Notes on : Schreel, L. (2016) Pure Designation. Deleuze's reading of Hjemslev in The Time Image. Retrieved from :

Dave Harris

The problem is to link cinematic images and signs to conventional language, something that will make them discursive. Saussure and semiotics will be unsuitable in grasping the specificity of cinema. Instead the notion of a relation of designation will be required. This will be based on Lyotard as well as Hjemslev.  Both argued that there is something antecedent and heterogeneous to the signifying capacities of language.  This will involve redefining the idea of the sign and how it is constituted, and reestablish the project of semiotics not just semiology, studying images and signs which are independent of language and express a material which is not captured by language.

In Deleuze's books on cinema, the image has 'an essential duality'—both visible and legible, showing an object of sensation and producing an idea or thought.  The image is composed of 'bodies, characters, parts, aspects, dimensions, distances etc.' that provides it with a content [49]. The frame provides the context and meaning of the elements, and this is how the film becomes a text or narration.  The frame can also justify or explain the contents, for example revealing them as 'normal or regular' (50). Similarly all the image sets are 'integrated into a homogeneous continuity, a universe or a plane of principally unlimited content'.

It looks like we can just treat images or sequences of them as propositions, 'narrative utterance'.  Deleuze critiques Metz here though, and concludes that the image also expresses 'a "non language material"'.  [Typically], this material is then taken as something presupposed by language [transcendental deduction?] even though it is radically heterogeneous to language.  In this way, structural linguistics is limited because it ignores this material: however this '"pre verbal intelligible content"' motivates expression.  We can see it as '"signaletic material"' (51), containing all sorts of sensory features including the visual and the sonic, and also modulating features—'kinetic, intensive, affective, rhythmic, tonal and even verbal'.

The cinematic sign is constructed upon this a-signifying material.  We need a new kind of 'pure' semiotics that gives full account to the 'system of pre linguistic images and signs'.  [Of course] we will have to investigate the ontology beneath the dual nature of the image [as pre and as linguistic].  This is where Bergson fits, especially in terms of the equation of the image and matter, Peirce is tried out, and Nietzsche on the power of the false.  Deleuze himself summarized the project as trying to develop the logic of cinema and its signs and images [in Negotiations?].

The interest in Peirce soon gets replaced, however, in favour of Hjemslev, who is credited as having integrated non linguistic matter or sense into semiotic analysis—he called the former 'purport'.  Hjemslev appears in the book on the time image, in AntiOedipus  and in ATP.  In AO the work is also extended with Lyotard on the figural.

We can understand cinematic signs in various ways.  Traditional realism argues that the sign, or rather its expression, is a sign of something outside, so signs signify, designate or function.  It would be a mistake to think of the content as exterior to the sign, however, and Saussure and Hjemslev alike see the sign as a whole, containing both content and expression, a signified as content and the signifier as expression, for example.  There is an 'isomorphic relation' between a 'conceptual content plane' and a 'phonic expression plane', or two 'functives'.  The semiotic function involves interaction between these two planes producing thought and speech or concept and sound.  However, neither side can be defined without relation to the other.  Hence, the semiotic function is something 'solidary' for Hjemslev as well.  [in Saussure, I recall, the connection is arbitrary, contingent, although he never specified any social forces that might lie beneath this apparently arbitrary act].

Of particular interest is the further distinction between form, substance and '"sense" or "purport"' (54), distinction which applies to both content and expression.  This means that the emergence of signs also depends on 'what is thought of and expressed'[almost a referent, but not a realist one?].  This emerges in H's critique of Saussure.  We can define a formal system as relating just to the connections between ideas and sounds, which express and partition the '"shapeless and indistinct mass"'that is thought on its own.  Language formulates some order and ideas.  The same plastic qualities affect the 'phonic substance' too also requiring language.  So language develops units between these two shapeless masses—the combination produces the form from a vague substance.  [then there is a diagram (55) to show how this further subdivides the distinction between signified and signifier, adding substance and form, in Saussure].

This idea that language develops unities by 'constituting itself between two amorphous masses' implies that there are two planes or substances preceding language, both 'chronologically' and 'hierarchically'.  There is also an implication that substance 'depends exclusively on form' with no independent existence independent of language.  But this means we cannot know about this amorphous substance [I think].

Perhaps we should proceed instead by comparing different languages looking for a common factor, and by focusing on 'the structural principle of signification itself', the linguistic function.  The two will go together because the common factor will be defined in terms of the structural principle before we go on to see how languages differ from one another.  The common factor will be sense or purport [some abstract human intentionality?].  This is something separate from signification but they can only be defined in terms of its linguistic function.  [Don't we have the same problems identifying it if it is not language?]

Purport on the content plane 'refers to unformed and unanalysed thought'(56).  It gives us a factor of the content of the sign common to different languages [even though they represent that signed by arbitrary words]—the example is saying 'I do not know' in different European languages.  The claim is that there is a meaning factor in common, the thought, or the sense of the proposition, but this is articulated specifically in different languages.  Hjemslev thinks we can extract this unformed purport [looks like a dodgy inference to me, or perhaps it is another example of transcendental deduction?], although we can't label it, since once we do, we are [adding form] and involving the perspective of a particular language, adding a 'content substance'.  Each language will provide a different form, different territorial boundaries on the amorphous mass, adding qualities or of values.  [The example below shows that the terms for the colour green actually refer to different subdivisions of the spectrum of visible light in different languages, p.58].  This linguistically formed content is 'content form', and it operates independently of sense or purport [actually here 'content-purport'].  It is connected to sense in an arbitrary way [there's my problem with arbitrary again—I'm sure Guattari would not tolerate that].

We will make an error if we think that content purport is some independent preexisting object that is just being referred to differently.  Differences in language do not reflect different 'realizations' of an underlying substance.  They reflect different realizations of a 'principle of formation', enabling a different form to be realized from the 'identical but amorphous purport'. A particular language 'carves up' or forms content-purport—this is the semantic dimension enabling phrases to refer to something outside themselves or to 'designate' something non linguistic.  But this is also arbitrary [that is not determined by some real content]. In this way, content substance depends on content form and is never independent.  [Could easily be the usual weasel around what causes what—all of it missing social forces affecting form and content].

So the function of the sign is to add form to a content, to 'constitute a content-form', which can only be explained by looking at the function itself, with no independent influences.  Or content and expression become linked and 'solidary', a unit.  This will refer to purport, again in an arbitrary way [with a further example of how the actual signs in different languages refer to different physical to divisions of light.  But doesn't this assume that there is some real content against which the arbitrary nature of language is being asserted?  There is some real light, which makes the divisions of it into colours the arbitrary bit?].

Having looked at content, we can now look at expression, to examine relations between form, substance and purport.  Expression will lead us into phonics.  There is expression purport, in this case an amorphous mass of sounds, a whole 'phonetic sphere' out of which actual speech will be realized.  The mass takes the form of a continuum, which is both 'unanalysed but analysable'.  Once more different languages operate with an arbitrary limited number of phonemes or figures.  The example notes that the human mouth is capable of all sorts of sounds according to which of its zones are utilized, but that different languages use different particular variants.

As above, we require an expression-purport to move through an expression-form into an expression-substance.  This will be a sequence of sounds in a particular language, 'spoken by an individual person hic and nunc' (60) [the Latin phrase for 'here and now'  is used throughout].  Again the substance makes no sense except by relating it to the form, which is itself a sound sequence that can be interpreted as a legitimate use of acceptable phonemes.  Again, it is the sign function that links sound to content, expression-forms to content-forms.  [A diagram on page 61 offers a summary of Hjemslev on the 'form-substance-purport triad' at work within both semiotic planes of content and expression.]


Overall, we have to remember that there is a paradox here, that signs refer both to content substance and expression substance.  Particular sounds become expression substances only after a sign function which will classify it as an expression form, located with other expression substances within the accepted vocabulary.  In Hjemslev's terms, the sign faces both outwards to the expression substance and inwards to the content substance.  The sign is always a unit consisting of both content- and expression-forms, and this unity or solidarity provides the sign function.  A sign is not just an expression or signifier like a kind of label attached to already existing things.  Saussure reminds us that it is a concept and a sound that is united in the sign, and the sound itself is a representation of the material produced by the senses [so material in that sense].

Deleuze likes the distinction between matter and substance.  We have to remember that substance here is matter or sense that has been formed linguistically, the result of a linguistic 'transcoding' (62) of the characteristics of matter.  These forms are themselves 'projected' on to purport [Hjemslev compares it to  a shadow cast on an undivided surface].  For Hjemslev non linguistically formed matter has to be formed by 'physical, biological and phenomenological points of view'[no doubt capable of being rendered as percepts in Deleuze?].  In this way, any distinction between substance and form only makes sense as categories used in the study of language [maybe].  Purport or sense can be 'studied by other disciplines in different ways'.

Deleuze uses this to argue for a 'biophysical origin of language'(63) in Logic of Sense.  We can grasp sense in a 'primary' even an 'emotive' level—'noematic sense'.  Things like phonemes were originally 'a libidinal movement of the body'.  The sense at this level is a 'depersonalised, disunified experience', the experience of a BWO [so we can experience it directly?].  It is also the first of a series of orientations of the psyche of the child [shades of pre-symbolic communication in Kristeva or Ettinger?].  There is no mastery of language required.  In AO, Hjemslev is used to argue for the idea of phonemes as more than just the affects of the signifier, but rather something that challenges the conventional signifier—'"schizzes, point signs or flows breaks that collapse the wall of the signifier"' (64).

Again we have to get technical.  Conventionally [ie in straw men], the concept of the sign implies only a link to an expression substance or content substance, not to purport.  A sign can designate something, but only after it has interiorized it.  In the example, the word 'ring' is a sign for a thing, but the thing is not connected to the sign in the traditional realist senseConventionally, the thing would be seen only as an element of content-substance, linked to a content-form and connected to other kinds of content substances [the example is the sound that comes from a telephone].  What seems to be happening is that the content form of the sign monopolizes ['subsumes'] the thing as content substance [in the normal set of power relations].

Deleuze wants to preserve the notion of a presignifying expression because he wants to see cinema, language, and art in general as potentially critical, capable of delivering 'a violent shock' to return our experience to that of the child before entering language.  Cinema must develop a signifying language if it is to deliver affect.  But this can not mean a total abolition of the normal textual order of expression with all its regulation and codes  Instead, we have to 'reinvest' in the prelinguistic material that existed before coded distinctions.  This would energize a new form of seeing which would be the outside of  language.

The example of the use of colour in Godard's Weekend demonstrates that colour no longer refers to a particular object [not just to blood] but shows a more 'devouring' or 'absorbent' function, a quality common to completely different objects.  We can then replace immediate representations, metaphors and figures with a connection more to bodily intensity and even to pure matter.  This is still  a symbolism of colours, but not a regular connection between the colour and an affect, or one that goes on to establish further connotations, say between white and innocence.  The colour is the affect itself working 'directly on to the nervous system'[apparently Godard himself], some pure sensation of colour, a material base that can even help us grasp the affects of music if we connect them [apparently in the works of some composers such as Messiaen].

This further implies that the cinematic image or sign, even the shot, cannot simply be reduced to a representation or a signified. This is particularly evident since the shot is not an arbitrary sign [in some existing and arbitrary language], but attempts to, is motivated to, relate to sense.  This will remain obscure if we try to see it in conventional linguistic terms [because signs there do not attempt to relate to sense directly, but only as 'substance for a form'(66)].  We want a cinematic sign to make the referent visible, and this involves deliberate motivation [because it is an artificial language].

In AO, Hjemslev is seen to offer a grasp of linguistic flows rather than structural conditions.  In ATP, Hjemslev is used in connection with the BWO which is permeated by flows and unformed matters as opposed to being coded, stratified or territorialized.  These processes give form but at the price of domesticating intensity and stuffing singularities into various systems.  Hjemslev is seen to offer an immanent theory of language instead, but it is the flows of desire that underpin the formation of form, substance, content and expression.  The implication is that we need to destroy conventional signifiers, decode language, as a result of definite motivation [of a kind that exceeds the social order]: that will replace structural notions of the articulation of signs.

It is that structural articulation, an arbitrary relation between sign and purport, that distinguishes between signs and referents and grants autonomy to the linguistic sign, apparently requiring no motivation nor particular speakers and situations.  To reintroduce motivation would be to allow the collapse of signification into expression.  This would further produce a collapse between content and expression.  The actual shape of the signifier could only be grasped by looking at the situation in which it was produced—signs would become like 'a cry, a rush of breath or a chanted melody'(67).  [It is the old problem.  If language is autonomous it looks nonpolitical, but if everything is politicized, the astonishing autonomy and objectivity of language collapses].

Even so, this argument implies that there is some excess of sense after all, beyond organized signification.  How can it be grasped semiotically?  Here we apparently need Lyotard and the figural, especially his 'theory of pure designation', admired by D and G in AO.  For D and G, the signifier is exceeded by '"figurative images"' 'outside' and the 'pure figures' inside which compose it.  The figural itself is crucial to explain how signs manage to relate their various components.  We can see the consequences where letters or words become partial objects or 'a signifying signs', controlled by various kinds of desire.  They also use the term '"an order of connotation"' of conventional signification.

Like Lyotard, they argue for certain presignifying signs at work within signifying orders, appearing whenever signification becomes problematic, and operating at a 'libidinal' level.  Lyotard had argued that there was an antecedent level or order which was radically heterogeneous, involved in referentialty  or designation as a signifying act.  He takes designation to refer to an experience of the speaker independent of signification.  [It seems to involve having to use or apply a system of language to something apparently external].  It involves both 'intentionality' and 'distance', and implies '" a profound exteriority that resides at the limit of discourse"' (69). 

This designation takes place before signification as an expression of prelinguistic sense.  Normally, conventional signification and transparent communication serve to neutralize this libidinal order—it is important within ordinary language use, but it can not be experienced directly and therefore not expressed.  The closest we get to explicit expressions in the visual arts.  We also see it appearing in disrupted language as in poetry or the free association of words in psychiatry.  Lyotard uses the terms 'figure image' and 'figure form' to relate to these two forms of explicit expression.

The latter shows that there is a non language present even in language, another order involved in expression.  The example in Lyotard is colour again—apparently, to designate a tree as green is to go beyond signification, but add expression.  Designation inhabits a different space.  It shares with language 'figural space'(70).  Lyotard draws on both Hjemslev and Godard to argue that there is something in enunciation that can not just be grasped from a linguistic point of view.  It is posited by speech but lies beyond it in some 'originary spatialization'.  We can understand it if we consider a pure designation as something not entirely linguistic, nor even symbolic [defined,by straw man Saussure, as 'rendering sensible an absent thing'].

We turn at this point to an anthropologist, a certain Leroi-Gourhan who saw the origin of language as arising in emotive situations with definite motivations [the example seems to turn on a sacred ritual where a narrator designates, points to or gestures at various painted figures.  It shows that designation does not involve words being used in the absence of the designated thing but rather in its presence; that the designated thing is a symbol and an opaque one [requiring words to be added in order to explain it?]. However it is not a matter of naming some thing as in classic symbolism, but rather manifesting the thing, grasping the thing in experience, making it visible: there is no split between the utterance and the object, but rather a depth, an '"atmospheric spatiality"'.

We are leaving behind unmotivated [structural] signification and the usual notion of the symbol [apparently this is the only way in which motivation appear as in structural linguistics].  The word is not just the linguistic sign nor a symbol for an absent thing.  Designation does not just connect signs to things, but instead shows how things become symbols in their presence [how they acquire magical socially significant qualities there and then?].  Designation in the form of phonemes uttered by the speaker are not to be seen as linguistic, intended to create some representation of an object in our minds that conforms to the evidence of our senses [one of Saussure's arguments apparently]. 

What is designated itself becomes an opaque sign, something asignifying which will require a subsequent secondary process, what D&G call 'connotation'.  Designation reveals some unknown facet of the thing, and it is the designated thing which can become a sign.  Sometimes, apparently we can just see the word attached to the thing, feel it without actually reading it.  D and G suggest [in AO]  that this example shows the importance of [motivated] designation to close the gap between the word and the thing [which was what I was trying to say was missing before in the idea of a purely arbitrary relation].  They even go on to say the words themselves are not signs, but rather they transform things or bodies into signs through designation. Words that arise from designation also have a '" strange ability to be seen not read"'.

All this relates to the work on the cinema because it shows that the vocal and the visual forms of expression are independent and heterogeneous.  We cannot understand the visual by using conventional significations [except as an order word?] because visual elements point to something unknown, opaque, and distant from signification.  We can nevertheless see these qualities as visible and sensible but only through 'connotative, deictic' [
'Relating to or denoting a word or expression whose meaning is dependent on the context in which it is used (such as here, you, me, that one there, or next Tuesday)'] designation.  This can take the form of 'emotive force', appearing as deliberately different from signifying structures, apparently conveying sense from the object itself.

We see this in the discussion of optical and sound situations, as well as the 'any space whatever'.  These designate even while being emptied of any easily attributed content [in conventional linguistic terms].  The cinema works by exploiting the duality between two 'autonomous and heterogeneous planes of expression'—the visual plane with 'pure prelinguistic images', and a phonic plane with 'pure presignifying signs'.  Together these offer an account of what is potentialy utterable in a language system, the equivalent of purport.  Every language works through form and substance which actualize the potentials into raw material, but this breaks the connection with the prelinguistic.  Cinema shows this material in images and signs and thus 'reinvests' it.  Signification is disturbed, so 'corporeal, rhythmic, affective sensorial features resurface' (73). The result is to break with the speculative and pragmatic in favour of a depersonalised experience in which we can become absorbed by the world. 

We might experience things that are strange or incomprehensible, but that does not mean that they are entirely illusory deprived of content.  Instead it shows that there is a 'a field of non linguistically formed sense'(74).  But every intentional process of meaning is based on this condition ['projected upon' it], even though it escapes meaning itself—is the sense that preexists language, and we draw upon it to understand images and words.

back to Deleuze page