Intro      Writing       Les mots et les noms       The unspeakable Other       Painting       Mathematics       The Thing Itself       Bibliography

The unspeakable other

Having argued for language as 'the limits of our world', post-structuralism returns obsessively to the question of what has been left out of this schema, and what should — by its own logic — be both unthinkable and unnameable. Defending this generalisation would require a much larger work, and locating 'post-structuralism' in the work of individual theorists is no less problematic. Given the space available, and the flaws of metonymy (especially when even the part's status as part of that whole is debatable), I shall proceed as follows. As my premise, I shall privilege Barthes, Derrida, and Kristeva as exceptionally prominent theorists in the field of post-structuralism. I will highlight a few quotations from Barthes and Derrida, in which the unnameable raises its head; their texts and arguments will, however, remain tangential, as a suggestion only of the wider scope of the phenomenon. The focus will be on Kristeva's writings: in particular, her notion of the semiotic and to what degree this resolves the contradiction faced by post-structuralism.

We must remember that according to the most fundamental tenets of post-structural theory, there is no thought before language; without holding that banality in our minds, we risk falling into the equally banal assumption that of course there are things which cannot be described. If language is that which allows us to think, and which provides the concepts with which we think, how can a concept, a thought, elude language? Nevertheless, Barthes writes in A Lover's Discourse: Fragments:

Yet, at the same time that adorable says everything, it also says what is lacking in everything; it seeks to designate that site of the other to which my desire clings in a special way, but this site cannot be designated; about it, I shall never know anything; my language will always fumble, stammer in order to attempt to express it, but I can never produce anything but a blank word... (1979: 19; my emphasis)

Here, designation by name is still associated with knowledge: 'this site cannot be designated; about it, I shall never know anything'; however, it is able to exist, as a concept, even though he cannot know or express it and 'can never produce anything but a blank word'. Further down, the text continues:

Yet the more I experience the specialty of my desire, the less I can give it a name; to the precision of the target corresponds a wavering of the name; what is characteristic of desire, proper to desire, can produce only an impropriety of the utterance. Of this failure of language, there remains only one trace: the word 'adorable' (the right translation of adorable would be the Latin ipse: it is the self, himself, herself, in person).' (1979: 20)

With the increase of desire, the ability to name decreases, becoming a wavering, then an impropriety, then a failure. All that remains is a 'trace', 'the word 'adorable'', whose right translation is first an appeal to Latin (old origin of meanings) and then to material presence: 'the self, himself, herself, in person'. In these passages, the signified to which Barthes alludes is referred to as 'that site of the other to which my desire clings in a special way', as 'the specialty of my desire'; the possibility of a signifier is refused. (I respect his refusal: I write 'alludes', and 'referred to' rather than accuse of him of naming it.) The closest one can get to representing this signified is through the trace of a word, 'adorable', whose 'right translation' is not only another word, but also a physical being. The signified is aligned with the material presence of the beloved, 'in person': the signified is conflated with a referent. (Is this not always the case, in [love]?)

Later in A Lover's Discourse, we find that even when the signifier is written, the signified still eludes it:

... discourse on love though I may for years at a time, I cannot hope to seize the concept of it except 'by the tail': by flashes, formulas, surprises of expression, scattered through the great stream of the Image-repertoire... (1979: 59)

He is discoursing on love, but 'cannot hope to seize the concept of it': far from there being 'no pre-existing ideas' outside the system of signifiers, the concept exists, apparently independently, exceeding and escaping any names given to it. Having argued for language as the limits of the post-structural world, one cannot logically even speak about that which exceeds language, for any such thought, without the language, should be impossible. It becomes the unspeakable Other of post-structuralism.

Much fruitful material remains for analysis, in this and other of Barthes's works (notably Le Plaisir du Texte also), but we must continue to Derrida. Surprisingly, joyously, the unspeakable other does not emerge from his texts in a tortuously argued abyss of meaninglessness, as one might expect, but in yes! Once more, we find that a certain special signified insists on preceding its signifier:

I say the yes and not the word 'yes,' for there can be a yes without the word, which is precisely our problem. One ought, then, to have preceded all of this with a long, knowledgeable and thoughtful meditation on the meaning, the function, the presupposition above all of the yes: before language, in language, but also in an experience of the plurality of languages that perhaps no longer belong to linguistics in the strict sense.' (1991: 590; my emphasis)[1]

The enthusiasm of yes, eagerly leaping ahead of its signifier, is at first tempered by seeing it as a 'problem'. The following sentence, however, seems to be written by a Derrida drunk on yes, gently mocking as it does the earnest requirement of 'a long, knowledgeable and thoughtful meditation on the meaning, the function, the presupposition above all of the yes' — something the Derrida of other essays would certainly not neglect, and yet which here is restricted to a non-sentence following the colon. Similarly to Barthes's 'discourse of love' which can never 'seize the concept' that he has just named, Derrida has had to name yes, to communicate, even while insisting that it can exist without the word, and that it is 'before language' as well as 'in language'. That it pre-exists language is later made still more explicit:

I have had to yield to the rhetorical necessity of translating this minimal and undetermined, almost virgin, address into words, into words such as 'I', 'I am', 'language', at a point where the position of I, of being, and of language still remains derivative with regard to this yes. This is the whole problem for anyone wishing to speak on the subject of the yes' (1991: 593)

Translating implies pre-existence in another language (which is not, it seems, a language of words for it is being translated into words). Before the acquisition of subjectivity that 'I', 'I am' implies (the entrance into the mirror phase, in Lacan's schema; the separation of oneself from others in order to speak), before the acquisition of language (the symbolic order, again following Lacan), there is yes, which cannot therefore be spoken of (it precedes language and the preconditions of language). For Barthes, the unspeakable Other is (something we call) desire; for Derrida, it is yes, which is related to desire — not just in the mind's ear, which hears a gasping yes! yes!, but in Derrida's pun. Yes, he says, is a sending-to-oneself — in French, s'envoyer, what we would call getting oneself off.

Both Barthes and Derrida, then, locate this unspeakable Other before and outside of language, and both associate it with the sexual drive (for Barthes, desire and the material presence of the beloved; for Derrida, the satisfaction which concludes — or is the summit of — desire). One cannot explain it away as being something purely material, for both are in fact dealing with concepts — Barthes with desire, love; Derrida with yes — which have no material referent. However, the premise that thought is possible only through language forbids us to theorise a concept which exists in our minds independently of language. The 'triple cluster of referent, signified and signifier' (Kristeva, 1974: 68) is collapsing; the status of the signifier is crumbling.

The riddle stands: what is a signified, but has no signifier, can exist without language but is not a referent? In the face of this challenge, Kristeva has found a way of theorising this something without rejecting post-structural tenets. This is made possible by distinguishing two opposite modes in language: the semiotic and the symbolic[2]. The symbolic is the aspect we traditionally think of as language (the system of signs, syntax, and so forth) and continues to act in accord with post-structural principles. The semiotic is both in language and before it: a 'preverbal functional state' (Kristeva, 1974: 25) that 'precedes the establishment of the sign' (ibid: 26). As 'preverbal' suggests, it has a special place in infancy but nonetheless continues into adulthood:

While we can imagine it in the crying, vocalisation, or gestures of babies, the semiotic functions in adult discourse as rhythm, prosody, word-play, nonsensical logic, laughter. (Kristeva, 1977: 14)

It is part of adult discourse in the same way that the unconscious is part of thinking. Kristeva notes that

Our positing of the semiotic is evidently inseparable from a theory of the subject that takes into consideration the Freudian positing of the unconscious. (Kristeva, 1974: 30)

The debt to Freud extends beyond the 'unconscious'. Through the Freudian theory of drives, Kristeva is able to theorise the semiotic without falling into either side of the material / conceptual, body / mind binary. Drives, it must be remembered, 'can be reduced neither to the biological nor to the social; they operate in between these two realms and bring one realm into the other.' (Oliver, 1997: xvii) They are present before the mirror stage, before entrance into the symbolic order (which is not identical to but includes Kristeva's 'symbolic'), and thus before one is constituted as a social subject, but are nonetheless arranged by social constraints:

Discreet quantities of energy run through the body of he who will later be a subject and, in the course of his development, they arrange themselves according to the constraints imposed on this body... by family and social structures. (Kristeva, 1974: 23)

It is in the articulation of these drives that Kristeva locates the semiotic. Kristeva's theory also argues for a material basis to signification, further establishing the physical body within language.[3]

The riddle was, 'What is a signified, but has no signifier, can exist without language but is not a referent?' In the semiotic, we find a mode of language which precedes language, which can exist without language but which is not a referent, and which, moreover, is founded on drives, thus escaping the logical opposition of mind vs body, signified vs referent. The unspeakable Other is given, if not a name[4], a mode of existence, a location.

Has the post-structural dilemma been resolved? This is what Kristeva seems to suggest when she says that poetic language, in which the semiotic function is dominant, can 'make clear what is untenable in the symbolic, nominal, paternal function' (1977: 163). Kelly Oliver, editor of The Portable Kristeva, indisputably claims that it has:

By insisting that language expresses bodily drives through its semiotic element, Kristeva's articulation of the relationship between language and the body circumvents the traditional problems of representation. (1997: xvii)

However, there are two problems with the semiotic for our purposes. The first is specific to the written text: the relationship between the semiotic and the physicality of discourse is not made clear. On the one hand, the act of speaking is essential to the semiotic: 'with a material support such as the voice, this semiotic network gives 'music' to literature' (1974: 62) writes Kristeva, and elsewhere she says of the semiotic chora: 'dance, physical theatre or painting realise the semiotic chora the best, not words... a word becomes the flowing of a drive through its enunciation, and the only justification for a text is to give rise to this music of drives... .' (1977: 484) Literature, a written form and almost exclusively read with the eyes, not the voice, cannot benefit from this semiotic vocality. Nevertheless, Kristeva insists that the literary text exemplifies the semiotic: 'it is only in certain signifying practices, such as the text, that they [the processes and relations of the semiotic process] dominate the signifying process' (1974: 28); and 'It is poetic language that awakens our attention to this undecidable character of any so-called natural language' (1977: 161). How the literary text discharges drives (the semiotic function) without being enunciated — read aloud — is unclear; from this we must doubt whether the literary text can participate in the semiotic at all, much less be dominated by and awaken our attention to it.

The latter quotation on the 'undecidable' character of language raises the second and more general problem with the semiotic. Kristeva emphasises the 'heterogeneity to meaning and signification' (1977: 158) of the semiotic, and this heterogeneity runs a risk of meaninglessness which it does not always avoid. Something which can mean anything in fact means nothing, and thus the more heterogeneous to meaning, the more meaningless. Semiotic activity 'introduces wandering, fuzziness into language' (1977: 161); it 'departs from the signified' (1977: 158) and 'makes of what is known as 'literature' something other than knowledge' (ibid); its 'signifying disposition is not that of meaning or signification' (ibid). Kristeva foresees this problem and attempts to theoretically forestall it:

To retain the signifying process and avoid sinking into a limitless 'unsayable', and thus to posit the subject of a practice, the subject of poetic language clings to the help that fetishism offers him... the poetic function converges with fetishism but isn't identical to it... There is no text that would be deprived of meaning or signification if it were 'musicalised', on the contrary, musicalisation pluralises meaning, and therefore one can say that the text is not a fetish. Moreover, it's just like 'natural' language It is entirely different from a fetish, because it signifies The text signifies the unsignifying, it assumes, in the signifying practice, this function which ignores meaning and operates before or despite it. (1974: 64, my emphasis)

This paragraph advances through two sets of related contradictions, highlighted in colour. At the beginning, we are offered two theses: firstly, the text is a fetish (red); secondly, this will prevent us sinking into a limitless 'unsayable' (beige)[5]. The first thesis progresses into contradiction thus: the poetic function converges with fetishism; the text is not a fetish; the text is nothing like a fetish. The second thesis progresses first from avoiding the unsayable, into a musical pluralisation of meaning, into signification, into that which ignores meaning and operates before or despite it. As the text's fetishistic properties are increasingly denied, so we sink back into meaninglessness. [6]

The separation of language into a symbolic and a semiotic mode seemed to offer a resolution of the post-structural inability to approach the thing itself, by establishing a theory of language that included a material basis of signification and that could comprehend a concept that was before and in language. However, the binary Kristeva has constructed attributes all meaning and signification to the signifier, leaving the semiotic a locus of drives, non-sense, and meaninglessness. Rather than resolving our dilemma, the semiotic has only reinforced it: the disconnection of signifier and signified is strengthened by the opposition between symbolic and semiotic. What was previously an unbridgeable cleavage in language now runs through the core of our subjectivity. We have a mode of language which includes our physical reality, yes, but one that is irreconcilable with meaning, and as accessible as the unconscious. The claim that Kristeva resolves the traditional problems of representation is not borne out, for in the semiotic she has sidestepped representation altogether.

The problem of writing the thing itself which posited in Still Life is one which post-structural theory clarifies: it is a conflation of the signified (a concept) and the referent (a single material object). A referent is an extralinguistic material object, always specific, graspable, and has no place in language; a signified is a concept, is already language, and has no prelinguistic existence, so cannot be 'represented'. However, the same post-structural theory that argues for the logical impossibility of writing the thing itself repeatedly attempts nothing less, and finds that no amount of theoretical manoeuvres will turn this contradiction into mere paradox. There is, however, another solution posited, already hinted at in the prologue: 'Language was against him, for a start.' (SL, 2) The persistent contradiction in post-structuralism suggests that its premise is at fault: the premise that all conceptualisation is linguistic. If one cannot write the thing itself, one can perhaps nevertheless find other ways of representing it.

1 from 'Ulysses Gramophone: Hear Say Yes in Joyce' in Ulysse gramphone: Deux mots pour Joyce [1987] 

2 These are substantially explored in Part One of Kristeva's Révolution du langage poétique and to 'D'une identité l'autre' in Polylogue. 

3 The separation required by language is founded on anality: the first experience of separation, a crucial step on the path to language-acquisition is anal expulsion, a pleasurable and aggressive act. 

4 Kristeva notes the difficulty of naming anything presymbolic, and gives this as her reason for using Plato's term, the chora. She writes, 'Why then borrow an ontologised term in order to designate an articulation that antecedes positing? First, the Platonic term makes explicit an insurmountable problem for discourse: once it has been named, that functioning, even if it is presymbolic, is brought back into a symbolic position... ' (1997: 63; transl. Margaret Waller) 

5 Prior to this passage (Part One, section nine of Révolution du langage poétique), Kristeva offers the following reasoning for seeing poetic language as a fetish: the semiotic disrupts and attacks the symbolic order, and the poetry that does this simultaneously establishes itself as a substitute object for the symbolic order. It attempts to compensate for the corruption of the symbolic order by eroticising the speaking organs through the semiotic. 

6 I later extended this idea in Contradictions and the way of the Creator, filching the same textual analysis and finding further examples of similar contradictions.