Postulates of Linguistics

by Beth Metcalf

In chapter 4 of A Thousand Plateaus Deleuze and Guattari examine four postulates of traditional representational linguistics.  They challenge these postulates of representation and contrast them with their own linguistic postulates based on univocity.

I.                   “Language Is Informational and Communicational”

The first postulate of representational linguistics assumes that language is informational and communicational.  That is, language is postulated as a system of exchange among subjects about external facts in an intersubjective reference frame.  This assumes that language is structured in a manner that represents possibilities of an objectively shared world.   

However in contrast, D&G postulate that language transmits order-words to compel obedience.  (76) “Language is not life; it gives life orders.” Language does not represent information about an external world.  There is no non-linguistic reference.  Language does not communicate something seen.  It goes from saying to saying.  Therefore, the “origin” of language is not a direct discourse of metaphor or symbol in direct correspondence to something seen, but indirect discourse of hearsay.  There are “uses” of metaphor (signifying similarity) and metonymy (subjectifying continguity) only if these uses are surface effects presupposing indirect discourse. 

Therefore, language is the transmission of order-words that does not represent or communicate information of something seen.  Language is hearsay that goes from saying to saying.  Austin showed that language is not reducible to the extrinsic relations of indicative or imperative modes.  There are performative intrinsic acts of saying.  D&G notice that this means language can’t be a code for conditions of possibility.  Pragmatics must be the presupposition intrinsic to language.  Pragmatics must no longer be the speech/language alternative --- no longer either the uses external to language or as determined by conditions of possibility.  Now, the semantic (language) and syntactic acts (speech) can no longer be defined extrinsically to each other, or by speech as individual application of language.  Rather, speech-acts are presupposed by language in use.  For D&G, the pragmatic function of speech-acts now becomes the presupposition of all dimensions of language. Therefore, there are no prior relations of individuals or objects.  There are only interactions that make themselves seen in speech-acts.  These interactions are not structural relations (or relations derived from a structure).  Rather speech-acts, as pre-individual, create interactions between individuals or groups. 

With direct discourse there may be a performative self-referentiality of shifters in a structure of subjective or intersubjective communication (Benveniste).  Or, there may be the realization that intersubjective communication is no better than significance of information in accounting for performative assemblages (Ducrot).  But in any case, direct discourse cannot reach the performative assemblages of speech-acts as the indirect discourse of order-words.  Order-words are not merely commands referring to something seen.  They are any statement acting as a “social obligation”.

The order-word is language-function coextensive with language.  The order-words are the relation of every statement to the presuppositions of speech acts accomplished in the statement.  The statement is the order-word in redundancy.  I take this to mean that every order-word of a statement (form of expression) and its presupposed statements in a current language (form of content) is in a relation of redundancy where the form of content and the form of expression relate (without identity) as a new form of content presupposed by further acts of expression.

D&G say (180), “Whatever the differences between significance [information] and subjectification [communication]….they have it in common to crush all polyvocality….and operate by signifying biunivocalization and subjective binarization.”   Language is not information or communication.  Rather, language is the transmission of order-words.  The redundancy of the order-word is prior to the significance of information and the subjectivity of communication.  Signifiance and subjectification presuppose the redundancy of the order-word in a social field. 

Information is individuated and communication is subjectified only because a prior collective assemblage actualizes a use.  The collective-assemblage is the redundancy of act and statement.  These acts are the set of incorporeal transformations in a society attributed to bodies in that society.  The incorporeal transformations are not corporeal actions or passions.  Rather, they are the incorporeal speech-acts expressed of statements and attributed to bodies.  Collective-assemblages are in continuous variation of these transformations.  The instantaneous transformation of the order-word transforms bodies to which the speech-act is attributed.  I take the redundancy of the order-word, then, to be intensity of incorporeal transformation that changes the nature of the statement with each act of redundancy.

Indirect discourse is the reported statement within a reporting statement --- order-word within the word.  Direct discourse is extracted from indirect discourse only insofar as a collective-assemblage distributes variables that enter into temporary constant relations for an actualized use.  There are pragmatic variables of use internal to enunciation coextensive with language.  Pragmatics is presupposed by uses of significance (information) and subjectification (communication). 

There is no individual enunciation.  There are only collective-assemblages of speech acts performed in redundancy with statements of order-words.  The repetitions of the order-words are the internal redundancy that changes the nature of the assemblage.  Expressive acts are the incorporeal transformations attributed to corporeal bodies in a given society.  This is consistent with Deleuze’s univocity.  These “acts” are the sense that inheres in the statement (form of expression) used in a given society and attributed to bodies (corporeal content) that can possibly be seen by a given society.  Order-words are the intensive incorporeal transformations of the event.  Order-words are the enunciation of incorporeal transformations.  This is the immediacy of intensive redundancy that changes the nature of the bodies to which the incorporeal transformation is attributed.

Now, this new pragmatics is not merely a residue of external situations or contexts in opposition to an internal form of possibility.  Indirect discourse is a reported statement (attributed to corporeal content) within a reporting statement (incorporeal expression of transformation).  Language is indirect discourse.  Uses of direct discourse are extracted from prior indirect discourse because significance and subjectification of an assemblage are distributed in expression and attributed to bodies.  The new pragmatics has variables of usage internal to enunciation as conditions of temporary use. 

The statement is individuated and enunciation subjectified only when collective-assemblages determine it to be an actualized use.  What are the speech-acts in redundancy with statements that actualize temporary uses?  These acts are not actions/passions of bodies.  They are incorporeal transformations attributed to bodies.  When linguistics is reduced to constants, it links the statement to a signifier and enunciation to subjectification.  This puts significance and subjectification into a relation of constant variability.  Then, pragmatics is merely a residue of external circumstances.  But the order-word is a new pragmatics of language-function.  There is a use of order-words as the event or act of intensive incorporeal transformation. 

II.                “There Is an Abstract-Machine of Language That Does Not Appeal to Any ‘Extrinsic’ Factor”

As we have seen, the first postulate of representation assumes that language is direct discourse with extrinsic reference.  That is, representation assumes an internal form of language corresponding to the external structure of reality.  Now, we see that the second postulate of representation assumes that language is an abstract-machine that does not appeal to any factor extrinsic to language.  Together, these first two postulates of representation mean that language refers to something extrinsic without appeal to any extrinsic factor.  But how can this correspondence happen unless there is miraculous appeal to some transcendent ground?

D&G counter these first two representational postulates with their own postulates of univocal being.  Their first postulate says that being is the language of indirect discourse without extrinsic reference.  Being is saying.  Their second postulate says that language is an abstract-machine that does appeal to a factor (modal acts) extrinsic and heterogeneous to any supposed constant of language.

Representational linguistics postulated an abstract-machine of form-content duality.  The form of language represented the form of things in one homogeneous structure.  Nothing could appeal to any factor extrinsic to the structure of the already formed content.  However, D&G counter this postulate by making a distinction between corporeal modifications of content and incorporeal transformations of expression.  There is now parallelism between content and expression, each having its own heterogeneous formalization.  Therefore, expression never represents content.  The two formalizations are heterogeneous (like Spinoza’s attributes).  The incorporeal acts are the expressed events that inhere in statements and are attributed to the corporeal bodies of content.  The body or state of things is not a referent of the expression.  There is only intervention of a speech act (intervention of sense). 

(86-7) “The warp of the instantaneous transformation is always inserted into the woof of the continuous modifications….expressions are inserted into contents….In short, there are degrees of deterritorialization that quantify the respective forms and according to which contents and expression are conjugated, feed into each other, accelerate each other, or on the contrary become stabilized and perform a reterritorialization….In short, the way an expression relates to content is not by uncovering or representing it.  Rather, forms of expression and forms of content communicate through a conjunction of their quanta of relative deterritorialization, each intervening and operating in the other.”

The form of content and the form of expression are heterogeneous parallel series.  A formalization of content may intermingle with a formalization of expression for a temporary effect or use.  Each such heterogeneous intermingling is inseparable.  And that is because, with the separating movement of deterritorialization, the degree of intensity separates and necessarily changes the nature of the formalization.  That is, contrary to the second postulate of representational linguistics, the abstract-machine of language does appeal to a factor extrinsic to any already formalized assemblage. 

It is a mistake to think that content causes expression.  Content presupposes a social machinic assemblage of bodies.  Expression presupposes a collective-assemblage of enunciation.  It is also a mistake to see the form of expression as a linguistic system or as a deep structure which would only make the abstract-machine of language into a synchronic set of constants.  D&G say (90-1), “We will not object that the machine thus conceived is too abstract.  On the contrary, it is not abstract enough….allowing it to consider linguistic factors in themselves, independently of nonlinguistic factors, and to treat those linguistic factors as constants.  But if the abstraction is taken further, one necessarily reaches a level where the pseudoconstants of language are superseded by variables of expression internal to enunciation itself; these variables of expression are then no longer separable from the variables of content with which they are in perpetual interaction.”  If language does not appeal to any heterogeneous extrinsic factor, then it will be reduced to a homogeneous constant relation of variability.  In opposition to the second postulate of representation, linguistic factors of expression must not be thought independently of extrinsic and heterogeneous nonlinguistic factors of content. 

D&G envision a new abstract-machine that is the diagram of an assemblage.  (91) “Content is not a signified nor expression a signifier; rather, both are variables of the assemblage.  We get nowhere until the pragmatic, but also semantic, syntactical, and phonological determinations are directly linked to the assemblages of enunciation upon which they depend…..The abstract machine as it relates to the diagram of the assemblage is never purely a matter of language, except for lack of sufficient abstraction.  It is language that depends on the abstract machine, not the reverse….”  The diagram of the assemblage appeals to an extrinsic speech-act that intervenes as the incorporeal transformation that is the expressed of the statement attributed to material content.

III.              “There Are Constants or Universals of Language That Enable Us to Define It as a Homogeneous System”

Structural invariance (whether atomic or relational) is the basis for claims of a scientific study of linguistics.  Although linguists recognize that language is heterogeneous, representational postulates still tie the abstract machine to universals or constants in order to carve out a homogeneous system for scientific study.  D&G credit Labov with seeing that every system is in variation and is not homogeneous, universal, or constant.  There is no reason to assume that variables are necessarily different points of view on one theme that keeps the principle of the statement constant.  There is no reason to assume constant relations when we reach the heterogeneity of form of content and form of expression.  D&G envision a new pragmatics of ‘uses’ immanent in language.  They think of language not defined by invariants, but by lines of continuous variation that are different within each language.    

(94-5) “….there is a constant tendency to seek a “reduction”: everything is explained by the situation….But this is to content oneself with extracting a pseudoconstant of content, which is no better than extracting a pseudoconstant of expression.  Placing-in-variation allows us to avoid these dangers, because it builds a continuum or medium without beginning or end.  Continuous variation should not be confused with the continuous or discontinuous character of the variable itself….A constant or invariant is defined less by its permanence and duration than by its function as a center, if only relative”. 

Placing linguistic elements in continuous variation creates new concepts and new elements.  But none are given in advance and none are final.  There is no principle of invariance that would maintain a prior structure in a constant relation of variability.  There is a new pragmatic function of invariance which is only a relative and temporary ‘use’.  Each invariant use is a singular and fragile difference incommensurable with any other.  All language is in immanent continuous variation.  Placing all linguistic elements of both content and expression in variation is to make language itself stammer with disparate styles or assemblages of enunciation.  There is no prior concept of how content and expression are to be conjoined.  There is a new form of redundancy (and…and…and).  Language becomes continuous variation of intensive assemblages.  Continuous variation is not an actual determination of constant relation.  Language places variables in a new state of continuous variation without constant relation.  There is a (99) “cutting edge of deterritorialization of language” for the variation of variables themselves.

The abstract machine of language is not universal or general.  Its rules vary with its variation, because it is the dice game of univocity.  The abstract machine is a diagram of continuous variation.  The concrete collective assemblage organizes variables as a function of the diagram --- which variables will take on a constant relation (a temporary ‘use’) and which will be more fluid.  The abstract machine and collective assemblage function together as singular variety in continuous collective variation.

IV.              “Language Can Be Scientifically Studied Only under the Conditions of a Standard or Major Language”      

This postulate of representational linguistics assumes there is a Major Language that can bring linguistic variables under the domination of homogeneous conditions.  Although everyone knows that language is heterogeneous and variable, linguists have carved out a homogeneous system of universals and constants in order to study language as an object of science.  But D&G see this scientific model as one with the political power of a dominant language.  The assumption of homogeneous universals and constants of language brings with it a presumption of a metalanguage or common default language which must be used in a scientific study of linguistics. 

But D&G reject the Representational assumptions.  Such a Representational Major Language does not exist.  They tell us (103) “You will never find a homogeneous system that is not still or already affected by a regulated, continuous, immanent process of variation.  The major and minor are not two kinds of language, but two uses of the same language.  Constants do not exist apart from variables.  Constants are drawn from variables in two ways.  There are two treatments of variables --- the “major” treatment and the “minor” treatment.  D&G tell us (106-7) that the major treatment of language consists of extracting constants from variables.  It gives orders to be obeyed in an extensive system of judgment. The minor treatment is a placing in continuous variation. It is the intensive composition of ordinality.  The order-word is the variable of enunciation that defines either a major or a minor usage.  D&G call the order-words the only “metalanguage” that accounts for this double treatment.  There is no metalanguage of universals or constants, as the Representational scientific model assumes.  The order-word is the variable that determines which treatment (major or minor) is being used.

With the major use, the order-word is a death sentence.  With the first treatment of the order-word, death is the expressed of the statement.  Death is the incorporeal transformation that is attributed to bodies.  It reterritorializes extensive systems of actions and passions.  It separates and distinguishes bodies.  If the incorporeal transformation is the expressed of order-words attributed to extensive bodies, then there is an act of transformation where enunciation fuses with the statement to become a reterritorialized sentence.    

D&G write about this major use (107), “Now if we consider the first aspect of the order-word, in other words, death as the expressed of the statement, it clearly meets the preceding requirements [of incorporeal transformation]: even though death essentially concerns bodies, is attributed to bodies, its immediacy, its instantaneousness, lends it the authentic character of an incorporeal transformation.  What precedes and follows it may be an extensive system… is neither action nor passion, but a pure act, a pure transformation that enunciation fuses with the statement, the sentence.” 

With the minor use, the order-word may be a warning to flee.  Then, there is still an incorporeal transformation attributed to bodies, but the variables are in a state of continuous variation --- a pragmatic placing-in-variation.  There is an intensive transformation --- an intensive change in the nature of the assemblage.  The plane of consistency has an assemblage that becomes absolutely deterritorialized drawing the two forms of content and expression together in conjunction of a new plane of consistency with new cutting edges of deterritorialization --- a new abstract machine --- a diagram of a new assemblage.

D&G say of this minor use (108-9), “If we consider the other aspect of the order-word, flight rather than death, it appears that variables are in a new state, that of continuous variation.  An incorporeal transformation is still attributed to bodies, but it is now a passage to the limit….In continuous variation the relevant distinction is no longer between form of expression and a form of content but between two inseparable planes in reciprocal presupposition.  The relativity of the distinction between them is now fully realized on the plane of consistency, where the assemblage is swept up by a now absolute deterritorializaiton  Absolute, however, does not mean undifferentiated….The relation of presupposition between variables of content and expression no long requires two forms: the placing-in-variation of the variables instead draws the two forms together and effects the conjunction of cutting edges of deterritorialization on both sides; this occurs on the plane of a single liberated matter….”  

Therefore, just as D&G postulate (265-6) “two planes, or two ways of conceptualizing [one] plane”; there are two languages that are really just two treatments of one language.  These two planes with their treatments are not on the horizontal axis of content and expression, but the vertical axis of territorialization-reterritorialization or deterritorialization (88).  There is (281-2) the plane of organization and development (the transcendent plane of disparate and incommensurable uses of major language) and there is the plane of immanence and univocality (the plane of consistency of minor language).  “It is a jumping from one plane to the other, or from the relative thresholds [of territorialization or reterritorialization] to the absolute thresholds [of deterritorialization] that coexists with them….”  We can see that the Representational postulates of a Dominant Major Language have no relation to the postulates of univocal being-saying where the two treatments of language intersect.  The old linguistics of Representational pragmatism has no relation to the new linguistics of univocal being-saying with its new pragmatic uses.  The intersection of the two treatments is a relation of vice-diction that opens the forms and allows no pseudoconstants to subsist.

The Representational postulates would lead us to assume that a Major Language can be used for scientific study of linguistics.  But this is to merely bring minor dialects under the overcoding domination of the homogeneous conditions of the Major Language.  A dominant use of one standard Major Language is the death sentence of closed Representation.  But D&G tell us that the question is not how to flee death, or flee a major language.  Rather, it is how to make both major language of death and minor language of flight creative.  Minor language exists only in relation to a major language that it makes minor by lines of flight in intensive continuous variation.  It is not a question of reterritorializing a major language.  Rather, becoming minor is deterritorializing a major language.  If we try to use the Representational postulates of Major Language to reach the univocal being-saying of D&G, we merely block the lines of flight.

Return to Home Page