Deleuze for the Desperate #13c: Sign regimes

Dave Harris

Both Deleuze and Guattari have discussed matters of language and linguistics in their own solo pieces -- eg Deleuze (2006) or Guattari's articles in Genosko (1996).  Guattari has a detailed discussion in his Machinic Unconscious They also have a substantial section in their Anti-Oedipus. Most of what follows will be based on Plateau 4 however.

With the other issues, it's been convenient to look at some examples first before zeroing in on the concept itself, but I must say I found the examples rather obscure here, things like the history of psychoanalysis or Jewish history. Instead, I found the summaries, lists and diagrams, were more helpful, and I refer to those as we go through.

M Harris

The argument here begins with the notion of coding. Linguists seem to vary about what a linguistic code actually does, but the interest here is on the ways in which signs are actually attached to objects events or people, how things are given particular names. Coding for structural linguists is not particularly interesting, and seems to be a matter of simply applying culturally acceptable names that we have all learned. This is part of a general disinterest in content in favour of examining the ways in which signs relate to each other. However, Deleuze and Guattari are more interested in the political implications. There are obvious cases where choosing names involve politics. Are we to call the Palestinian activists terrorists or freedom fighters, for example? Shall we describe the main influences on policies as market forces,national interests or class politics?

When we talk about sign regimes, we are talking about combinations of linguistic and political features, political in the broadest sense. There are several main types of sign regimes, four in this case. Deleuze and Guattari are keen to assure us that these are to be seen as maps rather than specific descriptions of historical stages. They have several other maps which are similar in intent, such as regimes of subjectification (Guattari 2013)  or of capitalist regimes (in Genosko 1996). These maps overlap and display different focuses.

end M Harris

The first sign regime is called pre-signifying (130). We have fairly simple connections between signs and events, almost 'natural' forms of coding they tell us. Codes are found not just in speech but in gesture and bodily expressions, in dances or ritual. In Anti-Oedipus the system is described as undercoding. There is no tight control over coding, and instead we have a more polyvocal system. This regime depended on there being tight local territories, and as social change weakens those territories, so pre-signifying sign regimes tended to be replaced. Deleuze and Guattari argue that this replacement tended to be total and complete, and this is in contrast to modern sign regimes.

Second, we find a countersignifying regime found among nomads and existing in war machines. We have already discussed the war machine in a video. The implication here is that there is a refusal to operate with fully positivistic or quantitative systems and definitions, but to prefer intensive measures -- of nearness rather than measured distance, for example. The implication is that we also find loose political coalitions and local leaders rather than a bureaucratic modern state. As we saw with the war machine, this regime is never fully conquered by the state regime, not least because the modern State needs it to generate new ideas and practices.

The third regime is developed in early despotic political systems. It could be called the signifying regime (of the sign), and there are eight characteristics of it listed on p.129. There is also  a useful diagram on page 151. To summarize, language and coding is controlled very tightly by a single central authority, often a human being like king or chief, sometimes acting as a god. There are some implications for faciality here. Approved codes are extended to different parallel circles of society that all radiate out from the centre. There is a constant effort to interpret new events in the light of the central codes, and a specialist body of priests to do this. This is also known as overcoding. We start to see important implications for the construction of subjectivity as well. Language offers a form of strong signifiance, language designed specifically to construct the ideal subject, someone totally obedient to the central authority, referring to themselves only in identities approved by the dominant language, obeying what God or the King wills. There is a clear link with the paranoid personality type. Dissent is still possible, because even overcoded language can never be fully controlled. Flows or packets of signs can be detached from central systems, sometimes after the work of a charismatic prophet, and can become associated with deviant groups. This involves a break with dominant forms of faciality, a double turning away: God turns away his face so his human subjects reciprocate. Usually that meets with punishment in the form of death or banishment. Dissidents are given the choice of obeying central authority or going into exile as scapegoats, choosing between the face of God or the anus of the goat as they put it. They seem to have the history of the Jewish people in mind here and spend several pages developing their particular take on it.

The fourth one describes modern society -- the post-signifying regime. There is no central control over coding, but rather a number of linguistic flows and codes. Individuals in these societies see themselves as freely acting persons with no obligations to God or to any despotic social arrangements. Instead they can pursue their own passions. This term is used after a quick survey of French psychoanalysis which identified a passional personality, a rational form of delusion. The subtypes of that delusion included monomania and erotomania, which made me think of the mad passions pursued by young gentlemen in Proust, amour fou as the French call it. Here, any external object, event or person can act as 'a point of subjectification'(141) and develop into subjective activity. However such activity will still be limited and takes the form of segmented lines that stop and then start again. These lines are further described in Plateau 8 on the novella. They want to develop rhizomatic forms of subjective activity instead.

The post signifying regime looks more liberating because there are no central dominant codings and no sustained despotic signifiance, but there is a downside in endless relativism. Individuals do not lack signs with which to describe themselves adequately, despite Lacan, but rather suffer from an excess of possibilities. Sociologists like Durkheim would call this state anomie (the whole scheme looks rather like Durkheim). Individuals in post-signifying regimes are endlessly restless, have endlessly segmented lives, with no possibilities ever definitively excluded. Deleuze and Guattari refer to the importance of guilt and betrayal, in a literary reference to Borges which I have not followed (see p.138 and note 20)  Deleuze (1990) refers to the notion of a society of control, with an awful combination of the freedom to develop however you want, but everything requiring access to information, the right sort of legitimation and entitlement. You find your subjective identity only as a category within more general types -- not an individual but a  'dividual'.

The notion of the endlessly restless consumer, or the human individual awash in a sea of overlapping flows of signs seems to offer some promise in explaining some of the pathologies of current social media. There is an article by Genosko (2008)  that explores some of those possibilities. We can at least see that post-signifying systems imply a deep dependence on current forms of communication, and that when we operate with them we are always open to the actions of others, we experience eternal debt is how it is put. We might add that we are particularly open to externally-provided flows of signs.

So we have four main types of sign regime, but, with the exception of the first one, it is common to find them mixed. A particular mixture might encourage the development of individuals extending their lines of subjectivity, only to have a priest step in now and then and insist on interpreting what they do in terms of dominant codings. The examples in this plateau include modern psychoanalysts as priests, and again there is a reference to a slogan by Lacan on page 138. Patients freely develop their subjective thoughts, only to have the psychoanalyst intervene and insist that they are really displaying classic symptoms of neurosis or psychosis. I must say I think that the interventions of any experts in social fields can be criticized like this. University academics are also good examples. To paraphrase Goffman (1968), educational experts are people who identify problems with your thoughts that you never even knew you had. Bourdieu (2000) uses the term 'symbolic violence' to describe the ways in which researchers in social science treat the views of their respondents as mere 'data'.

Let's trace some implications for the human subject. The paradox of the subject is well developed by French Marxists like Althusser (1977) , who is cited in this Plateau p. 144). In order to be accepted as a a fully responsible human subject, you have to subject yourself to socially accepted forms. For Deleuze and Guattari there is a linguistic twist. People can experience themselves as human subjects when they enunciate, initiate an act of communication. They are subjects of enunciation. But they also find themselves are subjects of the statement made about them and others. This probably illustrates the difference between the I and the me, which sociologists might have encountered in the work of GH Mead, but which is attributed to Kant here. In the examples in this Plateau, the I is the thinking subject that says 'I believe…' But there are also statements in which the I stands for a general person, a me, which might equally be about a he or a she — 'I breathe...'.

As usual it is tempting to think of a more local example. When I was a full-time academic I was a free thinker able to pursue argument wherever it led according to my own passions, but I was also an employee and found myself in statements made by my employer, along with all the other employees. We were redundant, had to work harder, had to produce more research or do more teaching or whatever. The subject of the statement dominates in stratified social life say Deleuze and Guattari, while the subject of enunciation persists, perhaps as one of those rational delusions, or perhaps only in purely private thought, we might add. The subject of enunciation 'recoils' into the subject of the statement is how they put it. The human subject is 'doubled'.

It's not surprising that post signifying regimes are highly compatible with capitalism with its endless flows of subjectivity and illusory forms of subjective freedom. They are also taken as natural by structural linguistics, where endlessly flexible signifiers just link to each other with an almost infinite capacity to apply to almost anything.
In capitalism, exchange value is abstracted from use values for any commodity whatsoever and is used to set up a system of relations of exchange which apply to everything everywhere. Any Marxist could see the political implications of this similarity, and Guattari spells them out in several places in Genosko (1996).

The Plateau ends with what is a familiar argument referring to abstract machines of language as a way forward. Actual languages are based on actual regimes of signs, which means that language gets stratified by political and economic power. But actual languages also face towards the plane of consistency as they put it, that is they point to an underlying abstract machine that has produced them. Abstract machines, operating at the virtual level, produce all actual systems as possibilities, but there are additional possibilities which have not been actualized and those will help us break out of the current sign regime. Apart from anything else, they will help us deterritorialize current possibilities  As they put it: 'there is no universal logic [behind language] nor is there grammatical reality in itself, any more than there is signifier for itself'
(p. 164). Instead, there is something behind statements and more general semiotic or sense making, activities — 'machines, assemblages, and movements of deterritorialization'. We have to get back to the abstract machinic level in order to realize the full potentials of language and semiotic activity. The only constraint on the operation of those machines and assemblages is the plane of consistency, roughly that any possibilities have to reflect the characteristics of the machine. Later on they are going to say that there may be, for example, certain 'traits' ultimately affecting both content and expression.

Actual linguistic expressions are going to have four basic functions or components (160 – 1).

(1) The generative function means that any particular form of expression can be located in different regimes as in the example of mixed semiotics, so any subjective enunciation can be recaptured by some expert interpreter.
(2) Recaptures are possible because there is a transformational component in language, where one term is transformed into another, again in mixed regimes. This makes it look as if forms of expression alone are the important components, but we want to also restore content as an independent variable.
(3) There are more critical elements too, however including diagrammatic ones where we can take particular particles (emanating from contents?) or signs and see them instead as something more unformed, as traits capable of combining with one another (161). This process of abstraction actually approaches the real state which includes a virtual dimension. We are to do this abstraction both with contents and expressions, and Plateau 3 gives more clues about how to do this with content, as in the next audiofile.
(4) There is the machinic component, as discussed above, showing how abstract machines are 'effectuated in concrete assemblages'. We can do this with the forms of both expression and content. Again the point about content is explained later in this series, but the idea is that matter is formed into more understandable substances before any human semiotic activity takes place.

So what philosophers do, and this might be seen as a critical kind of pragmatics, also known as schizoanalysis, is to look at mixed semiotics and see how they have been generated. Then we map transformations and understand them as 'buds'. Then we develop the diagram of the abstract machine as a matter of 'potentialities or as effective emergences', trying to get back to the 'semiotically unformed matters in relation to physically unformed matters'. Finally, we outline a programme of assemblages that will distribute everything and that 'effectuate abstract machines, simultaneously semioticizing matters of expression and physicalizing matters of content' (162).

We have no further developments of these abstract processes here, but we do have an example. We might take a particular propositions such as 'I love you'. Then we ask about the statements underpinning this proposition which might be different for different groups. We try to examine how a particular regime of signs particularises this proposition. We need to ask about nonlinguistic elements as well, 'variables of enunciation', how motives and social constraints prompt us to enunciate such a proposition. Rather bizarrely Henry Miller is cited as an authority here to show how the term 'I love you' shows that the type of love varies in different societies — romantic love, collective love and so on. That proposition can be captured by a despot and worked into both signifiance and interpretation to produce a signifying chain that controls people — decisions by leaders is also a form of love, for example. There is also a passional form where individual subjects develop particular points of subjectification, as when they fall madly in love as we saw. Plateau  6 on the body without organs also notes developments such as courtly love as we saw. By looking at these different examples we are led to ask questions about abstract machines and assemblages. We would not even get close if we just worked with the formal syntax of propositions, or some transcendental model, working on the apparent universal logic of language or grammar.

We have located particular propositions in regimes of signs and noted possible mixtures, translations and transformations. Pursuing classic 60s sexual politics, we can also try to develop new unknown statements about love, perhaps turning on 'sensual delight, physical and semiotic systems in shreds, a-subjective affects, signs without signifiance' (163). There might be 'feverish improvisations, becomings-animal, becomings-molecular, real transsexualities, continuums of intensity, constitutions of bodies without organs'


Althusser, L. (1977) 'Ideology and Ideological State Apparatuses (Notes Towards an Investigation)'. In Althusser L . 'Lenin and Philosophy' and Other Essays, London, New Left Books (notes here
Deleuze, G. (2006) Two Regimes of Madness: texts and interviews 1975--95. Ed D Lapoujade. Trans. Ames Hodges & Mike Taormina. New York: Semiotext(e) Foreign Agents Series (notes here)
Genosko, G. (2008) A-Signifying Semiotics. The Public Journal of Semiotics II(I), pp.11--21. Online:
Genosko G (Ed) (1996) The Guattari Reader. Oxford: Blackwell. (my notes)
Goffman, E  (1968)  ,Stigma: Notes on the Management of Spoiled Identity. Harmondsworth:  Pelican Books. (notes here)
Guattari, F. ( 2013) Schizoanalytic Cartographies.  Translated by Andrew Goffey.  London: Bloomsbury Academic. (my notes)
Guattari, F.  (2011) The Machinic Unconscious.  Essays in Schizoanalysis, translated by Taylor Adkins.  Los Angeles: Semiotext(e) Foreign Agents. (notes here