The Meaninglessness of Meaning

Michael Wood

  • The Grain of the Voice: Interviews 1962-1980 by Roland Barthes, translated by Linda Coverdale
    Cape, 368 pp, £25.00, October 1985, ISBN 0 224 02302 0
  • Writing Degree Zero and Elements of Semiology by Roland Barthes, translated by Annette Lavers and Colin Smith
    Cape, 172 pp, £8.95, September 1984, ISBN 0 224 02267 9
  • The Fashion System by Roland Barthes, translated by Matthew Ward and Richard Howard
    Cape, 303 pp, £15.00, March 1985, ISBN 0 224 02984 3
  • The Responsibility of Forms: Critical Essays on Music, Art and Representation by Roland Barthes, translated by Richard Howard
    Blackwell, 312 pp, £19.50, January 1986, ISBN 0 631 14746 2
  • The Rustle of Language by Roland Barthes, translated by Richard Howard
    Blackwell, 373 pp, £27.50, May 1986, ISBN 0 631 14864 7
  • A Barthes Reader edited by Susan Sontag
    Cape, 495 pp, £15.00, September 1982, ISBN 0 224 02946 0
  • Barthes: Selected Writings edited by Susan Sontag
    Fontana, 495 pp, £4.95, August 1983, ISBN 0 00 636645 7
  • Roland Barthes: A Conservative Estimate by Philip Thody
    University of Chicago Press, 203 pp, £6.75, February 1984, ISBN 0 226 79513 6
  • Roland Barthes: Structuralism and After by Annette Lavers
    Methuen, 300 pp, £16.95, September 1982, ISBN 0 416 72380 2
  • Barthes by Jonathan Culler
    Fontana, 128 pp, £1.95, February 1983, ISBN 0 00 635974 4

A diary, Roland Barthes suggested, provokes in its writer not the tragic question, ‘Who am I?’ but the comic question: ‘Am I?’ This elegant and amused remark goes some way towards explaining why Barthes, who reflected much on his life and published a book called Roland Barthes par Roland Barthes, should not have kept a diary. The comic question can’t be confronted, it can only be circumvented, played with, smiled at. Of course, if you didn’t think it was funny, or if you were sure of the answer, you could keep a diary: but you wouldn’t be Barthes.

A novel might be an attractive alternative, and later in his career Barthes flirted with the notion of what he called the novelistic, le romanesque. Camera Lucida, the last of his works to be published in his lifetime (he died in 1980, at the age of 65), is described by Annette Lavers as ‘his only novel, and a love story’. Opening Roland Barthes par Roland Barthes we read, inside the front cover, white on black in Barthes’s handwriting: Tout ceci doit être considéré comme dit par un personnage de roman. What follows is text and pictures, a fractured autobiography, a dictionary of personal themes, arranged alphabetically. A fiction? Not exactly. Barthes on the page is a character in the novel of his life, not because he is an invented or rearranged figure, but because the writer, caught in the act, is always someone else, a creature whose home is words. Interviewers like to plot Barthes’s career by means of labels, and the labels are useful enough as long as we don’t fall in love with them. First, there was the critic (Writing Degree Zero, On Racine, Critical Essays), then the mythologist (Mythologies, La Tour Eiffel), then the semiologist (Elements of Semiology, The Fashion System), then the literary theorist (S/Z, The Pleasure of the Text), then the writer (Empire of Signs, A Lover’s Discourse, Camera Lucida). The writer? Didn’t the other avatars write? Does writer here mean, in the grovelling Anglo-Saxon phrase, ‘creative writer’? Barthes never produced anything we would readily regard as creative writing, or what he or we would really call a novel, except in some stretch of metaphor. He was too keen on fragments, too suspicious of character, and of big stories. But he was always a writer in one sense – ‘someone’, as he said in Critique et Vérité, not yet translated, ‘for whom language is problematic, who experiences its depth, not its usefulness or its beauty’. ‘I have an illness,’ he wrote elsewhere, ‘I can see language.’ And he became a writer in another sense, converted writing –‘his one great subject’, as Susan Sontag says – into his central practice, an embodiment or evocation of his concerns rather than a description of them. He stopped, as he said, writing on subjects – ‘It is that “on” which bothers me’ – even on the subject of writing. His prose then became fluid and imaginative like the best criticism, un-paraphraseable, not detachable from its topic. The writer: not necessarily a novelist but someone who knows he is a character in a novel, that the way out of words is made of words.

There is a special difficulty in interviewing such a figure, a prescription for a misfire, and the printed record of such an interview must seem sorry and belated: the translated trace of a man who wasn’t there, All the more reason for not being too much of a purist about this line of thought, since the 39 interviews with Barthes collected in The Grain of the Voice, if they cannot deliver Barthes the writer, do conjure up a very engaging person: courteous, quick-witted, unemphatic, willing to keep talking until he is understood. Annette Lavers sneers justifiably at those who had to wait for Barthes’s later work to realise he was human, but there may be some stragglers still, and this book should help them. Barthes discusses his debts to Sartre and Brecht, and the linguist Benveniste, remembers his early bouts of tuberculosis as hints of what might have been a vocation. The sanatorium, he says, was ‘a form of culture, surely’, intensifying all friendships and producing in him ‘the strange feeling of being always five or six years younger than I really am’. His first published piece of writing was on Gide’s Journal, and, asked if he knew Gide, Barthes replies that he saw him only once, from a distance, at the Lutétia: ‘he was eating a pear and reading a book.’ What interested him about Gide? Barthes’s answer might be taken as a swift epitaph on himself, four brief sentences wonderfully afloat on all they don’t say: ‘He was a Protestant. He played the piano. He talked about desire. He wrote.’

An earlier, more militant remark is worth pondering too: ‘between jargon and platitudes, I prefer jargon.’ Of course we fervently hope that is not the choice, but if it were? ‘It’s shameful to judge someone on his vocabulary,’ Barthes adds. The dream of clarity (say) is itself an ideological dream, an issue confronted with notable honesty by Philip Thody in his book on Barthes, now reissued with an afterword. Having struggled throughout the earlier edition to translate Barthes into an English critical idiom, helped by much good will and intelligence, hampered by a bewildered sense of how much he was losing in the process, Thody now memorably says that Barthes has taught him to be ‘more suspicious of the idea of tolerance’ and less insistent on the idea that all philosophers ought to write like Gilbert Ryle. It’s not that tolerance is not a virtue, only that all kinds of things which are not virtues can hide in its skirts, and that tolerance itself may be indistinguishable from condescension.

Thody’s problem is Barthes’s Frenchness, or rather the Frenchness of his rhetoric. Thody’s very description of what he calls ‘a well-established French rhetorical tradition’ shows how far he is from its mood: ‘ideas are stated in what is sometimes rather an exaggerated form in order to produce more of an effect.’ ‘Sometimes’, ‘rather’ – the words mime a caution and a reasonableness the tradition wouldn’t give the time of day to. In much of Barthes’s writing, as in much of Foucault and Derrida, the effect is the idea, and the idea is extreme, exorbitant: it has no prior, sensible, ‘English’ form which could be worked up into an exaggeration. This is not to say that the writing doesn’t mean anything, or that form and content are identical in it, only that the rhetorical high wire has its own airy relation to the ground.

Annette Lavers, on the other hand, takes Barthes’s Frenchness, his location in what she calls ‘the Darwinian struggle’ for literary survival, as her point of departure for a substantial piece of intellectual history. ‘This is meant to be a book, not a textbook,’ she says sternly: but I don’t see why it can’t be both. It does helpfully identify a number of stars in the recent Paris sky, so that Lévi-Strauss, Lacan, Althusser, Foucault, Derrida, Kristeva, Blanchot, Greimas and others are all placed in a kind of orbit around Barthes. Lavers is not much taken with what she sees as the temporary mindlessness of Barthes’s next-to-last works, and she is more persuasive than anyone has ever been about the interest of The Fashion System. She argues that Barthes was more loyal to his early views than most accounts of his career suggest, although she also says that he ‘connived’ at his later, popular ‘humanisation’, because to be ‘human’ – as Brecht’s Mother Courage, say, is taken to be ‘human’ – is no longer to be political or troubling. So those who miss Barthes’s humanity are wrong, and those who insist on it are wrong too. It’s not easy to win in such a competition.

Jonathan Culler’s job is to portray him as a Modern Master – even Thody speaks of Barthes as a ‘major French writer’ without any sort of qualification – and he does this with subtlety and skill. There are many modes of mastery and Culler suggests Barthes possesses ‘a special sort ... suited to experimentation with the intelligibilities of our time’. This phrase, an adaptation of Barthes’s definition of criticism (construction de l’intelligible de notre temps), is not as unintelligible as it looks. Barthes’s continuing project is meaning, the way societies make meaning, the way meaning surrounds and imprisons us, like some Wittgensteinian fly-bottle. ‘The notion of meaning,’ Lavers remarks, ‘can be said to govern the whole of his thinking,’ and she shrewdly comments that his attacks on single meaning may well be attacks on ‘meaning altogether’. His works and conversations are full of eloquent and dramatic images: the temptation of meaning, a festival of meaning, the adventure of the intelligible, the itinerary of meaning, the dream of meaning. ‘The act of writing is racked by the need for meaning.’ Barthes likes Buñuel’s Exterminating Angel because it keeps suspending meaning, and worries about Robbe-Grillet’s films because they don’t: ‘In fact, Robbe-Grillet doesn’t kill meaning at all, he scrambles it; he thinks it’s enough to mix up a meaning for it to die. It takes more than that to kill meaning.’ Why should we want to kill meaning? Of course Barthes doesn’t really want to. Meaning both enchants and tires him, tires him because it enchants him so. He spent his life looking at our compulsion to make sense, and so his life at times looked to him like an endless spell of sentry duty. He was attracted by the thought of an ‘exemption from meaning’, as one is exempt from military service, and he found in Japan a world of empty meanings – that is, a world full of meanings he didn’t understand, that ritually referred only to each other, or their place in a ceremony, not to a deep or ultimate meaning-beneath-the-meaning. He called this a ‘happiness of signs’, a holiday from meaning, and this is what his brief and elegant Empire of Signs celebrates.

For Culler, S/Z is Barthes’s summa: ‘a compendium of his views on literature and a meeting ground for projects often held to be contradictory ... The fact that S/Z has been seen as an extreme example of both structuralism and post-structuralism suggests that we ought to regard this distinction with suspicion.’ Culler goes on to articulate this sensible suspicion succinctly. Structuralism dreams of science and general theories, post-structuralism seeks to expose the flimsy, compromised nature of all structures of thought: but in practice, as Culler says, structuralism tends to concentrate on deviant or exceptional cases, so that the general theory becomes a ‘methodological horizon’ rather than a goal, while there is a strong universalising streak in post-structuralism’s insistence on putting everything into question. Culler’s scepticism deserts him when he moves into the higher theory and argues that boredom ‘is a major theoretical category with a role in any theory of reading’. This is a view which is likely to win theory only some very sleepy friends.

Is S/Z Barthes’s summa? It is, if the Barthes you cling to is the theorist (Thody leans distinctly to the mythologist, Lavers slightly to the semiologist). The trouble may lie with the hefty idea of the summa itself, the notion that that is what a master must leave. Barthes is a real master, but a slight one, and what makes him a master is not some single great work, and not one or more of his separate avatars, but the multiple career: the life of the mind lived with energy and variety and grace, leaving traces in texts. This, effectively, is the Barthes that Culler’s book ends with, and when Barthes in an interview calls himself a philosopher this is what he means. He is not identifying an ambition or a role or a job (as a writer he only dabbles in philosophy), but confessing an addiction to the charms of thought.

Barthes’s death has not stopped the flow of his publications, either in English or in French. His first book, Writing Degree Zero, with Elements of Semiology, was translated by Annette Lavers and Colin Smith in 1967, and this volume is now reissued. The Fashion System appears in English for the first time, and Barthes’s solemnity here looks more and more like a deadpan act, or a feat of self-discipline, or perhaps both: Buster Keaton at the semiologists’ circus. There are two new collections of essays by Barthes, first published in French in 1982 and 1984, and both now translated; and there is Susan Sontag’s anthology, which first saw life as A Barthes Reader in 1982 and returned as Barthes: Selected Writings.

In Writing Degree Zero, as the title suggests, Barthes begins to juggle with one of his magical words. ‘Writing’, he says, names a zone of language which has neither the broadness of general usage nor the quirkiness of personal style. It involves the choice, not of what to say or how to say it, but of how to signal one’s relation to literature. Communist fiction, Barthes thinks, is very literary, full of fussy metaphors; so-called realist writing is more literary still, ‘loaded with the most spectacular signs of fabrication’. The French revolutionaries of the 1790s used obscenities to point to the politics of what they wrote; Camus emptied out language into an affect-less neutrality, which unfortunately became just another literary gesture almost before you could say l’étranger. Queneau and Céline seem to abandon literature for speech, and if this too, in a book, must be a form of literature, at least it no longer looks like a separate, high-class cantonment: ‘it really represents the writer’s descent into the sticky opacity of the condition he is describing.’

Barthes overworks the pathos of this situation (‘literature is openly reduced to the problematics of language; and indeed, that is all it can now be’), and is inclined to use the word ‘tragic’ when he means ‘compromised’ or ‘troubled’. ‘Revolution must of necessity borrow, from what it wants to destroy, the very image of what it wants to possess.’ There is no doubt that revolutions have done this, or that it is quite hard to see what else they could do. But Barthes’s formulation turns this dismal observation into a comfortable law, and we seem, in such a sentence, already to hear the elegantly despairing voices of Jacques Derrida and Paul de Man. Some of Barthes’s other generalities are rather glib too (‘a modern masterpiece is impossible,’ ‘order ... is always a murder in intention’), and his eager, stabs at historians’ history (‘Now the 1850s bring the concurrence of three new and important facts in History: the demographic expansion in Europe, the replacement of textiles by heavy industry ..., the scission ... of French society into three mutual hostile classes’) are likely to make us smile. Nevertheless, Writing Degree Zero remains a vivid and fast-moving work, full of delicate ironies and marks of affection, and it can still surprise us. On this reading, for example, I was much taken with Barthes’s evident fondness for what he attacks. The classical age (which in Barthes’s sweeping view extends to 1848) is repressive and deceitful, the playground of the old bourgeois enemy (Barthes’s bourgeoisie includes the aristocracy), but it is also a sort of structuralist paradise, a world of relations not objects, provoking ‘euphoria’ and ‘delight’.

And of course his overall aim in Writing Degree Zero, the celebration of an extraordinary marriage between Sartre and formalism, is brilliantly achieved. For Barthes, a writer commits himself through his formal choices, through his tone, for instance, rather than his content, ‘making form a kind of behaviour and giving rise to an ethics of writing’. This argument frees the writer from the dutiful Philistinism which has characterised so much working and thinking on the left, but at the same time it makes all formal matters available for ethical scrutiny, takes them out of the aesthetic hothouse where they are usually thought to flourish. ‘Writing,’ he says, ‘is thus essentially the morality of form’; and the morality of form, in one shape or another, preoccupied him for the rest of his life.

‘Often,’ he later wrote of himself, ‘he felt stupid: it was that he had only a moral intelligence (that is, neither scientific, nor political, nor practical, nor philosophical etc).’ He said wryly that he raided technical vocabularies for metaphors and comparisons, so that he sounded vaguely like a linguist, or a theoretician, when he was performing as an artist – trying the dials of the radio of language, to use his own figure. Jonathan Culler nicely speaks of Barthes’s ‘piratical approach to linguistics’. I notice now how often ‘algebra’ and ‘algebraic’ come up in Writing Degree Zero, meaning a set of clean and slender signs, unhampered by muddy emotions, or the need to represent what they indicate. ‘Moral’, of course, in this context needs the expanded sense it doesn’t always have in English. Barthes is a moralist, as John Sturrock says in Structuralism and Since, because ‘moral passions and distinctions excite him,’ and his apparent flightiness is often a form of scruple, a refusal to lay down a badgering law. Sometimes it is just flightiness.

Elements of Semiology both picks up Saussure’s challenge regarding a science of signs, and inverts its priorities. For Barthes, linguistics is not, as it became for Saussure, a branch of the projected discipline of semiology. On the contrary, semiology is a branch of linguistics, since ‘every semiologieal system has its linguistic admixture.’ This is not a metaphor. Barthes is saying that language creeps into every signifying system which advances beyond the most primitive stages. Traffic lights are just lights, but clothes are talked about as well as worn. And even with the traffic lights, some of the associations of the words ‘red’ and ‘green’ may play a role. This is a risky claim, and heresy for real semiologists; and I can’t see myself that a linguistic admixture would necessarily imply a supremacy for linguistics. But the point is worth arguing about, and it is Barthes’s main personal point in this book, which for the rest is a straightforward primer, a map of the ground rather than an act of exploration. We may note that Barthes here refers to a zero degree of meaning in language, which is not a total absence of meaning, but ‘a significant absence’. Meaning haunts us even when it’s gone: it means absence.

Barthes’s act of exploration takes place in The Fashion System, an extraordinarily patient and technical look, not at clothes on people or at clothes in magazines, but at the language magazines use about clothes. And not a lot of magazines at that: only Elle and La Jardin des Modes, with a few glances at Vogue and L’Echo de la Mode, for the year 1958 to 1959, from June to June. Each chapter has a fragile or funny epigraph (‘Gauze, organza, voile and cotton muslin, summer is here,’ ‘The twin set makes a noted appearance’) suspended over tough talk about utterances, signifiers, syntagms, functions, variants and taxonomies. Barthes must have intended this lightly ironic effect: the truly shocking thing to do with fashion language was to take it absolutely seriously. He thought the work was dated when he published it in 1967, and it is dated for another reason now: the promises of semiology have either been kept or forgotten, and either way have faded as a context. Even so the book has a lot to tell us about the way we manufacture and consume meaning – ‘it is not the object but the name that creates desire; it is not the dream but the meaning that sells’ – and its diligent method, beyond the irony, allows us to see in great detail how names and meaning work.

Fashion, like literature, constructs a world out of signs, and one of the polemical points of the book is to make us wonder what the difference is. The world of literature is deeper, more durable? That is to reach beyond the signs, rather like deciding a football match on the crowd’s behaviour. Wouldn’t it help to know how these worlds are constructed? Fashion language for Barthes is a kind of laboratory, a cleared space, where signs are legible because they are sterilised and empty. My own scepticism concerns, not the project or the method in themselves, but their application: an empty sign is not just like a full sign, only empty. The emptiness/fullness makes all the difference. Or if not all the difference – mustn’t be too theological in these suspicious times – enough of a difference to alter the nature of the questions we want to ask.

Writing as form, form as morality. These ways of seeing, as well as other old habits (slight touches of pedantry, an acute fashion-consciousness, a tendency to think of the Rue Jacob as the world), persist in the two new collections of Barthes’s work, The Responsibility of Forms (a translation of L’Obvie et l’Obtus), and The Rustle of Language (a translation of Le Bruissement de la Langue). These books are billed (in French) as Essais Critiques III and IV (I was Essais Critiques, 1964, II was eight essays added to a reprint of Writing Degree Zero in 1972). Ten of the pieces these two books collect have already appeared in English in Stephen Heath’s valuable selection Image-Music-Text, and three are also to be found in Sontag’s anthology. An editor’s note to The Responsibility of Forms manages to miscount these previously translated items, and to turn Heath’s text into a test. It also contains a curious defence of this duplication. To subtract these items would be to distort the structure of the book. This structure respects ‘the integral architecture of R.B.’s work’. But the structure isn’t Barthes’s: he published the pieces all over the place, and what looks like respect is really the editor patting himself on the back – the architecture is all his.

Barthes’s worry about adjectives continues in The Responsibility of Forms. Is there nothing else, he moans, are we condemned either to the predictable or the ineffable? He insists on the materiality of the voice (‘the body speaking its mother tongue’) and of rhythm (‘I hear what beats in the body, what beats the body, or better: I hear this body that beats’). These essays concentrate on music (Schumann, Romantic lieder, the difficulty of describing sound, playing music as distinct from listening to it) and on what his editor calls ‘writing the visible’ – that is, photography, painting, cinema, theatre, design. There are pieces on Eisenstein, advertising, Greek theatre, Réquichot, Arcimboldo, Erté. Perhaps the most interesting, and certainly the most delicate essay in the book discusses the graphic work of Cy Twombly, those scribbles and smudges and absences which suggest an art which cannot come into being without betraying its finest elements. It can scarcely be collected and therefore, Barthes says, ‘blows up the Museum’. This art is not an ‘act’, Barthes says, because an act is transitive, seeks a result. What Twombly offers is a gesture, the sketch of the trace of an act, ‘the indeterminate and inexhaustible total of reasons, pulsions, indolences which surround the action with an atmosphere (in the astronomical sense of the word)’. This art is thrown away, a form of negligence, and brings us, Barthes believes, close to the ‘truth of things’. ‘Take an ordinary object: it is not its new, virgin state which best accounts for its essence, but its worn, lopsided, soiled, somewhat forsaken condition.’ Barthes, like Nabokov, has a tenderness for whatever is neglected. Not that he collects the items individually. He defends them generally, presents their apologies, would like to save them from the depredations he knows they can’t escape. One feels Barthes himself would like to be remembered as a sort of Twombly drawing: a smoky presence, unforgettable but insubstantial; nothing like a monument.

In the essay on Eisenstein, Barthes tries to elaborate, not really successfully, a theory of this tenderness. There is an ‘obvious’ sense to an image, he suggests, which includes both its literal and its symbolic meanings. Then there is a third, devious, elusive meaning, a semantic Harry Lime, which Barthes calls ‘obtuse’, accepting the pejorative overtone as part (but only part) of the implication. The ‘obtuse’, he says, is a scar on the ‘obvious’ meaning, and needless to say it is much more interesting. The strategy here is similar to the one Barthes adopts in Camera Lucida, where the studium of a photograph (what it sets out to show, what a competent reading would find) is distinguished from its punctum (its point or edge, what moves me about it, ‘bruises me’). In both cases, Barthes wants to shift an old arguing practice of his – what he once called banalité corigée, the orthodoxy uncovered or invented, and then attacked – into a more conciliatory mood. There will be what one is supposed to see (and does see), and there will be a supplement, an extravagance, which will cheer us up, saving us from the dreariness of doing what is expected of us. It strikes me not as obtuse but as perverse to try so hard to be eccentric, and no doubt a real eccentric wouldn’t have to try at all. In any event, Barthes’s theory doesn’t do justice to his own tenderness, which is less edgy and more agile than these laborious distinctions suggest.

The other collection, The Rustle of Language, returns to language and literature, and includes some brilliant work on Proust, a fine tribute to Jakobson (‘he converted prejudice into anachronism’), several gleeful references to the Marx Brothers’ Night at the Opera (‘an allegory of many textual problems’), some sharp and grateful pages on Brecht, said to be a practitioner not of subversion but of the shake-up, and a moving essay on Stendhal, which is Barthes’s last piece of writing – the second page of his clean copy was in his typewriter when he died. Stendhal cannot talk of what he loves, cannot evoke his beloved Italy. ‘Stendhal does not describe the thing, he does not even describe the effect,’ Barthes says, adapting Mallarmé (peindre non la chose mais l’effet qu’elle produit). ‘He simply says: there, there is an effect ... we must read Stendhal’s Italian discourse like a figured bass.’ Except – the exception is spectacular – for the Chartreuse de Parme, where the figured bass becomes a full score, and Stendhal writes at last like a character in a novel, encounters the full force and truthfulness of what Barthes calls the novel’s mode of lying, le mensonge romanesque.

This volume also contains, dotted among essays on other topics, some late and personal notes on death and aging. The middle of your life, Barthes says, arrives when you not only know you are mortal, but feel mortal. ‘This is not a natural feeling; the natural one is to believe yourself immortal; whence so many accidents due to carelessness.’ We discover that ‘death is real, and no longer merely dreadful.’ And the most violent death, Barthes adds, is not the death one inflicts or wants to inflict, but ‘the death that comes all by itself.

The rather odd ‘rustle’ in the title-essay of this collection is a translation of bruissement, ‘murmur’, ‘rumbling’, ‘hum’, ‘rustle’. Barthes opposes it to bredouillement, the retroactive stammer or mumble with which we try to correct speech that has gone wrong. Bruissement is the purr of an engine going right, and Barthes wonders if we can imagine such a noise for language, just ticking over, perfectly tuned. Could there really be such a sound, wouldn’t it always be mixed up in mistakes, and excesses of meaning? It would. ‘But what is impossible is not inconceivable.’ (And as we have recently learned, what is inconceivable is perfectly possible.) ‘The rustle of language forms a utopia.’ It is the utopia of the empty (not meaningless) sign, an echo of the dream Barthes expresses in his book on Japan: ‘The dream: to know a foreign (alien) language and yet not to understand it.’ We know there is a language, and that it is working. But it is not working for us: not harassing us with its solicitations and stereotypes.

In this context it may help to look at what Sontag calls Barthes’s ‘instantly notorious hyperbole’, his assertion that language is ‘quite simply fascist’. We shouldn’t see this as a slip, or even as just the waving of a red rag to an audience of academic bulls. Nor can we really say, as Phlip Thody wants to, that ‘when looked at in Barthes’s own terms’, the remark ‘makes perfectly good sense’. The remark is crazy in any terms, but its craziness, the extremity of its logic, is its point.

Here is the setting of the hyperbole. Barthes speaks of power and its many masks and faces, and says: ‘I call the discourse of power any discourse which engenders blame, hence guilt, in its recipients.’ We can agree that all language does this, if we are sensitive enough to the way it bends and breaks things in order to name them, but happily not all of us are so sensitive (or we are not all so sensitive all of the time). ‘We do not see the power which is in speech,’ Barthes continues, ‘because we forget that all speech is a classification, and that all classifications are oppressive.’ Speech can’t get along without classifications – but is that all speech is? – and any classification may be oppressive. Some, like race and gender, immediately and alarmingly are. But language? Won’t it depend on who is doing the classifying, and on what violence is offered to the classified world? Barthes will have none of this relativism. He is out for mischief and swashbuckling, not a parliamentary debate. Language is fascist, he says, not because of what it denies us but because of what it forces upon us: ‘for fascism does not prevent speech, it compels speech ... Speech is immediately assertive: negation, doubt, possibility, the suspension of judgment require special mechanisms which are themselves caught up in a play of linguistic masks ...’

If we imagine a disaster, Barthes writes elsewhere, we are on the way to accepting it: ‘to utter it is to assert it (again the fascism of language).’ The idea (without the political colouring) already appears in The Fashion System, where Barthes points to the ‘astounding phenomenon that language does not allow for the distinction between the simple utterance of a thing and the affirmation that it exists,’ and asks: ‘What new Borges among us will imagine a language in which to say things would be by rights to negate them, and in which an affirmative particle would have to be added in order to make them exist?’ This seems to be the old philosophical problem of the bald king of France, and surely represents one of the lures of language rather than a fact about it. For some reason we long to believe that the possession of a name confers existence on a thing, when everything we know about language and lying ought to prompt scepticism. Utterances don’t have to be assertions, and often aren’t. But it is true that our speech contains plenty of implicit assertions, even when we imagine we are guessing or soothing or asking questions. And it is true that much speech rests on concealed power relations. Think of all the suggestions and recommendations which can’t really be refused, all the command performances we benevolently require of our neighbours and children and friends. ‘Don’t you think you ought to?’ ‘If I were you, I’d ...’ We have the freedom Barthes denies, but we have less of it than we think, and his hyperbole points sharply to the bullying aspects of a community we like to think peaceable, to the clout behind the consensus. Asked by Bernard-Henri Lévy if he rejects power, Barthes says he is sensitive to its ubiquity and its endurance. ‘It never gets tired, it goes on and on, like a calendar.’ Is power itself fascist? No but ‘fascism is the constant temptation of power, its natural element, what comes in through the back door after it has been tossed out the front.’ A political activist could not talk in this way, but then a political activist might not dream of freedom as Barthes does: ‘If we call freedom not only the capacity to escape power but also and especially the capacity to subjugate no one, then freedom can exist only outside language. Unfortunately, human language has no exterior: there is no exit.’ If we were mystics or supermen, Barthes goes on, there would be an exit, but since we are not, our only option is to cheat speech, or to cheat with speech. This subversion of language under the very nose of the tyrant, this dragging of words towards their impossible freedom, is what he used to call ‘writing’, and now calls ‘literature’.