveal is the workings of the rhizome as a kind of a 'repressive jouissance', where school repression is somehow connected to mechanisms of faciality preventing masturbation, for example.  These particular connections can then be globalised, territorialized on a particular activity like multiple division, which then also offers a potential line of flight.  Other choices are equally possible.  Educational institutions, in families and schools can only be affective by working on other forms of creativity, leaving out particular territorializations such as stammering and other 'neuroses related to etiquette' (164). [Pretty uncritical view of schools as repressive here? No points of deterritorialization or lines of fligbht in the material offered in schools -- no potential to resist offered by social science or philosophy, or the formal equality of school regimes as in Gintis and Bowles?]

There can be no full repression of individual semiotic activity [there you are then you prat!]   Repression can only take place if desire first crystallizes on particular indexes and points, and if these are then connected to a repressive social activity.  This always takes place accompanied with the continued activity of the 'abstract unconscious machinism' which can proceed to generate alternative assemblages.  Adolescents, for example can see the point of adjusting to domesticated assemblages, and balance the intensity of desire with the dominant form, which is why those forms are 'metastable'.  Not all machinic indexes are actualized, and the virtual ones are potentially threatening and revolutionary.  However, capitalistic abstractions are more durable.  But nothing can be predicted in advance, and 'diagrammatic reactions' or 'machinic flights of desire' can break with dominant significations and reconnect sign machines and social ensembles (165).  In facts, three possibilities arise: abstract unconscious machines can dissociate fully and returned to anarchy; stratifications can be relatively deterritorialized, if they are seen as too rigid;  full stratification can take place 'through the affects of diagramatization'.

Unconscious abstract machines are not confined to particular stages or modalities.  Their activity is not exhausted in objectified structures or representations, including those of subject and objects.  They are not always recuperated by strata, and can destratify.  There is a tendency for them to manifest themselves constantly, including in stratified forms, but the matter that results is not simply passive as in formalism.  Thus 'homeostatic equilibrium' is never guaranteed (166).  Strata are threatened by machines which deterritorialize from the outside, from other strata, and from the inside by lines of flight.

The unconscious is not just a purely logical construct.  It results from the interactions of codings, including chemical and genetic ones.  It uses different matters of expression, some more territorialized than others.  It produces its own plane of consistency to develop its machinic potentialities.  Unconscious deterritorialization 'constitutes the essence of politics' (167) but not just human politics—'a transhuman, transsexual, transcosmic politics'.  Deterritorialization always produces a remainder, residual possibilities, new connections, and 'never stops midway'.  Abstract machines are not a matter of psychological processes, or ideologies or teachings.  They arise 'from a politics of desire "before" objects and subjects have been specified'.  There is no intrinsic potential freedom among humans, nor is it found in the qualities outside the concept in objects [to turn this into a point made against Adorno].  We should not see deterritorialization as a matter of causality, nor is it inherently on the side of open possibility.

We can use this notion to rethink Freudian theory, including the work on the latency period: Freud himself tells us that repression does not extinguish all the infantile memories, even though these become incomprehensible 'for the white, civilized and normal adult' (168).  Infantile creativity and semiotization have been dismissed, and the whole issue of eternal antagonism between Eros and Thanatos really disguises 'the appearance of repressive social assemblages'. This is an interference in 'the child's semiotic politics'.  Why should such creativity be confined?  And how does the 'educational abstract machine' connect to the child's own abstract machine?  How do nurseries continue this process begun by parents [nurseries are particularly important for dividing work and recreation, 168-9].  Schools who demand 'writing detached from any real use' also domesticate the creative diagram, as does their division of space and time and the 'semiotics of discipline' [which include grading and competition].  Here, they link with the conditions of factories or offices and barracks.  The real role of education is not to transmit a culture but to transform 'the child's semiotic coordinates' (169).  So the latency period of infantile amnesia is really a kind of initiation into this later regime, it can last for 15 years, and it enslaves individuals.  This repression of memory, like later cases of amnesia, involves erasing infantile intensity the better to 'reconstruct a childhood according to the norms'. Neurotics can escape 'for one reason or another', but they are then continually subjected to normal values and significations.

Schizoanalysis examines the 'pragmatics of the machinic unconscious' (170). It is not focused on the person or on lived experience as in conventional psychoanalysis.  It is not focused on verbal material and its 'systematic " paradigmatization"'on the basis of some abstract structure.  Instead, groups, individuals, institutions or other social assemblages can become analytic, especially if this also involves 'other "non - human" flows (non - human sexuality, economic flows, material flows, etc.…)'. [Deleuze thinks Marx has got something on non-human sexuality but I have never traced it]

What is required is a matter of the whole unconscious with all its strata, lines and black holes instead of the oedipal triangle, which only ends in transference and 'interminable analysis'.  It is not easy to unblock the possibilities, but the potential of the abstract machine offers a way forward: 'machinic consistency is not totalizing but deterritorializing' (171).  We have to understand the possibilities as rhizomes not trees built around dichotomies [one of the consequences of imposing a linguistic semiology], and be aware of multiple connections, and the combination of stratification and lines of flight.  There is no need to represent the rhizome formally, in mathematical or particularly theoretical terms: there is no absolute representation.  The aim is not to trace empirical states back to the operations of an underlying unconscious which overcodes them.  Instead, we should understand the operation as 'oriented toward an experimentation in touch with the real'.  The point is to construct the unconscious as a series of connected fields or bodies [stratified, cancerous, or empty bodies without organs will have to be unblocked—we are talking about Spinozan substance here?] We will have to understand coding and semiotic activity at the levels of biological, perception, and thought.  And deal with images, categories, gestures and words, political and social fields, writing and the arts.  We will aim at 'systems of tracing capable of being articulated in a map of the unconscious' (172).

Tracings are useful because they show how sign particles are connected and mutate as a kind of experimentation.  Maps are also the result of experimental tracings.  Maps oppose structures because they are more open in all dimensions, not fixed, and adaptable 'to all kinds of assemblies'.  Either individuals or groups can develop a 'pragmatic map' in whatever form suits them—a painting or political action.  The performance can itself modify the map.  The maps can guide 'collective praxis' as with small anti- psychiatric communities, but will have no more general significance for other groups.  Maps can indicate more or less developed pragmatics, however, according to how machinic molecular and deterritorialized they are.  The point is to avoid conventional signification which limits the possibilities.  Any stratification needs to be disrupted, for example through the 'use of signs of a linguistic origin in domains that are aesthetic, scientific, etc.', as a 'deterritorializing line of flight' (173) [endorses delirious ramblings involving literature etc?]   Conventional forms cannot reveal the forces at work, because various segments, molar and molecular, are made to correspond.  Machinic models, however, depict 'diagrammatic processes generating a quantified deterritorialization via systems of signs - particles' [we describe things in unusual ways?]. 

However, there is always a link with existing segments, and 'no universal cartography exists' and there is no general map, no axiomatic.  Abstract machines are always tied to 'an intentional plane'[that is not extensional].  They're always tied to specific times and different realities.  Their deterritorialization is constantly interrupted by reterritorializations, and subsequent deterritorialization.  Nevertheless, they do dismantle 'dominant realities and significations: they constitute the navel, the point of emergence and creationism of the machinic phylum' (174).

However, subject groups and subjugated groups are never totally opposed.  The gap can be exploited by diagrammatic politics, breaking with any kind of fatalism, and seeing the unconscious as a creative force.  Investments of desire have to be localized if they are to take on bureaucracy or reified leadership and so on.  The group's body without organs should be put to new usages and transformations, to 'challenge every status of hegemony' in linguistics, psychoanalysis, the human sciences and so on.

We can trace out the diagram for little Hans [below, page 175], showing how his phobic assemblage developed at different levels.  First we have to understand all the semiotic productions, and how they have appeared in either trees or rhizomes, which in turn depends on the activities of his parents or of Professor Freud, as well as contingencies such as the connection with horses.  We can perceive his libido as 'constrained to find shelter in the semiotization of a becoming-horse, etc.' (176), and this helps us see his symptoms as 'libidinal pragmatics', of escape, to avoid the normal solutions.

Little Hans' rhizome

At the level of molecular revolution, schizoanalysis has a 'watchword'—'"Do it"" (176) we must break with conventional semiotic assemblages and their grammaticality, a refusal of 'legitimating the signifying power' of these assemblages.  Instead the need to 'fabricate a new map of competence and new asignifying diagrammatic coordinates'.  One example is the creation of an avant-garde party with Leninism, to transform the masses away from spontaneism.  The development of Stalinist bureaucracy shows the difficulty of avoiding inversions of maps and tracings. Leninist assemblages claim universality, by being grounded in the First International, 'a new type of deterritorialized working class' (177).

A micro political schizoanalysis should aim to 'abolish the individuated modes of subjectification', by inventing a new form of social pragmatics.  This would operates synchronically by developing connections between different systems of 'signifying generations'[in the sense of generating something], and might operate with any semiotic register, like the symbolic, intellectually signifying or "natural" modes of encoding.  The idea is to operate at the very root of the crystallization of power around a particular dominant 'transformational component' (178), such as a black hole or a particular semiotic branch or line of flight [some extraordinary examples of such components—'despotic signifying writing'in 'Asiatic empires', or 'the systematize signifying delire' in paranoia].  The idea is to see this particular enunciation as just one element in other machinic assemblages [this can be achieved, for example, by taking particular 'writing machines'from their original contexts in poetry, music or maths].  This will oppose both [a reliance on individual?] signifiance and individuation.  Political strategies will be needed to organise new assemblages, diachronically as well, by pursuing a machinic rhizome.  Again, there are no general laws, but sometimes we might find clusters of pragmatic assemblages, or "complexes", for example '" romantic"…  "Popular front"…  Resistance…  And…  Leninist"'complexes', [one of his rhizomatic diagrams on 180] and these can sustain themselves away from their original historical localizations, just as Freudian complexes do.

We need to consider common patterns such as the ways in which molar levels are dominated by territorialized segments.  This looks law-like, but it is confined nevertheless to particular historical periods, and can be disrupted by revolutionary situations: these reveal subterranean machines already at work.  Nevertheless, we can pursue politically useful links with deterritorializing tendencies, developing for example schizophrenic rather than paranoid lines to combat bureaucracy.  Again we need to avoid binaries ['Manichean alternatives'] and remain experimental, pursuing tracings rather than maps.  For example, it might be necessary to territorialize on a new body without organs in order to generate further struggles, or follow a particular line of flight.  Generally, if we want to generate alternatives, we should concern ourselves with 'cancerous and empty bodies without organs', while if we are interested in transformations, it will be more appropriate to work on 'full bodies without organs connected to the machinic plane of consistency' (179).  However, everything begins with tracing diagrammatic effects, and realizing some of them—like writing new dreams, designing new operations for the unconscious: this is quite unlike conventional psychoanalysis which attempts to explain and domesticate diagrammatic effects. New forms of signification like this can become 'a "war machine"'(181), although it needs to constantly guard against domesticating tendencies operating through 'the redundancy of resonance'.  We are not after novelty for its own sake, but interested in increasing 'the consistency of semiotic efficiency' in a particular pragmatic field.

The 'matters of expression' must also be chosen carefully, because they affects deterritorialization and other qualities such as viscosity, and we need to pick those which will help us fabricate a BWO.  Here it is a matter of the right sort of 'systems of intensity' rather than theoretical analyses or the role of affect.  For example, we can deploy a tree if it is generative, and rhizomes if we are interested in transformation.  A generative tree can produce a new rhizome, or 'a microscopic element of the tree' (182) which will help develop a new local competence permitting particular semiotic components to 'blossom'[more horticultural analogies!].  Their particular intensity can develop a new 'hallucinatory' perception, a new mutation, even 'a synsthesia', and these will force a reassessment 'in one stroke' of 'the hegemony of the signifier'[optimistic rubbish, assuming we're all philosophers]. 

So generative trees are useful if we open them up, even those criticized earlier, as in Chomsky.  We can use them to generate formative statements ['a promise or an order', or something that change is 'the bearing' of the situation, like an oath], which are still fully grammatical, but can have transformative power and become micropolitical.  Again, we must attend closely to pragmatic components, unlike structural linguists.  These pragmatic components and assemblages of enunciation are crucial, for example agreeing to swear an oath to tell the truth, while reserving the right to one's own truth.  We might use this, Guattari thinks, to reexamine the Moscow trials, or the operations of schools, tribunals, parties or families.

We can also see how the stages [of socialization] actually work, rather than just taking some simple model of increasing 'sublimation'.  Stages can sometimes engage simultaneously as a resource to threaten the system [as when childhood playfulness emerges to combat factory or police discipline?].  There are no law-like connections, but rather 'a coincidence between the maps' and certain conjunctions—and disjunctions.  There are differences in 'the range of a signifying power' over a system, sometimes completely overcoding or dominating, but not always: there are always 'dialects, indeed particular idiolects' (184), and these are subdued by power relations, not always successfully.

So schizoanalysis involves separating out these different components and vehicles, in order to traverse different strata and deterritorialize.  It is not a matter of psychoanalytic transference of its assumption of neutral analysis.  Schizoanalysis is 'implicated by its object even in its fundamental nucleus'.  It must always make micropolitical choices to develop or restrain mutations with a view to considering 'an inter-assemblage transition'. There are no underlying structures to uncover, but exploration and experimentation 'with an unconscious in action'. It will operate both diachronically and synchronically, as above, and interest itself in even [trivial] 'situational potentialities' (185).  In each case, it is necessary to ask why particular black holes or equilibria have appeared, and what might be done about them [all the dangers of 'extended politics' here—do we start with disputes between neighbours?].  Are there any credible 'rhizomatic openings'?  In each case, we must avoid reducing them to conventional theoretical abstractions and attempt instead to 'thoroughly grasp the points of singularity, points of non-sense, and semiological asperities which phenomenologically appear to be the most irreducible'[whatever takes your fancy, then?].

There are different types according to whether we are generating or transforming, although this is a 'slightly artificial'distinction.  We do not always do this in an "extremist" or "savage" way, however, and we have to be more careful than theoretical dogmatists are.  It is always a matter of observing 'compromises, retreats, advances, breaks and revolutions', and this is our task to 'semiotically and machinically assist them' (186).

Generative schizoanalysis involves 'dismantling or unraveling'the existing weak interactions between assemblages, sometimes in the name of avoiding catastrophe.  Then we need to find indexes and lines of flight, not deterritorializing, but rather reterritorializing over 'long durations', until machinic processes can eventually be deployed.  Inevitably, these efforts will be alienated in capitalism, and also threatened by existing 'micro-mega-machinisms' like the media.  Schizoanalytic interpretations might begin by insisting that the apparent univocal signification in these machinisms is only a selective point of view.  The slogan of the schizoanalysis might involve 'passwords', to show the possibility of 'new machinic sense in situations where everything seemed played out in advance' (187).  It will be necessary to oppose [foundational terms in theoretical analyses], conventional causes or geneses.  It would be a matter of showing how conventional symbols have been attached to present realities, and reawakening new possibilities 'congealed in the past'.

Transformational schizoanalysis is different and involves radical modifications of the mechanisms and assemblages to create new ones.  It will be necessary to pursue strong interactions with other assemblages, deterritorialization, and the development of 'mutant abstract machines'.  There might begin with already existing assemblages and inter-assemblage relations, or particular molecular populations, sometimes extracted from old assemblages, or brought together for a particular purpose.  The micropolitics of choice must be stressed in a molecular politics aimed at developing 'new machinic nuclei'.  Specific transformations need to be identified, and how they might change the functions of existing populations or matters of expression.  The aim is to reawaken 'components of passage' and this involves:
  1. making components discernible, through thinking or magnifying [and the examples are Proust or Kafka];
  2. the proliferation of components once detached from stratified positions in assemblages;
  3. the development of diagrams through interactions, say between social economic and artistic matters of expression.  These might produce controlled deterritorialization, aimed in each case at demonstrating new machinic possibles and active nuclei 'beginning from singularity points'. Alternative consistencies can then appear, alternative constructions of reality.  Everything depends on developing components 'traversing petrified stratifications'(189).  The possibility of traversing like this indicates that existing components look stable but in fact possess 'a surplus of deterritorialization'.  We must not see this as simply a matter of semiotic activity, though, since diagrammatic components must cross particular thresholds before they can become real.

In each case, we need to examine both 'existential consistency' and 'semiotic efficiency' of existing links and future transformations.  We need to think big, considering even 'the most fantastic possible', as well as 'the most irreversible materializations': 'everything in between is possible!' Transformation involves making these possibles consistent, by thinking of them as 'the interaction of heterogeneous components'(190), components of a machine.  We need not proceed immediately to an alternative reterritorialization, but understand the possibilities in deterritorialization, and produce transformational trajectories, running through 'different consistencies of the real'.  This would help us consider all those entities which have escaped immediate consciousness, as long as they are machinically consistent [which is equated with theoretical experiential, aesthetic or fantastical consistency].  We do not need to operate with fixed stages or universals, transcendental ideas or structures.  We do not have to start from nothing, nor need we be thwarted by 'the wall of the visible and the actual'.  Instead, we can isolate particular vectors of abstraction and reality, and the more we do this, the more we can support our analysis.  This will ground apparently abstract machinisms which will be detectable in and adequate to 'the body, the spirit, and the socius'.

Schizoanalysis is not a technique or a science, and nor does it require professional training.  Sometimes it is only possible if mutations have already begun, in social fields or machinic ones.  The comments above should not be seen as a set of basic principles, offering a cure.  They reflect personal experiences and a personal trajectory.

Conventional notions of social objects or psychiatric entities should be seen instead as an assemblage, to avoid reductionism.  Any fact-like qualities of these objects should be seen as arising from their machinic territoriality.  In particular, we should always bear in mind three dimensions:
  1. the dimension of components of passage which help crystallise and sustain machinic nuclei;
  2. the dimension of enunciation or semiotization, 'all the means of expression, representation, communication and indeed subjectification or conscientalization which grant them a particular capacity of recognition' (191);
  3. the dimension relative to machinic nuclei and their capacity to detach assemblages from their contexts, and connect them to 'the entire "mechanosphere"' (192).  Everything is located at the connection between 'a position on the objective phylum of concrete machines', and its position on 'a plane of consistency of abstract machines'.  The machinic nuclei integrate these two connections in order to enable abstract machines to manifest themselves, while already materialized machines can connect back with their 'metabolization', and both processes can then be understood and described ['semiotization'].  This is the way in which we can prevent anything living or conscious or imaginary becoming alienated.

We must focus on specific 'assemblage "analyzers"':
  • singularity points, 'contingencies irreducible to serial generations', found in actual history beyond structures, and including 'faciality traits, refrains, corporeality, landscapity, and territory escaping from the systems of dominant redundancy';
  •  the body of reference for enunciation, but as a BWO, not something closed in, a machinic not a dominant territoriality;
  • machinic nuclei which articulate heterogeneous components, turning them into an internal milieu.

Between these three analysers are three types of relation:
  • relations of subjection '(molar alienation)'relating singularity points' to machinic nuclei, and particular material machinisms to systematic modes of generation [historical processes and contexts];
  • molecular enslavement linking machinic nuclei and territorialities [individuals produced by stratified social relations and structures], to be combated by elaborating abstract possibles;
  • relations of desire, which underpin deterritorializing flows and components of passage, and which separate the stages of machinic creativity and incarnation [actualization], and abstract and concrete machinisms and territorialities.  [diagram p.193 -- loads more to come!].

Eight "principles" follow [god help us].  It would be wrong to see schizoanalysis as 'a new cult of the machine' (194), certainly not if we are thinking of capitalistic machines and their 'monstrous development'.  Is this politics still Marxist? -- history is not just driven by productive and economic machines, however, but by all machines, so they can be no single revolution.  We have to confront 'coercive material means and micro political means of disciplining thoughts and affects', whichever point of the compass we turn to [the origin of the pseudy bit about priests and points of the compass in Thousand Plateaus?].  Is repression always necessary in any social organization?  We can at least conceive of other machinic possibilities, but there is no general solution.  We need to establish 'highly differentiated assemblages'.  Classic revolutionary machines have to destroy capitalist exploitation and societies, and also break with all the values associated with 'muscle, the phallus, territorialized power, etc.' (195).  Schizoanalysis is a micropolitical practice, guided by 'a gigantic rhizome of molecular revolutions proliferating from a multitude of mutant becomings: becoming - woman, becoming - child, becoming - elderly, becoming - animal, becoming - plant, becoming - cosmos, becoming - invisible', in order to generate a new sensibility, awareness of existence and 'a new gentleness'.  We can summarise this in the following aphorisms:

1. 'Don't hold back', stick at the limit 'adjacent to the becoming in process' [do not engage in interminable analysis like current psychoanalysis]

2.'When something has happened, this proves that something has happened', unlike the symptomatic readings of nothing happening in psychoanalysis.  Such happenings are not common 'in the assemblages of desire' (196), but we have to understand them, and not in the privileged terms of conventional psychoanalysis.  Dreams have their own meanings.  Unconscious desire is expressed directly, without deception: 'No need here for spokespersons or interpreters'.  Most people have 'an inexhaustible unconscious wealth', but it is often domesticated by family and society.  At least conventional psychoanalysis reveals the unconscious, 'than one understands very well why psychoanalysts are paid so much!'

3. 'The best position for accessing the hiding place of the unconscious does not necessarily consist in remaining seated behind a couch' [no additional comment]

4. 'The unconscious drenches those who approach it', and the options it presents dominates everything else that happens.

5.'Important things never happen where we expect', so components which initiate the change can arise anywhere, and are not usually the same as those which actually effectuate change [so speech gets converted into something somatic or economic].  This means we should appreciate the heterogeneity of 'these sorts of rhizomatic transferences' (197) instead of sticking with set interpretations—these are only open the possibility of control.  Thus there are no necessarily common indicators of schizoanalytic subjectivity.  Particular components remain inactive because they've been dominated by existing interpretations.

6. Transferences can take different forms—'subjective resonance, personological identification'; machinic transferences operating 'below' signifiers and actual persons, featuring asignifying and machinic interactions, and producing new assemblages

7. 'Nothing is ever given' crossed or surpassed, but everything is available to reuse.  However we must beware some of the 'downfalls', as when one black hole conceals another.  Nothing is guaranteed.  It is a matter of assemblage and reassemblage, subject to consistency [so again no universal psychological complexes].

8 'Any principle idea must be held suspect' (198).  Theoretical elaborations are necessary, but must recognize that their subject matter ["the schizoanalytic  assemblage'] is precarious.

Annex: The Molecular Transmission of Signs

[Dear god, this is almost completely unmanageable.  It seems to be a more systematic discussion of the concepts, machinic nuclei. consciential components etc, that have popped up in the earlier examples, although it is hard to know whether the 'applications' actually came before the systematization.  It is Guattari at his worst, attempting to show how the various terms might be defined better, usually in such a way that they can justify his explanations.  Sometimes this is ludicrously systematic, as in endless subtypes of the main types.  Sometimes the definitions are simply circular, as when 'components of passage' are supposed to be effective, but only when assemblages have components of their own ready to receive these components.  On other occasions, 'helpful' diagrams really just illustrate what the words have already stated— subjectivity in black holes spirals down to nothingness, and the diagram shows a spiral inside a semiological triangle Some remarks seemed to be fairly easily translated into normal language: I have not done this systematically, but it seems, for example, that 'diagrammatic sign-particles' are probably the terms in the semiological triangle, as when signifiers get detached from one context and used to expand the possibilities in another, as with metaphors, analogies and tje like.  The general sources of creativity again are not too unfamiliar — signs relate both to specific objects and events, and are also part of sign systems, which can then be further elaborated {he doesn't like that notion as we will see -- OK transferred then} in a semiotic system, and eventually, a subjectification system.  Subjects emerge from particualr assemblages combining semiotic ( of all kinds) components and some give us consciousness ( the 'conscientalized' ones).

Overall, the effect is of a science fiction world again, with systematic taxonomies of aliens, artificial ecologies and the like.  I was also reminded of Roussel, whom Deleuze rather likes, with his detailed descriptions of machines that are perfectly plausible, as long as you accept the absurd and obsessive propositions and initial definitions on which they are based, and the occasional pseudo scientific description to cover or circularity.  Anglo readers might prefer to think of Heath Robinson.

Anyone reading these notes will have to forgive me, and go back to the original if you can handle it.]

Linguists are wrong to think that semiotics systems are necessarily more elaborate and creative than simpler codes.  The genetic code, for example seems capable of generating as much complexity as anything in the literary or scientific phylum.  Everything depends on the actions of assemblages in arranging and coordinating these codes [assemblages here are the more actual and functional 'side' of abstract machines].  Similarly, there is not a restriction of meanings in semiological systems contrasted with 'an amorphous mass of the possible' (199), since the possible is also structured [by rhizomes].  A machinic unconsciousness acknowledges extremely diversified components at all levels, and the 'multiple universes of machinic creativity' (200), but we have to move beyond the customary concepts of form, information and message, and develop new notions of how coding and semiotic systems work [hence the obsessive taxonomy which follows].

Iconic components, which can be anything from signals to refrains can be detached from one assemblage and attached to another.  This attachment might take the form of a virtual possibility, but it's more interesting when it generates 'a real possible'.  This is particularly how consistencies emerge.  The component becomes integrated into the semiological system of the new assemblage, and can take the form of 'a morpheme' of a referent in that system (201).  Further details will be required to understand the operation of this 'messenger entity' [below]  The real actualizations offer the maximum amount of consistency.

Components of encoding arise when iconic components are capable of being attached to a series of assemblages.  This requires that all these assemblages should be able to receive this component in the first place [!].  In this way, components become 'transitive', and feature 'machinic redundancy' (202).  Specific redundancies can be 'incarnated' into discursive chains, various figures and images, or particular systems referring to, say, 'molding, catalysis, field induction'.  Such transitivity assumes an abstract mechanism connecting assemblages [!] This in turn depends on: the internal relations of the component, the information it contains, the degree of variation it will accept, and the threshold of consistency required; the characteristics of the particular series it is being attached to, including their characteristic lines of flight, black holes, channels of transmission and 'the "machinic inertia" of the system'.  Again, assemblages might be virtual as well as actual.

Components of semiotization.  Some assemblages are capable of developing reflexive relations, and here, 'semiotic redundancies' can arise (203) [in the sense that we repeat both the original statement and the reflection upon it, or maybe reflect and then develop wider applications?].  This quality can affect a whole series of assemblages, and a ' subset messenger' can be produced from a whole series—this is the 'component of passage', and it can also belong to a different series: the same component of semiotization can produce a linear succession between the series [which seems to be a shape commonly imposed on the rhizomes occupied by components of passage—because semiotization itself is linear?].  The process is assisted when messengers are rendered as discrete or digitalized [an example of this annoying process of asignification again, which means anything that is not ordinary language?].  We then get [artificial?] 'syntagmatic chains of designation' (204).  This only works if 'intra-component relations' are suitable; if relations between components and components of passage of the kind mentioned above [reflexivity and linearity] develop [!]; if sign systems in the component of passage develop in a relatively autonomous way.  [All of which is simply another way of making the same points]

Components of subjectification appear when components of passage are deterritorialized and differentiated in two types of content and expression.  Components of subjectification appear if: the new morphemes of the referent also generate redundancies; if expressions generate 'asignifying redundancies'[referred to Hjelmslev]; if there are iconic redundancies of contents [equated to the signified in Saussure—redundancies here specifically meaning the ability to be applied to more than one case?].  Various kind of secondary redundancies are also required, and we need to refer to the semiological triangle which this time features symbol/reference/reference at its apexes.  The first one turns on the redundancy of designation, and here, a note (355) refers us to Barthes on denotation, as a first meaning, implying that there are other designations, and again some of them might be virtual.  Things like computerised systems, however do not feature denotation and connotations, because they '"deal" directly with the referent'—still not a direct relation to the 'real world' though, but only to a 'techno - scientific order'.  Apparently, this is an example of differences in 'figures of expression', and diagrammatic semiotics would include all these possibilities.  The second type of secondary redundancy refers to representation, and the third one to signification [both processes are represented by different sides of the semiological triangle].  Finally, there are 'subjective redundancies'which emerge particularly from the last three, and are depicted in the middle of the other redundancies in a diagram on 206.  [This is because we experience ourselves as subjects if we are able to manipulate these particular redundancies, when we creatively use language?]

The development of subjective redundancies depends first on dividing semiotic components along a continuum from' deterritorialized elements of expression and reterritorialized elements of content' (206).  Then we need to develop 'an angle of signifiance' (207), using one of these components on the continuum to engage in representation [maybe].  Some of these originating components used in signifiance might already itself be managed by a [domesticating] signification process—hence the development of 'an objectifying subjectification', where the subject has no meaning except as a signifier.  The angle of signifiance needs to establish a space for itself to operate [literally on the semiological triangle drawn on page 208, below], and this space needs to be somewhere between the standard icons [the apexes] [the apexes and the lines represent limits or threats to signifiance—divinely ordained asignification, for example, or purely nominalist designations].

angle of signifiance

Interactions between the redundancies develops semiological consistency relating to the referent as material reality; representations and concepts as conforming to reality; the same for signs; the same for 'individuated subjects'[I still prefer my version in the last sentence of the second paragraph above].  This consistencies can be diverse and coexist inside a semiological assemblage.  It is perceptions of 'the modes of dominant reality'(208) that help us harmonise them, by modulating particular consistencies.  This happens in every type of society and 'social subset'[the example given is the tyranny of the French ministry of education determining educational activities to the hour].  Signification in particular relates to 'brute Matter…  The reality of the living Soul…  The reality of the signifying Verb…  The reality of the individuated Subject', and these influence explicit codes and also 'models of the social formations of the unconscious' (209).  They also appear in 'institutions - agencies of power and the media, which are like so many operators of a "grammar" of the unconscious'.  [Very deterministic, so we can predict reservations and waffles to ensue -- yup!]

Particular redundancies can be emphasized in particular assemblages, but there is no general hierarchy, and so no univocal production of sense.  Linguistic categories are required, but only at an elementary stage of production.  Grammar is not the only influence on the operation of a language, and there are always more complex productions in social assemblages, because abstract machinisms are always more complex.  Capitalistic speech and semiology tries to 'miniaturize' and systematize language, but even here, there is a 'degree to which social relations and relations of production are complexified' [and must be unless we are to give up politics altogether].

Attempts to shape the production of sense produce subjective'"feelings of signification"'[quoting Thom].  This is because for him there is an '"acute resonance"'between concepts, words and objects in the outside world.  It is necessary to critique this argument, however and to open up possibilities—resonance like this arises from diagrammatic possibilities, including, of course, 'asignifying and asubjective interactions of a metasemiological nature' (210).  The relations of social forces deeply affect representations [they are"faked" for Guattari], but subjectification is never fully fixed, but operates between two limits [pessimism and optimism, of course] : 'a fusional feeling of appropriation', where everything is dominated by domesticating significations; 'a feeling of "cartographic" hyperlucidity' where we can see transversalities, fight off black holes, extract singularities, develop minor languages and liberate our creative memory [defined as Proustian 'involuntary memory'] [rare moments of ecstasy, although the latter intrudes on us all the time—but even so, we have to learn not to repress it but get a 12 volume novel out of it?].

Consciential components produce subjectification, and they face outside to referents and their morphemes, and types of representation, and inside to 'a sort of maelstrom or semiotic black hole'(210) [this must be the source of all the pessimism about the dangers of pursuing lines of flight to extremes, especially through alcohol or drugs].  Black holes arise from an excess of semiological redundancies which then become empty of meaning [a kind of excessive relativism?  A bit like the vertigo of confronting lots of philosophical positions?].  This shows the machine at its most powerful, even though it produces a personal powerlessness.  Nevertheless it has a creative side as well because new 'diagrammatic signs-particles' can be emitted.  The consciential is a counter point to the subjective [it occurs when consciousness and thought exceeds normal social and personal restraint?].

Dangers are presented by redundancies differently.  Redundancies of interaction can exhibit excessive deterritorialization, which is not controlled by relations with the usual more territorialized forms of signification and representation, or the persistence of iconic reference and conventional conceptualizations.  Redundancies of resonance arise when relations themselves produce relativism: this is not easily reattached to more territorialized forms, and can accelerate.  These redundancies acquire 'infinite power' but also 'nullified efficiency' (211).

Consciential components therefore lead to a situation where subjectivity is not grounded on conventional references, and they produce nothingness.  Ironically, this becomes a stage in a further 'process of subjectification which starts to spin around itself', producing a kind of free-floating subjectivity, and an illusion of autonomy from all the normal objects and supports [maybe], 'nuclei of pure self destruction' (212).  It seems impossible to escape from 'absolute deterritorialization', and this pure subjectivity goes on to infect all the other semiotic productions.  However, it is 'infinitely porous to sign - particles', emanating from machines, offering 'an "escape from [conventional thinking and] language"'.

Two kinds of micro politics are then possible, with their intermediaries found in concrete assemblages.  In the first one, we can construct an entire 'world of simulacra', where all redundancies resonate together [and confirm each other].  In the second case, machinic inputs lead to absolute deterritorialization and then a more relative version which we can use to find new possibilities in existing strata, codes and assemblages, to produce a more critical redundancy of interaction.  [A nice diagram with the semiological triangle containing a spiral appears on 213.  It helps us translate conventional psychoanalytic terms in to Guattari's own language.  Similarly, there are some pathological reterritorializations of subjectification illustrated in another diagram on 214—hysteria, paranoia, phobias and schizos].  Psychological pathologies arise when resonance centres on a particular node which then controls all the other components.

Diagrammatic components.  We have seen that the subjective black hole can produce two outcomes, and this is openly described as 'a new type of binary machinism' (215).  One set of operations lead to the world of simulacra at the molar level, while the other set 'diffuses a dust of diagrammatic signs - particles'[now this really is poetic!] at the molecular level, and this encourages machinic potentialities.  Signs-particles are not really semiotic entities any more, but components of referents and abstract machines, seeming to stand apart from semiological systems, but 'infinitely deterritorialized and deterritorializing' (216).  This makes them able to colonize[develop a foreign policy in G's terms]  in an 'almost unlimited semiotization', unlike icons or conventional indexes.  As a result, they are able to destratifiy and deform, and even desubjectify and desemiologize.  This is because they display autonomous figures of expression, produced by diagrams not existing representations.  They act like the artificial languages of logic or mathematics (or avant-garde music), not so much offering morphemes of conventional referents, but producing new ones.

This sort of diagrammatic operation breaks with the usual connections between formal substance and matter.  The conventional connections between signifier and signified are 'disaggregated', form and matter are linked in a new way, formalization itself is 'miniaturized 'and accelerated [an obscure reference to Virillo on the new speed of politics, assisted by miniaturized 'technical components of data processing', 357, and an equally obscure diagram on 217 showing some sort of short circuit between matter and form leaving out substance].

Different movements of deterritorialization can be articulated in different ways.  The first option affects 'semiological sign machines', which switch from conventional figures of expression to those generated by the abstract machine.  The second one affects 'mental icons' of relation, [possibly] by reconnecting them to 'asignifying diagrammatic figures' [what would these be? Levi-Straussian triangles?]  The third one affect morphemes of the referent as we saw, and can generate new realities, at the levels of chemistry and biology as well as social.  These are three vectors, and they interact or conjoin in a cycle so it becomes impossible to separate them, and they work to undermine conventional signification by emptying the semiological triangle [that is, each of these three deterritorializes one of the apexes of the triangle, as in the diagram on 219, and this can be cumulative].  Apparently all this is predicted by Hjelmslev's work on expression and content and their forms and how they interact—but for Guattari there is more interaction, and it is not mechanical, but produced by a 'micro politics which could be termed ontological' (218).  In particular, sense is not dependent just on form, but on 'a universe of abstract machines beyond all formalism'.  This explains how sense can emerge outside the normal forms of encoding and syntax [which is the point of looking at non - sense in Logic of Sense]: there is no amorphous mass outside awaiting the imposition of forms, but 'machinic sense', manifested as a 'spatial, temporal, substantial, multi dimensional, and deictic [meanings rely on context]  rhizome'. This represents all the possibilities, not just transformations and resonances, as signs-particles interact as 'pure potentiality' (219).

Those signs-particles have small amounts of actual consistency, but large amounts of potential consistency, and this helps them link the abstract machines to the more concrete figures, and morphemes to referents.  They are abstract.  They miniaturize semiotic vectors.  They are intermediate between absolutely abstract machines, subsets of those machines which appear in redundant systems, but which are not confined to them, and 'concrete machinic subsets' appearing in actual figures of expression, icons, faciality, or refrain and all the other things which guide 'territorialized redundancies' (220).  Those figures are also constrained by the redundancies.

We can now see that the consistency of an assemblage will vary, from the inconsistent, where concrete elements are emptied out by other assemblages, or where mechanisms decay or become abstract or reified, and this leads to a 'black hole of resonance' or just a disintegration.  Consistent assemblages emerge from abstract machines which can produce new connections and complexities providing an energetic'"machinic nucleus"', producing new singularities and thereby demonstrating new possibles.

With the latter, there are three stages [sigh] .  In the first one machinic redundancies are differentiated, and components specialized.  In the second phase, preceeding systems are neutralized, partly by increasing resonances through subjectification and conscientialization [just like Durkheim—an increase in moral density leads to more individualism and profane thinking]. In the third one, deterritorialized machinisms proliferate and miniaturize [computers give us more power to think of alternatives].  There is no dialectical path or career linking the stages.

The subject is also affected.  It is necessary to remind ourselves that subjectification is not the same as conscientialization.  We can develop a new subjectivity independent of conventional notions of consciousness as in Freud, and there is also a new one 'independent of individuated subjectivity' which is particularly important for schizoanalysis (221).  Machinic consciousness could become a component in new assemblages of enunciation, no longer confined to just human subjectivity, maybe even appearing in completely automated systems.  This is a new notion of the unconscious, not just something interior affecting subjectivity, but 'a hyper conscious diagrammatic unconscious no longer maintaining anything but a distant relationship with the significations of dominant semiologies' (222).

We can now recapitulate the discussion of components of passage.  In general, components are not the same as flows and strata because they belong to assemblages.  Components of passage actually help constitute the assemblage, because they belong to its machinic nucleus or can affect it.  Components of passage traverse the whole set of other components and guarantee machinic consistency, or manage the thresholds of 'sufficient consistency'. In this way, purely imaginary possibilities can become mathematical possibles, and theoretical possibles technical ones.  What happens can be explained as an increase in consistency [changes of state, thickness, visibility, actualization and so on], without entropy: there is no general principle of decline with these machines.  Consistency cannot easily be quantified, but must be seen as arising 'through assemblages of proliferating and fuzzy subsets'(223) [in other words he doesn't know].

Machinic possibles are turned into inputs for other assemblages in various ways or modalities.  There can be 'a catastrophic mode' of disintegration; a 'black hole mode' of relative deterritorialization; a 'quantum mode of deterritorialization', where forms, structures and systems are transmitted to other assemblages 'through molding, catalysis, crystals of code, informatic sequences, diagrammatic processes'; a general association of these modalities.  The process can be supported by 'matters of expression deterritorialized to differing degrees': signs-particles which are fully abstract and can escape all conventional coordinates; 'concrete machinic propositions' which traverse substances of expression; 'material flows and strata' which are also expressions with 'intrinsic codings'; the 'evental possible' which is singular enough to escape coordinates and which can attach themselves to coordinating assemblages: luckily 'an essential affinity [weasel] exists between the possible of the most abstract machines and those of the most singular points', because they are both on a continuum of an absolute impossible, in 'the sort of seat of radical creationism' (224), and they can interact, as when singulars generate laws, and 'the general is singularized as concrete manifestation'[all this is a particular example of how philosophers go back over what they have written in order to try and make it all consistent, more or less at the expense of adding characteristics to different objects so they can relate together].  Machinic nuclei can also support the process of input because they commonly provide an assemblage with options and the opportunity for 'the politics of subjectified choices', including those which enable thresholds to be crossed.  Our job in exploring the machinic unconscious is to examine how these components of passage are articulated by modalities and supports.

In assemblages of semiotization, there are molar structures and transformational machines which can bring 'into play asignifying diagrammatic semiotics'.  Assemblages of enunciation territorialize, however, usually by 'fixing them on an individuated, conscious, deliberate subject' and this helps stave off black hole effects.  However, are there are other matters of expression which are asignifying [not just computers this time, since objects talk to us and the like -- cakes or pink hawthorns talked to Proust, geological strata in TP?]. Assemblages of content do the opposite, limiting diagrammatic processes and fixing 'paradigmatic frameworks' (225), thereby tidying signified matters, linking to power formations and achieving social homogeneity. 

These two processes [relating to machinic possibilities and assemblages of semiotization] can also be schematized: they can produce substantial ensembles of content or expression; particular forms of content and expression, and matters of expression, which help extract forms and substances [unhelpful diagram on 225].  Further, signifying components of passage produce redundancies between matters of expression [another reference to Hjelmslev], but passing through a 'loop of significance' which can then generate strata or black holes.  Asignifying components do not move through signifiance but offer formal kinds of extraction, limited only by a plane of consistency [I am not sure what is being argued here, as is often the case.  Is it that the asignifying components are somehow more objective, less open to creative signifiance and thus capture by power, somehow directly intuited as machinic?]. Asignifying figures of expression 'work directly... with abstract and concrete machinisms and singularity points' (226) [the examples given are 'phonemes, graphemes, mathemes and informatemes', so back to artificial languages again, presumably tightly regulated by definitions and logical rules.  The implication is that this makes them less open to capture by power relations? If so, this is highly debatable to put it mildly]

Capitalist subjectification is a particular assemblage which features deterritorialized flows,strata and assemblages which in turn generates 'a global super-machinism' which then integrates every human activity.  Artificial objects are used to reterritorialize, create modes of subjectification which are then 'controlled by the dominant powers'.  As a result, components of passage take the role of integration, say between subjects and labour power, or the integration of the unconscious in an acceptable way, using power apparatuses of all kinds, including the media.  Components of passage also link individual and collective economies of desire, restricting escape through lines of flight, providing 'indexes of every nature'.

However, [and here comes the customary resistance option], 'In fact, on a real terrain' [as if he would know!] (227), this system does not operate directly and the same components can also produce possible deterritorialization, or stratify apparently common territories.  For example, capitalist signification territorializes both on 'neo - archaisms' and [very modern] specific asignifying elements [commodities as well as computers in this case?].  We can even detect 'components of "molecular revolution" or schizoanalysis', in the same processes' of deterritorialization and 'artificialization', and as usual, diagrammatic representation exceeds the conventional divisions between realities and territories [as anyone can see?] .  Further, capitalist politics attempts to reterritorialize in order to manage 'machinic turmoil', but an excess of deterritorialization threatens this process: we must think of a way to found human and social life on movement not on structures [sounds like a Trotskyite refrain to me].  The free radio movement of the early 1980s might represent such as social tendency, because it is 'a machinic mutation, a technical miniaturization'[how literal!], and this can help release diagrammatic processes by appropriating various bits of redundancy image and exploiting 'optimal conditions'[technological advance and permissive legislation, presumably].  This might allow the creation of 'new realities and new modes of subjectification and sociality'[so why didn't it happen?]

[I have only quickly skimmed the essays on Proust with which this massive volume of work concludes.  I have never really liked Proust, and I am trying to work my way through In Search of Lost Time  again { I never got past the first 20 pages before}, allowing myself an awful lot of skimming.  I am summarizing it, as an Monty Python fan would.  It is now June 2014 -- I might return and make more comments if I get through it. I have now --  July-August 2014.  In the process, I have become aware of a real problem in commenting on the text,apart from its massive length and detail: it is hard to keep separate in your mind the Narrator and Proust himself. I think other commentators have the same problem and I am not sure even Guattari manages it totally. Barthes says that
Proust never simply puts his life into words but rethinks and retextualises his life in order to make it into 'a work for which his own book was the model' (1977: 144). Some people say there are obviously biographical bits, and the text is realist and has the Narrator as a character and an off-page narrator too. Nevertheless, I have doubtless screwed up sometimes in saying that Proust says or does something, whereas really it is the Narrator who does: is Proust himself limited in his becoming-woman by his respectable and patriarchal stance towards Albertine, or is the Narrator who is, while Proust cleverly shows us the diagrammatic possibilities? Is Proust responsible for the distanciation and the repetitive pot-boiling bits or is he showing us how the Narrator's memory works? ]

Proust and the other novelists D&G admire {see the collection in the 'clinical' project} are good at producing 'hyper-deterritorialized mental objects' (231) , and it is useful for philosophy -- and science-- to study them. This would provide that creative '"perceptive overlapping"' that is crucial to the business of creating diagrams. I suspect you need an awful lot of cultural capital to do this successfully, of course,and there is always a danger of actually subordinating literature to philosophy. To pursue the latter -- Guattari isolates and illustrates themes and topics, but what of the 'surplus' that is omitted, such as the tremendous botanical detail or the obsession with place names, aristocratic lineages or trains? -- is this just a flow of redundancies? Guattari does return to the text as a kind of empirical confirmation of his readings.

It is quite possible to see what Guattari getting at, nonetheless, and the essays are really very clever analysis [much better than the usual drivel about how novelists' characters are unique but also so universally applicable, or how landscapes affect writers etc] of the structure of the great rambling Proustfest— which is a rhizome, or map of one, of course.  The 'little phrase' in the sonata composed by Vinteuil is obviously one of those refrains, and Swann's obsession with the face of Odette is clearly a matter of  faciality.  In fact, there is a lot of stuff on both, and how Odette uses the little phrase to ensnare and bind Swann, although Guattari says it does a lot more as well, helping the main characters to see the semiotic power of music (or in the case of de Charlus, a gay musician) . [The musical phrase does most of the semiotic and philosophical work for Guattari -- but then he is very keen on music himself. Music offers access to the diagram, but there are obstacles to seeing it or hearing it that way, including the way in which Morel abridges the music and renders it as Swann and Odette's own song, aided by Odette, so that we get conventional readings -and the rest is just mysterious and threatening. One potential is to show how to become woman, but Swann gets hetero, and later, deCharlus and Morel get bitchily homo: only Proust sees it as becoming-woman. Music is so threatening that Swann and many of the rest of us have to rely on faciality instead]

Faces and how they transform are frequent metaphors of relationships -- Odette's face is transformed by Swann in love and then back again as he falls out of love; the general's face is scarred and his features become an 'asignifying system' of their own,  de Charlus has a fierce face etc. I still think all this can be handled by considering metaphors rather than new concepts of faciality -- Guattari admits that faciality is of importance mostly to Swann, a visual artist, while the refrain covers all the cases.  The components of faciality become components of passage as they traverse different assemblages, Guattari's way of noting how females with their mysterious looks and glances appear in several contexts as bridges -- eg Gilberte attracts the Narrator by a mysterious look which he misunderstands; they meet near the place where he had spied on Mlle Venteuil at Combray .

The original obsession with Mama kissing the lad goodnight threatens to become a black hole. He copes by deterritorializing her face, partly by writing about it. We can see how a black hole develops around another face as Swann loses all his normal bearings as he falls in love, develops an all-consuming passion (really with his image of Odette's face, glamourized by having a portrait imposed on it). He has to find a way out of this 'semiotic collapse' emanating from his relationship with an unsuitable courtesan --he can't just screw and leave her because he is in love, he shouldn't really marry her because that would be unthinkable socially, and anyway he is not in love after all -- so a kind of marriage of convenience is the way out

We can also see 'landscapity' as a major theme -- so major that I wonder if reading Proust suggested these ideas in the first place: the terms in D&G are not poetry but hommage? The recovery of memory is associated with walks in the countryside ('ways') and the landscape inspires and organizes coherent memories, and delivers overall 'ecstasy' -- 'pragmatic fields' for Guattari (323), [but equally explicable as an organized bourgeois romantic gaze.]

The attempts to understand the impact of these events on the thinking of the characters are analyzed as various modes of semiotization, conducted in particular assemblages. Swann, fopr example, is not easily reduced to Oedipus, since his attempts to make sense of his relationship radiate in all directions, meet dead ends, deepen etc more in the manner of schizophrenia.   The narrative tension of the novel (such as it is, it is far too long to have any as far as I'm concerned)  is explained in terms of these assemblages gradually clarifying the meanings of things like the little phrase, becoming less fuzzy, as the characters come to some sort of self awareness, and make connections between events which eluded them before. It is also a break with conventional subjectification.  Various clues in the novel can then be understood, including some mysteries in the conduct of some of the characters.  In the case of Swann, discovery of his own ambivalent feelings lurking beneath his conventional sexuality for a man of his station (eg being predatory towards working class girls) and his apparently inconsistent tastes is  traced to some homosexual undertones, but opportunities are not taken up, and Swann's search for new semiotization and creativity (like the Narrator's and Prousts) is rendered as a process of 'becoming - woman'. This is one path in the rhizome, pursed as others atrophy {eg becoming homosexual}.

Proust also 'becomes-women' as he realizes the marvellous sensuous wholeness and reality of a group of young girls at Balbec (the 'little band') their philosophical sensuousness, as it were,not just sexual. In order to achive these insights, though, actual women have to be rejected and they usually die, [ as a kind of anticipation of Mulvey II. In fact the double standard is much in evidence, especially when discussing the difference between male and female sexual 'inversion'. The narrator is terribly jealous and dog-in-the mangerish about Albertine and her proclivities, to the extent of wanting to marry her despite not loving her, in order to keep her from lesbianism]

Meanwhile, Proust's own attempts to reconcile all these possibilities as a writer leads him to conceive of actual events as 'diagrammatic' (the many meanings and semiotic uses of the little phrase, including its original association with Odette, and then its capacity to re-energize Swann in his search for creativity) and as resulting from a multiplicity (the transformations of the faces, including those from actual to idealized objects and back again). Guattari quotes some actual bits of Proust which use terms similar to multiplicity, for example (286). Deterritorialization of music and faces etc  is also implied, of course. At the same time, the novel has to develop planes of consistency. So --Proust manages to deterritorialize his childhood memories to release their creative power and realize their effect on his later life; Swann manages to break out of specific hangups and passions and discover music as a means of expression; Venteuil  also produces a mighty Septet as well as his neat pretty Sonata and this explodes into joy (steady, Dave).

[However, I also think that meanings are stabilized in much more conventional ways too. The romantic gaze towards the countryside  and nature contexts,   dignifies and regulates, makes respectable the more adventurous and risky sexual semiotization and aestheticises  the more boring and challenging encounters in salons. Trains, lineages, place names, and the endless bitchy processes of social distanciation, like an early and horribly familiar version of Bourdieu on distinction,  remain constants which ground the speculative bits. The reader is assumed to be able to grasp the underlying issues that make some utterances, preferences or dispositions, gaffes. There are bits of realist commentary where an omniscient narrator explains motives or the meanings of events. There is a great deal of repetition and pot boiling, including tedious lists {not even trees!}  of place-name derivations or aristo genealogy -- pages of them -- possibly because the full 12 volumes are blown up from an early 3 volume effort: sometimes it reads as if it were written as a part-work.  There are appeals to the reader to keep going, and even farcical episodes]

Proust is able to semiotize much more productively, and he can manage his characters -- apparently, he makes Albertine narratively manageable [yes -- as above] as he falls out of love with her, by reducing her more and more to a mere appendage, and then she disappears altogether (becoming-imperceptible). [I think this fits Swann even better -- after two volumes based on him, he has only walkon parts in the others, and just dies without much further comment]. She was already 'several people' [like D&G in the intro to TP] ( 316) He wants us to realize that his early childhood memories are real, exploiting the asignifying properties of matters of expression (no doubt) of flowers, streams, buildings etc, and defends the detail in terms of showing us how it is all plausible and consistent{!}  (304) . The bastard is doing ontology! It is also a matter of developing ways to recall and semotize all the trivial incidents of his childhood,in a kind of therapy, reconsidering the near observation of lesbian sex by Mlle Venteuil at Montjouvain, which made him anxious,  as a part of a creative project to recall the past and understand it machinically . It is partly the rhizome itself driving things forward though, and the qualities of some of the matters of expression themselves, especially the musical refrains [we nearly permitted subjectivity there. Guattari says Proust himself deconstructs his fictional subjects showing them overwhelmed by bits of signs, faces, musical refrains etc., and comes to realize that individuated subjectivity is a pimple on the face of trans-subjectivity, 326].
We are talking about far more than subjective impressions, and the reality of these events is upheld by making them reappear and intervene in new assemblages  [the same reason D&G develop all those obsessive classifications, subtypes and taxonomies?].

The individual author is only a catalyst for an assemblage (305) .  Having deterritorialized faciality and refrain components, and amassed a whole collection of fragments, the author has to construct a new plane of consistency or risk a black hole from excess. Characters are seen as puppets controlled by time [machinic faciality for G], as singularities composed of non-individual material forces, in a new non-personal subjectivity. This is machinic time not empirical time, and the latter has to be managed in a novel [by seeing it as extension produced by intensive machinic time?] This makes the personal effects and the risks of a black hole likely when we root around in subjective recall, less dangerous because everything is more abstract and under control?. Certainly, disturbing events are now made 'common' and banal. This abstraction involves more than metaphor or metonym, Guattari insists , and resembles 'scientifiic work' ( 330) . It also shows the power of becoming-woman as a literary and philosophical process, seeing all the trivial aspects of ordinary life as 'quanta of the possible', whether at the social or psychological level.

Well, after all that -- back to Deleuze page